Much of the discussion about this has died down. There's been a
concern raised that we could generalize infrastructure across loop,
md, etc. This may be possible, in the future, but it isn't clear to me
how this would look like. I'm inclined to fix the existing issue with
loop devices now (this is a problem we hit at FB) and address
consolidation with other cases if and when those need to be addressed.
Note that patch 2 in this series (mm: support nesting
memalloc_use_memcg()) has also been submitted by Roman Guschchin to
the mm tree and can be seen here:
http://lkml.iu.edu/hypermail/linux/kernel/2008.2/09214.html - I
include it here for completeness, but this series works with either
version of the patch.
Changes since V7:
* Rebased against linus's branch
Changes since V6:
* Added separate spinlock for worker synchronization
* Minor style changes
Changes since V5:
* Fixed a missing css_put when failing to allocate a worker
* Minor style changes
Changes since V4:
Only patches 1 and 2 have changed.
* Fixed irq lock ordering bug
* Simplified loop detach
* Added support for nesting memalloc_use_memcg
Changes since V3:
* Fix race on loop device destruction and deferred worker cleanup
* Ensure charge on shmem_swapin_page works just like getpage
* Minor style changes
Changes since V2:
* Deferred destruction of workqueue items so in the common case there
is no allocation needed
Changes since V1:
* Split out and reordered patches so cgroup charging changes are
separate from kworker -> workqueue change
* Add mem_css to struct loop_cmd to simplify logic
The loop device runs all i/o to the backing file on a separate kworker
thread which results in all i/o being charged to the root cgroup. This
allows a loop device to be used to trivially bypass resource limits
and other policy. This patch series fixes this gap in accounting.
A simple script to demonstrate this behavior on cgroupv2 machine:
'''
#!/bin/bash
set -e
CGROUP=/sys/fs/cgroup/test.slice
LOOP_DEV=/dev/loop0
if [[ ! -d $CGROUP ]]
then
sudo mkdir $CGROUP
fi
grep oom_kill $CGROUP/memory.events
# Set a memory limit, write more than that limit to tmpfs -> OOM kill
sudo unshare -m bash -c "
echo \$\$ > $CGROUP/cgroup.procs;
echo 0 > $CGROUP/memory.swap.max;
echo 64M > $CGROUP/memory.max;
mount -t tmpfs -o size=512m tmpfs /tmp;
dd if=/dev/zero of=/tmp/file bs=1M count=256" || true
grep oom_kill $CGROUP/memory.events
# Set a memory limit, write more than that limit through loopback
# device -> no OOM kill
sudo unshare -m bash -c "
echo \$\$ > $CGROUP/cgroup.procs;
echo 0 > $CGROUP/memory.swap.max;
echo 64M > $CGROUP/memory.max;
mount -t tmpfs -o size=512m tmpfs /tmp;
truncate -s 512m /tmp/backing_file
losetup $LOOP_DEV /tmp/backing_file
dd if=/dev/zero of=$LOOP_DEV bs=1M count=256;
losetup -D $LOOP_DEV" || true
grep oom_kill $CGROUP/memory.events
'''
Naively charging cgroups could result in priority inversions through
the single kworker thread in the case where multiple cgroups are
reading/writing to the same loop device. This patch series does some
minor modification to the loop driver so that each cgroup can make
forward progress independently to avoid this inversion.
With this patch series applied, the above script triggers OOM kills
when writing through the loop device as expected.
Dan Schatzberg (3):
loop: Use worker per cgroup instead of kworker
mm: Charge active memcg when no mm is set
loop: Charge i/o to mem and blk cg
Johannes Weiner (1):
mm: support nesting memalloc_use_memcg()
drivers/block/loop.c | 248 ++++++++++++++++++++++-----
drivers/block/loop.h | 15 +-
fs/buffer.c | 6 +-
fs/notify/fanotify/fanotify.c | 5 +-
fs/notify/inotify/inotify_fsnotify.c | 5 +-
include/linux/memcontrol.h | 6 +
include/linux/sched/mm.h | 28 +--
kernel/cgroup/cgroup.c | 1 +
mm/memcontrol.c | 17 +-
mm/shmem.c | 4 +-
10 files changed, 253 insertions(+), 82 deletions(-)
--
2.24.1
The current code only associates with the existing blkcg when aio is
used to access the backing file. This patch covers all types of i/o to
the backing file and also associates the memcg so if the backing file is
on tmpfs, memory is charged appropriately.
This patch also exports cgroup_get_e_css so it can be used by the loop
module.
