For cgroup writeback support, all bdi-wide operations should be
distributed to all its wb's (bdi_writeback's).

This patch updates laptop_mode_timer_fn() so that it invokes
wb_start_writeback() on all wb's rather than just the root one.  As
the intent is writing out all dirty data, there's no reason to split
the number of pages to write.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

bdi_start_writeback() is a thin wrapper on top of
__wb_start_writeback() which is used only by laptop_mode_timer_fn().
This patches removes bdi_start_writeback(), renames
__wb_start_writeback() to wb_start_writeback() and makes
laptop_mode_timer_fn() use it instead.

This doesn't cause any functional difference and will ease making
laptop_mode_timer_fn() cgroup writeback aware.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

bdi->min/max_ratio are user-configurable per-bdi knobs which regulate
dirty limit of each bdi.  For cgroup writeback, they need to be
further distributed across wb's (bdi_writeback's) belonging to the
configured bdi.

This patch introduces wb_min_max_ratio() which distributes
bdi->min/max_ratio according to a wb's proportion in the total active
bandwidth of its bdi.

v2: Update wb_min_max_ratio() to fix a bug where both min and max were
    assigned the min value and avoid calculations when possible.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

bdi_has_dirty_io() used to only reflect whether the root wb
(bdi_writeback) has dirty inodes.  For cgroup writeback support, it
needs to take all active wb's into account.  If any wb on the bdi has
dirty inodes, bdi_has_dirty_io() should return true.

To achieve that, as inode_wb_list_{move|del}_locked() now keep track
of the dirty state transition of each wb, the number of dirty wbs can
be counted in the bdi; however, bdi is already aggregating
wb->avg_write_bandwidth which can easily be guaranteed to be > 0 when
there are any dirty inodes by ensuring wb->avg_write_bandwidth can't
dip below 1.  bdi_has_dirty_io() can simply test whether
bdi->tot_write_bandwidth is zero or not.

While this bumps the value of wb->avg_write_bandwidth to one when it
used to be zero, this shouldn't cause any meaningful behavior
difference.

bdi_has_dirty_io() is made an inline function which tests whether
->tot_write_bandwidth is non-zero.  Also, WARN_ON_ONCE()'s on its
value are added to inode_wb_list_{move|del}_locked().
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

cgroup writeback support needs to keep track of the sum of
avg_write_bandwidth of all wb's (bdi_writeback's) with dirty inodes to
distribute write workload.  This patch adds bdi->tot_write_bandwidth
and updates inode_wb_list_move_locked(), inode_wb_list_del_locked()
and wb_update_write_bandwidth() to adjust it as wb's gain and lose
dirty inodes and its avg_write_bandwidth gets updated.

As the update events are not synchronized with each other,
bdi->tot_write_bandwidth is an atomic_long_t.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, wb_has_dirty_io() determines whether a wb (bdi_writeback)
has any dirty inode by testing all three IO lists on each invocation
without actively keeping track.  For cgroup writeback support, a
single bdi will host multiple wb's each of which will host dirty
inodes separately and we'll need to make bdi_has_dirty_io(), which
currently only represents the root wb, aggregate has_dirty_io from all
member wb's, which requires tracking transitions in has_dirty_io state
on each wb.

This patch introduces inode_wb_list_{move|del}_locked() to consolidate
IO list operations leaving queue_io() the only other function which
directly manipulates IO lists (via move_expired_inodes()).  All three
functions are updated to call wb_io_lists_[de]populated() which keep
track of whether the wb has dirty inodes or not and record it using
the new WB_has_dirty_io flag.  inode_wb_list_moved_locked()'s return
value indicates whether the wb had no dirty inodes before.

mark_inode_dirty() is restructured so that the return value of
inode_wb_list_move_locked() can be used for deciding whether to wake
up the wb.

While at it, change {bdi|wb}_has_dirty_io()'s return values to bool.
These functions were returning 0 and 1 before.  Also, add a comment
explaining the synchronization of wb_state flags.

v2: Updated to accommodate b_dirty_time.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

In several places, bdi_congested() and its wrappers are used to
determine whether more IOs should be issued.  With cgroup writeback
support, this question can't be answered solely based on the bdi
(backing_dev_info).  It's dependent on whether the filesystem and bdi
support cgroup writeback and the blkcg the inode is associated with.