Signed-off-by: Dan Schatzberg <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
---
drivers/block/loop.c | 61 +++++++++++++++++++++++++-------------
drivers/block/loop.h | 3 +-
include/linux/memcontrol.h | 6 ++++
kernel/cgroup/cgroup.c | 1 +
mm/shmem.c | 1 -
5 files changed, 50 insertions(+), 22 deletions(-)
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index c8dae53f3914..102d236d1540 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -77,6 +77,7 @@
#include <linux/uio.h>
#include <linux/ioprio.h>
#include <linux/blk-cgroup.h>
+#include <linux/sched/mm.h>
#include "loop.h"
@@ -518,8 +519,6 @@ static void lo_rw_aio_complete(struct kiocb *iocb, long ret, long ret2)
{
struct loop_cmd *cmd = container_of(iocb, struct loop_cmd, iocb);
- if (cmd->css)
- css_put(cmd->css);
cmd->ret = ret;
lo_rw_aio_do_completion(cmd);
}
@@ -580,8 +579,6 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
cmd->iocb.ki_complete = lo_rw_aio_complete;
cmd->iocb.ki_flags = IOCB_DIRECT;
cmd->iocb.ki_ioprio = IOPRIO_PRIO_VALUE(IOPRIO_CLASS_NONE, 0);
- if (cmd->css)
- kthread_associate_blkcg(cmd->css);
if (rw == WRITE)
ret = call_write_iter(file, &cmd->iocb, &iter);
@@ -589,7 +586,6 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
ret = call_read_iter(file, &cmd->iocb, &iter);
lo_rw_aio_do_completion(cmd);
- kthread_associate_blkcg(NULL);
if (ret != -EIOCBQUEUED)
cmd->iocb.ki_complete(&cmd->iocb, ret, 0);
@@ -929,7 +925,7 @@ struct loop_worker {
struct list_head cmd_list;
struct list_head idle_list;
struct loop_device *lo;
- struct cgroup_subsys_state *css;
+ struct cgroup_subsys_state *blkcg_css;
unsigned long last_ran_at;
};
@@ -946,7 +942,7 @@ static void loop_queue_work(struct loop_device *lo, struct loop_cmd *cmd)
spin_lock_irq(&lo->lo_work_lock);
- if (!cmd->css)
+ if (!cmd->blkcg_css)
goto queue_work;
node = &lo->worker_tree.rb_node;
@@ -954,10 +950,10 @@ static void loop_queue_work(struct loop_device *lo, struct loop_cmd *cmd)
while (*node) {
parent = *node;
cur_worker = container_of(*node, struct loop_worker, rb_node);
- if (cur_worker->css == cmd->css) {
+ if (cur_worker->blkcg_css == cmd->blkcg_css) {
worker = cur_worker;
break;
- } else if ((long)cur_worker->css < (long)cmd->css) {
+ } else if ((long)cur_worker->blkcg_css < (long)cmd->blkcg_css) {
node = &(*node)->rb_left;
} else {
node = &(*node)->rb_right;
@@ -969,13 +965,18 @@ static void loop_queue_work(struct loop_device *lo, struct loop_cmd *cmd)
worker = kzalloc(sizeof(struct loop_worker), GFP_NOWAIT | __GFP_NOWARN);
/*
* In the event we cannot allocate a worker, just queue on the
- * rootcg worker
+ * rootcg worker and issue the I/O as the rootcg
*/
- if (!worker)
+ if (!worker) {
+ cmd->blkcg_css = NULL;
+ if (cmd->memcg_css)
+ css_put(cmd->memcg_css);
+ cmd->memcg_css = NULL;
goto queue_work;
+ }
- worker->css = cmd->css;
- css_get(worker->css);
+ worker->blkcg_css = cmd->blkcg_css;
+ css_get(worker->blkcg_css);
INIT_WORK(&worker->work, loop_workfn);
INIT_LIST_HEAD(&worker->cmd_list);
INIT_LIST_HEAD(&worker->idle_list);
@@ -1301,7 +1302,7 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
idle_list) {
list_del(&worker->idle_list);
rb_erase(&worker->rb_node, &lo->worker_tree);
- css_put(worker->css);
+ css_put(worker->blkcg_css);
kfree(worker);
}
spin_unlock_irq(&lo->lo_work_lock);
@@ -2102,13 +2103,18 @@ static blk_status_t loop_queue_rq(struct blk_mq_hw_ctx *hctx,
}
/* always use the first bio's css */
+ cmd->blkcg_css = NULL;
+ cmd->memcg_css = NULL;
#ifdef CONFIG_BLK_CGROUP
- if (cmd->use_aio && rq->bio && rq->bio->bi_blkg) {
- cmd->css = &bio_blkcg(rq->bio)->css;
- css_get(cmd->css);
- } else
+ if (rq->bio && rq->bio->bi_blkg) {
+ cmd->blkcg_css = &bio_blkcg(rq->bio)->css;
+#ifdef CONFIG_MEMCG
+ cmd->memcg_css =