This patch implements inode_congested() and its wrappers which take
@inode and determines the congestion state considering cgroup
writeback.  The new functions replace bdi_*congested() calls in places
where the query is about specific inode and task.

There are several filesystem users which also fit this criteria but
they should be updated when each filesystem implements cgroup
writeback support.

v2: Now that a given inode is associated with only one wb, congestion
    state can be determined independent from the asking task.  Drop
    @task.  Spotted by Vivek.  Also, converted to take @inode instead
    of @mapping and renamed to inode_congested().
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, all congestion functions take bdi (backing_dev_info) and
always operate on the root wb (bdi->wb) and the congestion state from
the block layer is propagated only for the root blkcg.  This patch
introduces {set|clear}_wb_congested() and wb_congested() which take a
bdi_writeback_congested and bdi_writeback respectively.  The bdi
counteparts are now wrappers invoking the wb based functions on
@bdi->wb.

While converting clear_bdi_congested() to clear_wb_congested(), the
local variable declaration order between @wqh and @bit is swapped for
cosmetic reason.

This patch just adds the new wb based functions.  The following
patches will apply them.

v2: Updated for bdi_writeback_congested.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, balance_dirty_pages() always work on bdi->wb.  This patch
updates it to work on the wb (bdi_writeback) matching memcg and blkcg
of the current task as that's what the inode is being dirtied against.

balance_dirty_pages_ratelimited() now pins the current wb and passes
it to balance_dirty_pages().

As no filesystem has FS_CGROUP_WRITEBACK yet, this doesn't lead to
visible behavior differences.

v2: Updated for per-inode wb association.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

Until now, all WB_* stats were accounted against the root wb
(bdi_writeback), now that multiple wb (bdi_writeback) support is in
place, let's attributes the stats to the respective per-cgroup wb's.

As no filesystem has FS_CGROUP_WRITEBACK yet, this doesn't lead to
visible behavior differences.

v2: Updated for per-inode wb association.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

For the planned cgroup writeback support, on each bdi
(backing_dev_info), each memcg will be served by a separate wb
(bdi_writeback).  This patch updates bdi so that a bdi can host
multiple wbs (bdi_writebacks).

On the default hierarchy, blkcg implicitly enables memcg.  This allows
using memcg's page ownership for attributing writeback IOs, and every
memcg - blkcg combination can be served by its own wb by assigning a
dedicated wb to each memcg.  This means that there may be multiple
wb's of a bdi mapped to the same blkcg.  As congested state is per
blkcg - bdi combination, those wb's should share the same congested
state.  This is achieved by tracking congested state via
bdi_writeback_congested structs which are keyed by blkcg.

bdi->wb remains unchanged and will keep serving the root cgroup.
cgwb's (cgroup wb's) for non-root cgroups are created on-demand or
looked up while dirtying an inode according to the memcg of the page
being dirtied or current task.  Each cgwb is indexed on bdi->cgwb_tree
by its memcg id.  Once an inode is associated with its wb, it can be
retrieved using inode_to_wb().

Currently, none of the filesystems has FS_CGROUP_WRITEBACK and all
pages will keep being associated with bdi->wb.

v3: inode_attach_wb() in account_page_dirtied() moved inside
    mapping_cap_account_dirty() block where it's known to be !NULL.
    Also, an unnecessary NULL check before kfree() removed.  Both
    detected by the kbuild bot.

v2: Updated so that wb association is per inode and wb is per memcg
    rather than blkcg.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: kbuild test robot <fengguang.wu@intel.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Jan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, a wb's (bdi_writeback) congestion state is carried in its
->state field; however, cgroup writeback support will require multiple
wb's sharing the same congestion state.  This patch separates out
congestion state into its own struct - struct bdi_writeback_congested.
A new field wb field, wb_congested, points to its associated congested
struct.  The default wb, bdi->wb, always points to bdi->wb_congested.

While this patch adds a layer of indirection, it doesn't introduce any
behavior changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@fb.com>

wb_init() currently always uses GFP_KERNEL but the planned cgroup
writeback support needs using other allocation masks.  Add @gfp to
wb_init().

This patch doesn't introduce any behavior changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: NJens Axboe <axboe@fb.com>

Now that bdi definitions are moved to backing-dev-defs.h,
backing-dev.h can include blkdev.h and inline inode_to_bdi() without
worrying about introducing circular include dependency.  The function
gets called from hot paths and fairly trivial.