+ cgroup_get_e_css(cmd->blkcg_css->cgroup,
+ &memory_cgrp_subsys);
+#endif
+ }
#endif
- cmd->css = NULL;
loop_queue_work(lo, cmd);
return BLK_STS_OK;
@@ -2120,13 +2126,28 @@ static void loop_handle_cmd(struct loop_cmd *cmd)
const bool write = op_is_write(req_op(rq));
struct loop_device *lo = rq->q->queuedata;
int ret = 0;
+ struct mem_cgroup *old_memcg = NULL;
if (write && (lo->lo_flags & LO_FLAGS_READ_ONLY)) {
ret = -EIO;
goto failed;
}
+ if (cmd->blkcg_css)
+ kthread_associate_blkcg(cmd->blkcg_css);
+ if (cmd->memcg_css)
+ old_memcg = memalloc_use_memcg(
+ mem_cgroup_from_css(cmd->memcg_css));
+
ret = do_req_filebacked(lo, rq);
+
+ if (cmd->blkcg_css)
+ kthread_associate_blkcg(NULL);
+
+ if (cmd->memcg_css) {
+ memalloc_use_memcg(old_memcg);
+ css_put(cmd->memcg_css);
+ }
failed:
/* complete non-aio request */
if (!cmd->use_aio || ret) {
@@ -2205,7 +2226,7 @@ static void loop_free_idle_workers(struct timer_list *timer)
break;
list_del(&worker->idle_list);
rb_erase(&worker->rb_node, &lo->worker_tree);
- css_put(worker->css);
+ css_put(worker->blkcg_css);
kfree(worker);
}
if (!list_empty(&lo->idle_worker_list))
diff --git a/drivers/block/loop.h b/drivers/block/loop.h
index 0162b55a68e1..4d6886d9855a 100644
--- a/drivers/block/loop.h
+++ b/drivers/block/loop.h
@@ -75,7 +75,8 @@ struct loop_cmd {
long ret;
struct kiocb iocb;
struct bio_vec *bvec;
- struct cgroup_subsys_state *css;
+ struct cgroup_subsys_state *blkcg_css;
+ struct cgroup_subsys_state *memcg_css;
};
/* Support for loadable transfer modules */
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index d0b036123c6a..fceac9f66d96 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -1031,6 +1031,12 @@ static inline struct mem_cgroup *get_mem_cgroup_from_page(struct page *page)
return NULL;
}
+static inline
+struct mem_cgroup *mem_cgroup_from_css(struct cgroup_subsys_state *css)
+{
+ return NULL;
+}
+
static inline void mem_cgroup_put(struct mem_cgroup *memcg)
{
}
diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index dd247747ec14..16d059a89a68 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -580,6 +580,7 @@ struct cgroup_subsys_state *cgroup_get_e_css(struct cgroup *cgrp,
rcu_read_unlock();
return css;
}
+EXPORT_SYMBOL_GPL(cgroup_get_e_css);
static void cgroup_get_live(struct cgroup *cgrp)
{
diff --git a/mm/shmem.c b/mm/shmem.c
index 77c908730be4..1139f52ac4ee 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -1696,7 +1696,6 @@ static int shmem_swapin_page(struct inode *inode, pgoff_t index,
struct address_space *mapping = inode->i_mapping;
struct shmem_inode_info *info = SHMEM_I(inode);
struct mm_struct *charge_mm = vma ? vma->vm_mm : NULL;
- struct mem_cgroup *memcg;
struct page *page;
swp_entry_t swap;
int error;
--
2.24.1
Existing uses of loop device may have multiple cgroups reading/writing
to the same device. Simply charging resources for I/O to the backing
file could result in priority inversion where one cgroup gets
synchronously blocked, holding up all other I/O to the loop device.
In order to avoid this priority inversion, we use a single workqueue
where each work item is a "struct loop_worker" which contains a queue of
struct loop_cmds to issue. The loop device maintains a tree mapping blk
css_id -> loop_worker. This allows each cgroup to independently make
forward progress issuing I/O to the backing file.
There is also a single queue for I/O associated with the rootcg which
can be used in cases of extreme memory shortage where we cannot allocate
a loop_worker.
The locking for the tree and queues is fairly heavy handed - we acquire
a per-loop-device spinlock any time either is accessed. The existing
implementation serializes all I/O through a single thread anyways, so I
don't believe this is any worse.
Signed-off-by: Dan Schatzberg <[email protected]>
---
drivers/block/loop.c | 207 ++++++++++++++++++++++++++++++++++++-------
drivers/block/loop.h | 12 ++-
2 files changed, 182 insertions(+), 37 deletions(-)
diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 2f137d6ce169..c8dae53f3914 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -70,7 +70,6 @@
#include <linux/writeback.h>
#include <linux/completion.h>
#include <linux/highmem.h>
-#include <linux/kthread.h>
#include <linux/splice.h>
#include <linux/sysfs.h>
#include <linux/miscdevice.h>
@@ -83,6 +82,8 @@
#include <linux/uaccess.h>
+#define LOOP_IDLE_WORKER_TIMEOUT (60 * HZ)
+
static DEFINE_IDR(loop_index_idr);
static DEFINE_MUTEX(loop_ctl_mutex);
@@ -922,27 +923,83 @@ static void loop_config_discard(struct loop_device *lo)
blk_queue_flag_clear(QUEUE_FLAG_DISCARD, q);
}
-static void loop_unprepare_queue(struct loop_device *lo)
-{
- kthread_flush_worker(&lo->worker);
- kthread_stop(lo->worker_task);
-}
+struct loop_worker {
+ struct rb_node rb_node;
+ struct work_struct work;
+ struct list_head cmd_list;
+ struct list_head idle_list;
+ struct loop_device *lo;
+ struct cgroup_subsys_state *css;
+ unsigned long last_ran_at;
+};
-static int loop_kthread_worker_fn(void *worker_ptr)
-{
- current->flags |= PF_LOCAL_THROTTLE | PF_MEMALLOC_NOIO;
- return kthread_worker_fn(worker_ptr);
-}
+static void loop_workfn(struct work_struct *work);
+static void loop_rootcg_workfn(struct work_struct *work);
+static void loop_free_idle_workers(struct timer_list *timer);
-static int loop_prepare_queue(struct loop_device *lo)
+static void loop_queue_work(struct loop_device *lo, struct loop_cmd *cmd)
{
- kthread_init_worker(&lo->worker);
- lo->worker_task = kthread_run(loop_kthread_worker_fn,
- &lo->worker, "loop%d", lo->lo_number);
- if (IS_ERR(lo->worker_task))
- return -ENOMEM;
- set_user_nice(lo->worker_task, MIN_NICE);
- return 0;
+ struct rb_node **node = &(lo->worker_tree.rb_node), *parent = NULL;
+ struct loop_worker *cur_worker, *worker = NULL;
+ struct work_struct *work;
+ struct list_head *cmd_list;
+
+ spin_lock_irq(&lo->lo_work_lock);
+
+ if (!cmd->css)
+ goto queue_work;
+
+ node = &lo->worker_tree.rb_node;
+
+ while (*node) {
+ parent = *node;
+ cur_worker = container_of(*node, struct loop_worker, rb_node);
+ if (cur_worker->css == cmd->css) {
+ worker = cur_worker;
+ break;
+ } else if ((long)cur_worker->css < (long)cmd->css) {
+ node = &(*node)->rb_left;
+ } else {
+ node = &(*node)->rb_right;
+ }
+ }
+ if (worker)
+ goto queue_work;
+
+ worker = kzalloc(sizeof(struct loop_worker), GFP_NOWAIT | __GFP_NOWARN);
+ /*
+ * In the event we cannot allocate a worker, just queue on the
+ * rootcg worker
+ */
+ if (!worker)
+ goto queue_work;
+
+ worker->css = cmd->css;
+ css_get(worker->css);
+ INIT_WORK(&worker->work, loop_workfn);
+ INIT_LIST_HEAD(&worker->cmd_list);
+ INIT_LIST_HEAD(&worker->idle_list);
+ worker->lo = lo;
+ rb_link_node(&worker->rb_node, parent, node);
+ rb_insert_color(&worker->rb_node, &lo->worker_tree);
+queue_work:
+ if (worker) {
+ /*
+ * We need to remove from the idle list here while
+ * holding the lock so that the idle timer doesn't
+ * free the worker
+ */
+ if (!list_empty(&worker->idle_list))
+ list_del_init(&worker->idle_list);
+ work = &worker->work;
+ cmd_list = &worker->cmd_list;
+ } else {
+ work = &lo->rootcg_work;
+ cmd_list = &lo->rootcg_cmd_list;
+ }
+ list_add_tail(&cmd->list_entry, cmd_list);
+ queue_work(lo->workqueue, work);
+ spin_unlock_irq(&lo->lo_work_lock);
}
static void loop_update_rotational(struct loop_device *lo)
@@ -1132,12 +1189,27 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
!file->f_op->write_iter)
lo->lo_flags |= LO_FLAGS_READ_ONLY;
- error = loop_prepare_queue(lo);
- if (error)
+ error = -EFBIG;
+ size = get_loop_size(lo, file);
+ if ((loff_t)(sector_t)size != size)
goto out_unlock;
+ lo->workqueue = alloc_workqueue("loop%d",
+ WQ_UNBOUND | WQ_FREEZABLE |
+ WQ_MEM_RECLAIM,
+ lo->lo_number);
+ if (!lo->workqueue) {
+ error = -ENOMEM;
+ goto out_unlock;
+ }
set_device_ro(bdev, (lo->lo_flags & LO_FLAGS_READ_ONLY) != 0);
+ INIT_WORK(&lo->rootcg_work, loop_rootcg_workfn);
+ INIT_LIST_HEAD(&lo->rootcg_cmd_list);
+ INIT_LIST_HEAD(&lo->idle_worker_list);
+ lo->worker_tree = RB_ROOT;
+ timer_setup(&lo->timer, loop_free_idle_workers,
+ TIMER_DEFERRABLE);
lo->use_dio = lo->lo_flags & LO_FLAGS_DIRECT_IO;
lo->lo_device = bdev;
lo->lo_backing_file = file;
@@ -1206,6 +1278,7 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
int err = 0;
bool partscan = false;
int lo_number;
+ struct loop_worker *pos, *worker;
mutex_lock(&loop_ctl_mutex);
if (WARN_ON_ONCE(lo->lo_state != Lo_rundown)) {
@@ -1222,6 +1295,18 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
/* freeze request queue during the transition */
blk_mq_freeze_queue(lo->lo_queue);
+ destroy_workqueue(lo->workqueue);