This patch makes inode_to_bdi() and sb_is_blkdev_sb() that the
function calls inline.  blockdev_superblock and noop_backing_dev_info
are EXPORT_GPL'd to allow the inline functions to be used from
modules.

While at it, make sb_is_blkdev_sb() return bool instead of int.

v2: Fixed typo in description as suggested by Jan.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJens Axboe <axboe@kernel.dk>
Cc: Christoph Hellwig <hch@infradead.org>
Signed-off-by: NJens Axboe <axboe@fb.com>

With the planned cgroup writeback support, backing-dev related
declarations will be more widely used across block and cgroup;
unfortunately, including backing-dev.h from include/linux/blkdev.h
makes cyclic include dependency quite likely.

This patch separates out backing-dev-defs.h which only has the
essential definitions and updates blkdev.h to include it.  c files
which need access to more backing-dev details now include
backing-dev.h directly.  This takes backing-dev.h off the common
include dependency chain making it a lot easier to use it across block
and cgroup.

v2: fs/fat build failure fixed.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: NJens Axboe <axboe@fb.com>

Move wb_shutdown(), bdi_register(), bdi_register_dev(),
bdi_prune_sb(), bdi_remove_from_list() and bdi_unregister() so that
init / exit functions are grouped together.  This will make updating
init / exit paths for cgroup writeback support easier.

This is pure source file reorganization.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, a bdi (backing_dev_info) embeds single wb (bdi_writeback)
and the role of the separation is unclear.  For cgroup support for
writeback IOs, a bdi will be updated to host multiple wb's where each
wb serves writeback IOs of a different cgroup on the bdi.  To achieve
that, a wb should carry all states necessary for servicing writeback
IOs for a cgroup independently.

This patch moves bdi->wb_lock and ->worklist into wb.

* The lock protects bdi->worklist and bdi->wb.dwork scheduling.  While
  moving, rename it to wb->work_lock as wb->wb_lock is confusing.
  Also, move wb->dwork downwards so that it's colocated with the new
  ->work_lock and ->work_list fields.

* bdi_writeback_workfn()		-> wb_workfn()
  bdi_wakeup_thread_delayed(bdi)	-> wb_wakeup_delayed(wb)
  bdi_wakeup_thread(bdi)		-> wb_wakeup(wb)
  bdi_queue_work(bdi, ...)		-> wb_queue_work(wb, ...)
  __bdi_start_writeback(bdi, ...)	-> __wb_start_writeback(wb, ...)
  get_next_work_item(bdi)		-> get_next_work_item(wb)

* bdi_wb_shutdown() is renamed to wb_shutdown() and now takes @wb.
  The function contained parts which belong to the containing bdi
  rather than the wb itself - testing cap_writeback_dirty and
  bdi_remove_from_list() invocation.  Those are moved to
  bdi_unregister().

* bdi_wb_{init|exit}() are renamed to wb_{init|exit}().
  Initializations of the moved bdi->wb_lock and ->work_list are
  relocated from bdi_init() to wb_init().

* As there's still only one bdi_writeback per backing_dev_info, all
  uses of bdi->state are mechanically replaced with bdi->wb.state
  introducing no behavior changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Writeback operations will now be per wb (bdi_writeback) instead of
bdi.  Replace the relevant bdi references in symbol names and comments
with wb.  This patch is purely cosmetic and doesn't make any
functional changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, a bdi (backing_dev_info) embeds single wb (bdi_writeback)
and the role of the separation is unclear.  For cgroup support for
writeback IOs, a bdi will be updated to host multiple wb's where each
wb serves writeback IOs of a different cgroup on the bdi.  To achieve
that, a wb should carry all states necessary for servicing writeback
IOs for a cgroup independently.

This patch moves bandwidth related fields from backing_dev_info into
bdi_writeback.

* The moved fields are: bw_time_stamp, dirtied_stamp, written_stamp,
  write_bandwidth, avg_write_bandwidth, dirty_ratelimit,
  balanced_dirty_ratelimit, completions and dirty_exceeded.

* writeback_chunk_size() and over_bground_thresh() now take @wb
  instead of @bdi.

* bdi_writeout_fraction(bdi, ...)	-> wb_writeout_fraction(wb, ...)
  bdi_dirty_limit(bdi, ...)		-> wb_dirty_limit(wb, ...)
  bdi_position_ration(bdi, ...)		-> wb_position_ratio(wb, ...)
  bdi_update_writebandwidth(bdi, ...)	-> wb_update_write_bandwidth(wb, ...)
  [__]bdi_update_bandwidth(bdi, ...)	-> [__]wb_update_bandwidth(wb, ...)
  bdi_{max|min}_pause(bdi, ...)		-> wb_{max|min}_pause(wb, ...)
  bdi_dirty_limits(bdi, ...)		-> wb_dirty_limits(wb, ...)

* Init/exits of the relocated fields are moved to bdi_wb_init/exit()
  respectively.  Note that explicit zeroing is dropped in the process
  as wb's are cleared in entirety anyway.

* As there's still only one bdi_writeback per backing_dev_info, all
  uses of bdi->stat[] are mechanically replaced with bdi->wb.stat[]
  introducing no behavior changes.

v2: Typo in description fixed as suggested by Jan.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Jaegeuk Kim <jaegeuk@kernel.org>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, a bdi (backing_dev_info) embeds single wb (bdi_writeback)
and the role of the separation is unclear.  For cgroup support for
writeback IOs, a bdi will be updated to host multiple wb's where each
wb serves writeback IOs of a different cgroup on the bdi.  To achieve
that, a wb should carry all states necessary for servicing writeback
IOs for a cgroup independently.

This patch moves bdi->bdi_stat[] into wb.

* enum bdi_stat_item is renamed to wb_stat_item and the prefix of all
  enums is changed from BDI_ to WB_.

* BDI_STAT_BATCH() -> WB_STAT_BATCH()

* [__]{add|inc|dec|sum}_wb_stat(bdi, ...) -> [__]{add|inc}_wb_stat(wb, ...)

* bdi_stat[_error]() -> wb_stat[_error]()

* bdi_writeout_inc() -> wb_writeout_inc()

* stat init is moved to bdi_wb_init() and bdi_wb_exit() is added and
  frees stat.

* As there's still only one bdi_writeback per backing_dev_info, all
  uses of bdi->stat[] are mechanically replaced with bdi->wb.stat[]
  introducing no behavior changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: Trond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, a bdi (backing_dev_info) embeds single wb (bdi_writeback)
and the role of the separation is unclear.  For cgroup support for
writeback IOs, a bdi will be updated to host multiple wb's where each
wb serves writeback IOs of a different cgroup on the bdi.  To achieve
that, a wb should carry all states necessary for servicing writeback
IOs for a cgroup independently.

This patch moves bdi->state into wb.

* enum bdi_state is renamed to wb_state and the prefix of all enums is
  changed from BDI_ to WB_.

* Explicit zeroing of bdi->state is removed without adding zeoring of
  wb->state as the whole data structure is zeroed on init anyway.

* As there's still only one bdi_writeback per backing_dev_info, all
  uses of bdi->state are mechanically replaced with bdi->wb.state
  introducing no behavior changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reviewed-by: NJan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: drbd-dev@lists.linbit.com
Cc: Neil Brown <neilb@suse.de>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Implement mem_cgroup_css_from_page() which returns the
cgroup_subsys_state of the memcg associated with a given page on the
default hierarchy.  This will be used by cgroup writeback support.

This function assumes that page->mem_cgroup association doesn't change
until the page is released, which is true on the default hierarchy as
long as replace_page_cache_page() is not used.  As the only user of
replace_page_cache_page() is FUSE which won't support cgroup writeback
for the time being, this works for now, and replace_page_cache_page()
will soon be updated so that the invariant actually holds.

Note that the RCU protected page->mem_cgroup access is consistent with
other usages across memcg but ultimately incorrect.  These unlocked
accesses are missing required barriers.  page->mem_cgroup should be
made an RCU pointer and updated and accessed using RCU operations.

v4: Instead of triggering WARN, return the root css on the traditional
    hierarchies.  This makes the function a lot easier to deal with
    especially as there's no light way to synchronize against
    hierarchy rebinding.

v3: s/mem_cgroup_migrate()/mem_cgroup_css_from_page()/

v2: Trigger WARN if the function is used on the traditional
    hierarchies and add comment about the assumed invariant.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