+ spin_lock_irq(&lo->lo_work_lock);
+ list_for_each_entry_safe(worker, pos, &lo->idle_worker_list,
+ idle_list) {
+ list_del(&worker->idle_list);
+ rb_erase(&worker->rb_node, &lo->worker_tree);
+ css_put(worker->css);
+ kfree(worker);
+ }
+ spin_unlock_irq(&lo->lo_work_lock);
+ del_timer_sync(&lo->timer);
+
spin_lock_irq(&lo->lo_lock);
lo->lo_backing_file = NULL;
spin_unlock_irq(&lo->lo_lock);
@@ -1259,7 +1344,6 @@ static int __loop_clr_fd(struct loop_device *lo, bool release)
partscan = lo->lo_flags & LO_FLAGS_PARTSCAN && bdev;
lo_number = lo->lo_number;
- loop_unprepare_queue(lo);
out_unlock:
mutex_unlock(&loop_ctl_mutex);
if (partscan) {
@@ -2025,7 +2109,7 @@ static blk_status_t loop_queue_rq(struct blk_mq_hw_ctx *hctx,
} else
#endif
cmd->css = NULL;
- kthread_queue_work(&lo->worker, &cmd->work);
+ loop_queue_work(lo, cmd);
return BLK_STS_OK;
}
@@ -2055,26 +2139,82 @@ static void loop_handle_cmd(struct loop_cmd *cmd)
}
}
-static void loop_queue_work(struct kthread_work *work)
+static void loop_set_timer(struct loop_device *lo)
+{
+ timer_reduce(&lo->timer, jiffies + LOOP_IDLE_WORKER_TIMEOUT);
+}
+
+static void loop_process_work(struct loop_worker *worker,
+ struct list_head *cmd_list, struct loop_device *lo)
{
- struct loop_cmd *cmd =
- container_of(work, struct loop_cmd, work);
+ int orig_flags = current->flags;
+ struct loop_cmd *cmd;
- loop_handle_cmd(cmd);
+ current->flags |= PF_LOCAL_THROTTLE | PF_MEMALLOC_NOIO;
+ spin_lock_irq(&lo->lo_work_lock);
+ while (!list_empty(cmd_list)) {
+ cmd = container_of(
+ cmd_list->next, struct loop_cmd, list_entry);
+ list_del(cmd_list->next);
+ spin_unlock_irq(&lo->lo_work_lock);
+
+ loop_handle_cmd(cmd);
+ cond_resched();
+
+ spin_lock_irq(&lo->lo_work_lock);
+ }
+
+ /*
+ * We only add to the idle list if there are no pending cmds
+ * *and* the worker will not run again which ensures that it
+ * is safe to free any worker on the idle list
+ */
+ if (worker && !work_pending(&worker->work)) {
+ worker->last_ran_at = jiffies;
+ list_add_tail(&worker->idle_list, &lo->idle_worker_list);
+ loop_set_timer(lo);
+ }
+ spin_unlock_irq(&lo->lo_work_lock);
+ current->flags = orig_flags;
}
-static int loop_init_request(struct blk_mq_tag_set *set, struct request *rq,
- unsigned int hctx_idx, unsigned int numa_node)
+static void loop_workfn(struct work_struct *work)
{
- struct loop_cmd *cmd = blk_mq_rq_to_pdu(rq);
+ struct loop_worker *worker =
+ container_of(work, struct loop_worker, work);
+ loop_process_work(worker, &worker->cmd_list, worker->lo);
+}
- kthread_init_work(&cmd->work, loop_queue_work);
- return 0;
+static void loop_rootcg_workfn(struct work_struct *work)
+{
+ struct loop_device *lo =
+ container_of(work, struct loop_device, rootcg_work);
+ loop_process_work(NULL, &lo->rootcg_cmd_list, lo);
+}
+
+static void loop_free_idle_workers(struct timer_list *timer)
+{
+ struct loop_device *lo = container_of(timer, struct loop_device, timer);
+ struct loop_worker *pos, *worker;
+
+ spin_lock_irq(&lo->lo_work_lock);
+ list_for_each_entry_safe(worker, pos, &lo->idle_worker_list,
+ idle_list) {
+ if (time_is_after_jiffies(worker->last_ran_at +
+ LOOP_IDLE_WORKER_TIMEOUT))
+ break;
+ list_del(&worker->idle_list);
+ rb_erase(&worker->rb_node, &lo->worker_tree);
+ css_put(worker->css);
+ kfree(worker);
+ }
+ if (!list_empty(&lo->idle_worker_list))
+ loop_set_timer(lo);
+ spin_unlock_irq(&lo->lo_work_lock);
}
static const struct blk_mq_ops loop_mq_ops = {
.queue_rq = loop_queue_rq,
- .init_request = loop_init_request,
.complete = lo_complete_rq,
};
@@ -2162,6 +2302,7 @@ static int loop_add(struct loop_device **l, int i)
atomic_set(&lo->lo_refcnt, 0);
lo->lo_number = i;
spin_lock_init(&lo->lo_lock);
+ spin_lock_init(&lo->lo_work_lock);
disk->major = LOOP_MAJOR;
disk->first_minor = i << part_shift;
disk->fops = &lo_fops;
diff --git a/drivers/block/loop.h b/drivers/block/loop.h
index af75a5ee4094..0162b55a68e1 100644
--- a/drivers/block/loop.h
+++ b/drivers/block/loop.h
@@ -14,7 +14,6 @@
#include <linux/blk-mq.h>
#include <linux/spinlock.h>
#include <linux/mutex.h>
-#include <linux/kthread.h>
#include <uapi/linux/loop.h>
/* Possible states of device */
@@ -54,8 +53,13 @@ struct loop_device {
spinlock_t lo_lock;
int lo_state;
- struct kthread_worker worker;
- struct task_struct *worker_task;
+ spinlock_t lo_work_lock;
+ struct workqueue_struct *workqueue;
+ struct work_struct rootcg_work;
+ struct list_head rootcg_cmd_list;
+ struct list_head idle_worker_list;
+ struct rb_root worker_tree;
+ struct timer_list timer;
bool use_dio;
bool sysfs_inited;
@@ -65,7 +69,7 @@ struct loop_device {
};
struct loop_cmd {
- struct kthread_work work;
+ struct list_head list_entry;
bool use_aio; /* use AIO interface to handle I/O */
atomic_t ref; /* only for aio */
long ret;
--
2.24.1
memalloc_use_memcg() worked for kernel allocations but was silently
ignored for user pages.
This patch establishes a precedence order for who gets charged:
1. If there is a memcg associated with the page already, that memcg is
charged. This happens during swapin.
2. If an explicit mm is passed, mm->memcg is charged. This happens
during page faults, which can be triggered in remote VMs (eg gup).
3. Otherwise consult the current process context. If it has configured
a current->active_memcg, use that. Otherwise, current->mm->memcg.
Previously, if a NULL mm was passed to mem_cgroup_try_charge (case 3) it
would always charge the root cgroup. Now it looks up the current
active_memcg first (falling back to charging the root cgroup if not
set).
Signed-off-by: Dan Schatzberg <[email protected]>
Acked-by: Johannes Weiner <[email protected]>
Acked-by: Tejun Heo <[email protected]>
Acked-by: Chris Down <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
---
mm/memcontrol.c | 11 ++++++++---
mm/shmem.c | 5 +++--
2 files changed, 11 insertions(+), 5 deletions(-)
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index b2468c80085d..79c70eef3ec3 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -6676,7 +6676,8 @@ void mem_cgroup_calculate_protection(struct mem_cgroup *root,
* @gfp_mask: reclaim mode
*
* Try to charge @page to the memcg that @mm belongs to, reclaiming
- * pages according to @gfp_mask if necessary.
+ * pages according to @gfp_mask if necessary. if @mm is NULL, try to
+ * charge to the active memcg.
*
* Returns 0 on success. Otherwise, an error code is returned.
*/
@@ -6712,8 +6713,12 @@ int mem_cgroup_charge(struct page *page, struct mm_struct *mm, gfp_t gfp_mask)
rcu_read_unlock();
}
- if (!memcg)
- memcg = get_mem_cgroup_from_mm(mm);
+ if (!memcg) {
+ if (!mm)
+ memcg = get_mem_cgroup_from_current();
+ else
+ memcg = get_mem_cgroup_from_mm(mm);
+ }
ret = try_charge(memcg, gfp_mask, nr_pages);
if (ret)
diff --git a/mm/shmem.c b/mm/shmem.c
index 271548ca20f3..77c908730be4 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -1695,7 +1695,8 @@ static int shmem_swapin_page(struct inode *inode, pgoff_t index,
{
struct address_space *mapping = inode->i_mapping;
struct shmem_inode_info *info = SHMEM_I(inode);
- struct mm_struct *charge_mm = vma ? vma->vm_mm : current->mm;
+ struct mm_struct *charge_mm = vma ? vma->vm_mm : NULL;
+ struct mem_cgroup *memcg;
struct page *page;
swp_entry_t swap;
int error;
@@ -1809,7 +1810,7 @@ static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
}
sbinfo = SHMEM_SB(inode->i_sb);
- charge_mm = vma ? vma->vm_mm : current->mm;
+ charge_mm = vma ? vma->vm_mm : NULL;
page = find_lock_entry(mapping, index);
if (xa_is_value(page)) {
--
2.24.1
From: Johannes Weiner <[email protected]>
The memalloc_use_memcg() function to override the default memcg
accounting context currently doesn't nest. But the patches to make the
loop driver cgroup-aware will end up nesting:
[ 98.137605] alloc_page_buffers+0x210/0x288
[ 98.141799] __getblk_gfp+0x1d4/0x400
[ 98.145475] ext4_read_block_bitmap_nowait+0x148/0xbc8
[ 98.150628] ext4_mb_init_cache+0x25c/0x9b0
[ 98.154821] ext4_mb_init_group+0x270/0x390
[ 98.159014] ext4_mb_good_group+0x264/0x270
[ 98.163208] ext4_mb_regular_allocator+0x480/0x798
[ 98.168011] ext4_mb_new_blocks+0x958/0x10f8
[ 98.172294] ext4_ext_map_blocks+0xec8/0x1618
[ 98.176660] ext4_map_blocks+0x1b8/0x8a0
[ 98.180592] ext4_writepages+0x830/0xf10
[ 98.184523] do_writepages+0xb4/0x198
[ 98.188195] __filemap_fdatawrite_range+0x170/0x1c8
[ 98.193086] filemap_write_and_wait_range+0x40/0xb0
[ 98.197974] ext4_punch_hole+0x4a4/0x660
[ 98.201907] ext4_fallocate+0x294/0x1190
[ 98.205839] loop_process_work+0x690/0x1100
[ 98.210032] loop_workfn+0x2c/0x110
[ 98.213529] process_one_work+0x3e0/0x648
[ 98.217546] worker_thread+0x70/0x670
[ 98.221217] kthread+0x1b8/0x1c0
[ 98.224452] ret_from_fork+0x10/0x18
where loop_process_work() sets the memcg override to the memcg that
submitted the IO request, and alloc_page_buffers() sets the override
to the memcg that instantiated the cache page, which may differ.