Add global mem_cgroup_root_css which points to the root memcg css.
This will be used by cgroup writeback support.  If memcg is disabled,
it's defined as ERR_PTR(-EINVAL).
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Johannes Weiner <hannes@cmpxchg.org>
aCc: Michal Hocko <mhocko@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

When modifying PG_Dirty on cached file pages, update the new
MEM_CGROUP_STAT_DIRTY counter.  This is done in the same places where
global NR_FILE_DIRTY is managed.  The new memcg stat is visible in the
per memcg memory.stat cgroupfs file.  The most recent past attempt at
this was http://thread.gmane.org/gmane.linux.kernel.cgroups/8632

The new accounting supports future efforts to add per cgroup dirty
page throttling and writeback.  It also helps an administrator break
down a container's memory usage and provides evidence to understand
memcg oom kills (the new dirty count is included in memcg oom kill
messages).

The ability to move page accounting between memcg
(memory.move_charge_at_immigrate) makes this accounting more
complicated than the global counter.  The existing
mem_cgroup_{begin,end}_page_stat() lock is used to serialize move
accounting with stat updates.
Typical update operation:
	memcg = mem_cgroup_begin_page_stat(page)
	if (TestSetPageDirty()) {
		[...]
		mem_cgroup_update_page_stat(memcg)
	}
	mem_cgroup_end_page_stat(memcg)

Summary of mem_cgroup_end_page_stat() overhead:
- Without CONFIG_MEMCG it's a no-op
- With CONFIG_MEMCG and no inter memcg task movement, it's just
  rcu_read_lock()
- With CONFIG_MEMCG and inter memcg  task movement, it's
  rcu_read_lock() + spin_lock_irqsave()

A memcg parameter is added to several routines because their callers
now grab mem_cgroup_begin_page_stat() which returns the memcg later
needed by for mem_cgroup_update_page_stat().

Because mem_cgroup_begin_page_stat() may disable interrupts, some
adjustments are needed:
- move __mark_inode_dirty() from __set_page_dirty() to its caller.
  __mark_inode_dirty() locking does not want interrupts disabled.
- use spin_lock_irqsave(tree_lock) rather than spin_lock_irq() in
  __delete_from_page_cache(), replace_page_cache_page(),
  invalidate_complete_page2(), and __remove_mapping().

   text    data     bss      dec    hex filename
8925147 1774832 1785856 12485835 be84cb vmlinux-!CONFIG_MEMCG-before
8925339 1774832 1785856 12486027 be858b vmlinux-!CONFIG_MEMCG-after
                            +192 text bytes
8965977 1784992 1785856 12536825 bf4bf9 vmlinux-CONFIG_MEMCG-before
8966750 1784992 1785856 12537598 bf4efe vmlinux-CONFIG_MEMCG-after
                            +773 text bytes

Performance tests run on v4.0-rc1-36-g4f671fe2.  Lower is better for
all metrics, they're all wall clock or cycle counts.  The read and write
fault benchmarks just measure fault time, they do not include I/O time.

* CONFIG_MEMCG not set:
                            baseline                              patched
  kbuild                 1m25.030000(+-0.088% 3 samples)       1m25.426667(+-0.120% 3 samples)
  dd write 100 MiB          0.859211561 +-15.10%                  0.874162885 +-15.03%
  dd write 200 MiB          1.670653105 +-17.87%                  1.669384764 +-11.99%
  dd write 1000 MiB         8.434691190 +-14.15%                  8.474733215 +-14.77%
  read fault cycles       254.0(+-0.000% 10 samples)            253.0(+-0.000% 10 samples)
  write fault cycles     2021.2(+-3.070% 10 samples)           1984.5(+-1.036% 10 samples)

* CONFIG_MEMCG=y root_memcg:
                            baseline                              patched
  kbuild                 1m25.716667(+-0.105% 3 samples)       1m25.686667(+-0.153% 3 samples)
  dd write 100 MiB          0.855650830 +-14.90%                  0.887557919 +-14.90%
  dd write 200 MiB          1.688322953 +-12.72%                  1.667682724 +-13.33%
  dd write 1000 MiB         8.418601605 +-14.30%                  8.673532299 +-15.00%
  read fault cycles       266.0(+-0.000% 10 samples)            266.0(+-0.000% 10 samples)
  write fault cycles     2051.7(+-1.349% 10 samples)           2049.6(+-1.686% 10 samples)