Make memalloc_use_memcg() return the old memcg and convert existing
users to a stacking model. Delete the unused memalloc_unuse_memcg().
Signed-off-by: Johannes Weiner <[email protected]>
Reviewed-by: Shakeel Butt <[email protected]>
Acked-by: Roman Gushchin <[email protected]>
Reported-by: Naresh Kamboju <[email protected]>
---
fs/buffer.c | 6 +++---
fs/notify/fanotify/fanotify.c | 5 +++--
fs/notify/inotify/inotify_fsnotify.c | 5 +++--
include/linux/sched/mm.h | 28 +++++++++-------------------
mm/memcontrol.c | 6 +++---
5 files changed, 21 insertions(+), 29 deletions(-)
diff --git a/fs/buffer.c b/fs/buffer.c
index d468ed9981e0..804170cb59fe 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -842,13 +842,13 @@ struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size,
struct buffer_head *bh, *head;
gfp_t gfp = GFP_NOFS | __GFP_ACCOUNT;
long offset;
- struct mem_cgroup *memcg;
+ struct mem_cgroup *memcg, *old_memcg;
if (retry)
gfp |= __GFP_NOFAIL;
memcg = get_mem_cgroup_from_page(page);
- memalloc_use_memcg(memcg);
+ old_memcg = memalloc_use_memcg(memcg);
head = NULL;
offset = PAGE_SIZE;
@@ -867,7 +867,7 @@ struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size,
set_bh_page(bh, page, offset);
}
out:
- memalloc_unuse_memcg();
+ memalloc_use_memcg(old_memcg);
mem_cgroup_put(memcg);
return head;
/*
diff --git a/fs/notify/fanotify/fanotify.c b/fs/notify/fanotify/fanotify.c
index c942910a8649..c8fd563e02a3 100644
--- a/fs/notify/fanotify/fanotify.c
+++ b/fs/notify/fanotify/fanotify.c
@@ -533,6 +533,7 @@ static struct fanotify_event *fanotify_alloc_event(struct fsnotify_group *group,
unsigned int fid_mode = FAN_GROUP_FLAG(group, FANOTIFY_FID_BITS);
struct inode *child = NULL;
bool name_event = false;
+ struct mem_cgroup *old_memcg;
if ((fid_mode & FAN_REPORT_DIR_FID) && dirid) {
/*
@@ -580,7 +581,7 @@ static struct fanotify_event *fanotify_alloc_event(struct fsnotify_group *group,
gfp |= __GFP_RETRY_MAYFAIL;
/* Whoever is interested in the event, pays for the allocation. */
- memalloc_use_memcg(group->memcg);
+ old_memcg = memalloc_use_memcg(group->memcg);
if (fanotify_is_perm_event(mask)) {
event = fanotify_alloc_perm_event(path, gfp);
@@ -608,7 +609,7 @@ static struct fanotify_event *fanotify_alloc_event(struct fsnotify_group *group,
event->pid = get_pid(task_tgid(current));
out:
- memalloc_unuse_memcg();
+ memalloc_use_memcg(old_memcg);
return event;
}
diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c
index a65cf8c9f600..8017a51561c4 100644
--- a/fs/notify/inotify/inotify_fsnotify.c
+++ b/fs/notify/inotify/inotify_fsnotify.c
@@ -66,6 +66,7 @@ static int inotify_one_event(struct fsnotify_group *group, u32 mask,
int ret;
int len = 0;
int alloc_len = sizeof(struct inotify_event_info);
+ struct mem_cgroup *old_memcg;
if ((inode_mark->mask & FS_EXCL_UNLINK) &&
path && d_unlinked(path->dentry))
@@ -87,9 +88,9 @@ static int inotify_one_event(struct fsnotify_group *group, u32 mask,
* trigger OOM killer in the target monitoring memcg as it may have
* security repercussion.
*/
- memalloc_use_memcg(group->memcg);
+ old_memcg = memalloc_use_memcg(group->memcg);
event = kmalloc(alloc_len, GFP_KERNEL_ACCOUNT | __GFP_RETRY_MAYFAIL);
- memalloc_unuse_memcg();
+ memalloc_use_memcg(old_memcg);
if (unlikely(!event)) {
/*
diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h
index f889e332912f..b8fde48d44a9 100644
--- a/include/linux/sched/mm.h
+++ b/include/linux/sched/mm.h
@@ -312,31 +312,21 @@ static inline void memalloc_nocma_restore(unsigned int flags)
* __GFP_ACCOUNT allocations till the end of the scope will be charged to the
* given memcg.