* CONFIG_MEMCG=y non-root_memcg:
                            baseline                              patched
  kbuild                 1m26.120000(+-0.273% 3 samples)       1m25.763333(+-0.127% 3 samples)
  dd write 100 MiB          0.861723964 +-15.25%                  0.818129350 +-14.82%
  dd write 200 MiB          1.669887569 +-13.30%                  1.698645885 +-13.27%
  dd write 1000 MiB         8.383191730 +-14.65%                  8.351742280 +-14.52%
  read fault cycles       265.7(+-0.172% 10 samples)            267.0(+-0.000% 10 samples)
  write fault cycles     2070.6(+-1.512% 10 samples)           2084.4(+-2.148% 10 samples)

As expected anon page faults are not affected by this patch.

tj: Updated to apply on top of the recent cancel_dirty_page() changes.
Signed-off-by: NSha Zhengju <handai.szj@gmail.com>
Signed-off-by: NGreg Thelen <gthelen@google.com>
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@fb.com>

cancel_dirty_page() had some issues and b9ea2515 ("page_writeback:
clean up mess around cancel_dirty_page()") replaced it with
account_page_cleaned() which makes the caller responsible for clearing
the dirty bit; unfortunately, the planned changes for cgroup writeback
support requires synchronization between dirty bit manipulation and
stat updates.  While we can open-code such synchronization in each
account_page_cleaned() callsite, that's gonna be unnecessarily awkward
and verbose.

This patch revives cancel_dirty_page() but in a more restricted form.
All it does is TestClearPageDirty() followed by account_page_cleaned()
invocation if the page was dirty.  This helper covers all
account_page_cleaned() usages except for __delete_from_page_cache()
which is a special case anyway and left alone.  As this leaves no
module user for account_page_cleaned(), EXPORT_SYMBOL() is dropped
from it.

This patch just revives cancel_dirty_page() as a trivial wrapper to
replace equivalent usages and doesn't introduce any functional
changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: NJens Axboe <axboe@fb.com>

Stop abusing struct page functionality and the swap end_io handler, and
instead add a modified version of the blk-lib.c bio_batch helpers.

Also move the block I/O code into swap.c as they are directly tied into
each other.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NPavel Machek <pavel@ucw.cz>
Tested-by: NMing Lin <mlin@kernel.org>
Acked-by: NPavel Machek <pavel@ucw.cz>
Acked-by: NRafael J. Wysocki <rjw@rjwysocki.net>
Signed-off-by: NJens Axboe <axboe@fb.com>

Do not perform cond_resched() before the busy compaction loop in
__zs_compact(), because this loop does it when needed.
Signed-off-by: NSergey Senozhatsky <sergey.senozhatsky@gmail.com>
Acked-by: NMinchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

There is no point in overriding the size class below.  It causes fatal
corruption on the next chunk on the 3264-bytes size class, which is the
last size class that is not huge.

For example, if the requested size was exactly 3264 bytes, current
zsmalloc allocates and returns a chunk from the size class of 3264 bytes,
not 4096.  User access to this chunk may overwrite head of the next
adjacent chunk.

Here is the panic log captured when freelist was corrupted due to this:

    Kernel BUG at ffffffc00030659c [verbose debug info unavailable]
    Internal error: Oops - BUG: 96000006 [#1] PREEMPT SMP
    Modules linked in:
    exynos-snapshot: core register saved(CPU:5)
    CPUMERRSR: 0000000000000000, L2MERRSR: 0000000000000000
    exynos-snapshot: context saved(CPU:5)
    exynos-snapshot: item - log_kevents is disabled
    CPU: 5 PID: 898 Comm: kswapd0 Not tainted 3.10.61-4497415-eng #1
    task: ffffffc0b8783d80 ti: ffffffc0b71e8000 task.ti: ffffffc0b71e8000
    PC is at obj_idx_to_offset+0x0/0x1c
    LR is at obj_malloc+0x44/0xe8
    pc : [<ffffffc00030659c>] lr : [<ffffffc000306604>] pstate: a0000045
    sp : ffffffc0b71eb790
    x29: ffffffc0b71eb790 x28: ffffffc00204c000
    x27: 000000000001d96f x26: 0000000000000000
    x25: ffffffc098cc3500 x24: ffffffc0a13f2810
    x23: ffffffc098cc3501 x22: ffffffc0a13f2800
    x21: 000011e1a02006e3 x20: ffffffc0a13f2800
    x19: ffffffbc02a7e000 x18: 0000000000000000
    x17: 0000000000000000 x16: 0000000000000feb
    x15: 0000000000000000 x14: 00000000a01003e3
    x13: 0000000000000020 x12: fffffffffffffff0
    x11: ffffffc08b264000 x10: 00000000e3a01004
    x9 : ffffffc08b263fea x8 : ffffffc0b1e611c0
    x7 : ffffffc000307d24 x6 : 0000000000000000
    x5 : 0000000000000038 x4 : 000000000000011e
    x3 : ffffffbc00003e90 x2 : 0000000000000cc0
    x1 : 00000000d0100371 x0 : ffffffbc00003e90
Reported-by: NSooyong Suk <s.suk@samsung.com>
Signed-off-by: NHeesub Shin <heesub.shin@samsung.com>
Tested-by: NSooyong Suk <s.suk@samsung.com>
Acked-by: NMinchan Kim <minchan@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In putback_zspage, we don't need to insert a zspage into list of zspage
in size_class again to just fix fullness group. We could do directly
without reinsertion so we could save some instuctions.
Reported-by: NHeesub Shin <heesub.shin@samsung.com>
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Ganesh Mahendran <opensource.ganesh@gmail.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Juneho Choi <juno.choi@lge.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

A micro-optimization.  Avoid additional branching and reduce (a bit)
registry pressure (f.e.  s_off += size; d_off += size; may be calculated
twise: first for >= PAGE_SIZE check and later for offset update in "else"
clause).

scripts/bloat-o-meter shows some improvement

add/remove: 0/0 grow/shrink: 0/1 up/down: 0/-10 (-10)
function                          old     new   delta
zs_object_copy                    550     540     -10
Signed-off-by: NSergey Senozhatsky <sergey.senozhatsky@gmail.com>
Acked-by: NMinchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Do not synchronize rcu in zs_compact(). Neither zsmalloc not
zram use rcu.
Signed-off-by: NSergey Senozhatsky <sergey.senozhatsky@gmail.com>
Acked-by: NMinchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Signed-off-by: NYinghao Xie <yinghao.xie@sumsung.com>
Suggested-by: NMinchan Kim <minchan@kernel.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Create zsmalloc doc which explains design concept and stat information.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

During investigating compaction, fullness information of each class is
helpful for investigating how the compaction works well.  With that, we
could know how compaction works well more clear on each size class.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We store handle on header of each allocated object so it increases the
size of each object by sizeof(unsigned long).

If zram stores 4096 bytes to zsmalloc(ie, bad compression), zsmalloc needs
4104B-class to add handle.

However, 4104B-class has 1-pages_per_zspage so wasted size by internal
fragment is 8192 - 4104, which is terrible.

So this patch records the handle in page->private on such huge object(ie,
pages_per_zspage == 1 && maxobj_per_zspage == 1) instead of header of each
object so we could use 4096B-class, not 4104B-class.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Curretly, zsmalloc regards a zspage as ZS_ALMOST_EMPTY if the zspage has
under 1/4 used objects(ie, fullness_threshold_frac).  It could make result
in loose packing since zsmalloc migrates only ZS_ALMOST_EMPTY zspage out.

This patch changes the rule so that zsmalloc makes zspage which has above
3/4 used object ZS_ALMOST_FULL so it could make tight packing.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch provides core functions for migration of zsmalloc.  Migraion
policy is simple as follows.

for each size class {
        while {
                src_page = get zs_page from ZS_ALMOST_EMPTY
                if (!src_page)
                        break;
                dst_page = get zs_page from ZS_ALMOST_FULL
                if (!dst_page)
                        dst_page = get zs_page from ZS_ALMOST_EMPTY
                if (!dst_page)
                        break;
                migrate(from src_page, to dst_page);
        }
}

For migration, we need to identify which objects in zspage are allocated
to migrate them out.  We could know it by iterating of freed objects in a
zspage because first_page of zspage keeps free objects singly-linked list
but it's not efficient.  Instead, this patch adds a tag(ie,
OBJ_ALLOCATED_TAG) in header of each object(ie, handle) so we could check
whether the object is allocated easily.

This patch adds another status bit in handle to synchronize between user
access through zs_map_object and migration.  During migration, we cannot
move objects user are using due to data coherency between old object and
new object.

[akpm@linux-foundation.org: zsmalloc.c needs sched.h for cond_resched()]
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

In later patch, migration needs some part of functions in zs_malloc and
zs_free so this patch factor out them.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Recently, we started to use zram heavily and some of issues
popped.

1) external fragmentation

I got a report from Juneho Choi that fork failed although there are plenty
of free pages in the system.  His investigation revealed zram is one of
the culprit to make heavy fragmentation so there was no more contiguous
16K page for pgd to fork in the ARM.

2) non-movable pages

Other problem of zram now is that inherently, user want to use zram as
swap in small memory system so they use zRAM with CMA to use memory
efficiently.  However, unfortunately, it doesn't work well because zRAM
cannot use CMA's movable pages unless it doesn't support compaction.  I
got several reports about that OOM happened with zram although there are
lots of swap space and free space in CMA area.

3) internal fragmentation

zRAM has started support memory limitation feature to limit memory usage
and I sent a patchset(https://lkml.org/lkml/2014/9/21/148) for VM to be
harmonized with zram-swap to stop anonymous page reclaim if zram consumed
memory up to the limit although there are free space on the swap.  One
problem for that direction is zram has no way to know any hole in memory
space zsmalloc allocated by internal fragmentation so zram would regard
swap is full although there are free space in zsmalloc.  For solving the
issue, zram want to trigger compaction of zsmalloc before it decides full
or not.

This patchset is first step to support above issues.  For that, it adds
indirect layer between handle and object location and supports manual
compaction to solve 3th problem first of all.

After this patchset got merged, next step is to make VM aware of zsmalloc
compaction so that generic compaction will move zsmalloced-pages
automatically in runtime.

In my imaginary experiment(ie, high compress ratio data with heavy swap
in/out on 8G zram-swap), data is as follows,

Before =
zram allocated object :      60212066 bytes
zram total used:     140103680 bytes
ratio:         42.98 percent
MemFree:          840192 kB

Compaction

After =
frag ratio after compaction
zram allocated object :      60212066 bytes
zram total used:      76185600 bytes
ratio:         79.03 percent
MemFree:          901932 kB

Juneho reported below in his real platform with small aging.
So, I think the benefit would be bigger in real aging system
for a long time.

- frag_ratio increased 3% (ie, higher is better)
- memfree increased about 6MB
- In buddy info, Normal 2^3: 4, 2^2: 1: 2^1 increased, Highmem: 2^1 21 increased

frag ratio after swap fragment
used :        156677 kbytes
total:        166092 kbytes
frag_ratio :  94
meminfo before compaction
MemFree:           83724 kB
Node 0, zone   Normal  13642   1364     57     10     61     17      9      5      4      0      0
Node 0, zone  HighMem    425     29      1      0      0      0      0      0      0      0      0

num_migrated :  23630
compaction done

frag ratio after compaction
used :        156673 kbytes
total:        160564 kbytes
frag_ratio :  97
meminfo after compaction
MemFree:           89060 kB
Node 0, zone   Normal  14076   1544     67     14     61     17      9      5      4      0      0
Node 0, zone  HighMem    863     50      1      0      0      0      0      0      0      0      0

This patchset adds more logics(about 480 lines) in zsmalloc but when I
tested heavy swapin/out program, the regression for swapin/out speed is
marginal because most of overheads were caused by compress/decompress and
other MM reclaim stuff.

This patch (of 7):

Currently, handle of zsmalloc encodes object's location directly so it
makes support of migration hard.

This patch decouples handle and object via adding indirect layer.  For
that, it allocates handle dynamically and returns it to user.  The handle
is the address allocated by slab allocation so it's unique and we could
keep object's location in the memory space allocated for handle.

With it, we can change object's position without changing handle itself.
Signed-off-by: NMinchan Kim <minchan@kernel.org>
Cc: Juneho Choi <juno.choi@lge.com>
Cc: Gunho Lee <gunho.lee@lge.com>
Cc: Luigi Semenzato <semenzato@google.com>
Cc: Dan Streetman <ddstreet@ieee.org>
Cc: Seth Jennings <sjennings@variantweb.net>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Mel Gorman <mel@csn.ul.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

mm/compaction.c:250:13: warning: 'suitable_migration_target' defined but not used [-Wunused-function]
Reported-by: NFengguang Wu <fengguang.wu@gmail.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>