*
- * NOTE: This function is not nesting safe.
+ * NOTE: This function can nest. Users must save the return value and
+ * reset the previous value after their own charging scope is over
*/
-static inline void memalloc_use_memcg(struct mem_cgroup *memcg)
+static inline struct mem_cgroup *
+memalloc_use_memcg(struct mem_cgroup *memcg)
{
- WARN_ON_ONCE(current->active_memcg);
+ struct mem_cgroup *old = current->active_memcg;
current->active_memcg = memcg;
-}
-
-/**
- * memalloc_unuse_memcg - Ends the remote memcg charging scope.
- *
- * This function marks the end of the remote memcg charging scope started by
- * memalloc_use_memcg().
- */
-static inline void memalloc_unuse_memcg(void)
-{
- current->active_memcg = NULL;
+ return old;
}
#else
-static inline void memalloc_use_memcg(struct mem_cgroup *memcg)
-{
-}
-
-static inline void memalloc_unuse_memcg(void)
+static inline struct mem_cgroup *
+memalloc_use_memcg(struct mem_cgroup *memcg)
{
+ return NULL;
}
#endif
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index b807952b4d43..b2468c80085d 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -5271,12 +5271,12 @@ static struct cgroup_subsys_state * __ref
mem_cgroup_css_alloc(struct cgroup_subsys_state *parent_css)
{
struct mem_cgroup *parent = mem_cgroup_from_css(parent_css);
- struct mem_cgroup *memcg;
+ struct mem_cgroup *memcg, *old_memcg;
long error = -ENOMEM;
- memalloc_use_memcg(parent);
+ old_memcg = memalloc_use_memcg(parent);
memcg = mem_cgroup_alloc();
- memalloc_unuse_memcg();
+ memalloc_use_memcg(old_memcg);
if (IS_ERR(memcg))
return ERR_CAST(memcg);
--
2.24.1
On Mon, Aug 24, 2020 at 11:36:00AM -0400, Dan Schatzberg wrote:
> From: Johannes Weiner <[email protected]>
>
> The memalloc_use_memcg() function to override the default memcg
> accounting context currently doesn't nest. But the patches to make the
> loop driver cgroup-aware will end up nesting:
>
> [ 98.137605] alloc_page_buffers+0x210/0x288
> [ 98.141799] __getblk_gfp+0x1d4/0x400
> [ 98.145475] ext4_read_block_bitmap_nowait+0x148/0xbc8
> [ 98.150628] ext4_mb_init_cache+0x25c/0x9b0
> [ 98.154821] ext4_mb_init_group+0x270/0x390
> [ 98.159014] ext4_mb_good_group+0x264/0x270
> [ 98.163208] ext4_mb_regular_allocator+0x480/0x798
> [ 98.168011] ext4_mb_new_blocks+0x958/0x10f8
> [ 98.172294] ext4_ext_map_blocks+0xec8/0x1618
> [ 98.176660] ext4_map_blocks+0x1b8/0x8a0
> [ 98.180592] ext4_writepages+0x830/0xf10
> [ 98.184523] do_writepages+0xb4/0x198
> [ 98.188195] __filemap_fdatawrite_range+0x170/0x1c8
> [ 98.193086] filemap_write_and_wait_range+0x40/0xb0
> [ 98.197974] ext4_punch_hole+0x4a4/0x660
> [ 98.201907] ext4_fallocate+0x294/0x1190
> [ 98.205839] loop_process_work+0x690/0x1100
> [ 98.210032] loop_workfn+0x2c/0x110
> [ 98.213529] process_one_work+0x3e0/0x648
> [ 98.217546] worker_thread+0x70/0x670
> [ 98.221217] kthread+0x1b8/0x1c0
> [ 98.224452] ret_from_fork+0x10/0x18
>
> where loop_process_work() sets the memcg override to the memcg that
> submitted the IO request, and alloc_page_buffers() sets the override
> to the memcg that instantiated the cache page, which may differ.
>
> Make memalloc_use_memcg() return the old memcg and convert existing
> users to a stacking model. Delete the unused memalloc_unuse_memcg().
>
> Signed-off-by: Johannes Weiner <[email protected]>
> Reviewed-by: Shakeel Butt <[email protected]>
> Acked-by: Roman Gushchin <[email protected]>
> Reported-by: Naresh Kamboju <[email protected]>
Hi Dan,
JFYI: I need a similar patch for the bpf memory accounting rework,
so I ended up sending it separately (with some modifications including
different naming): https://lkml.org/lkml/2020/8/21/1464 .
Can you please, rebase your patchset using this patch?
I hope Andrew can pull this standalone patch into 5.9-rc*,
as Shakeel suggested. It will help us to avoid merge conflicts
during the 5.10 merge window.
Thanks!
On Mon, Aug 24, 2020 at 09:19:01AM -0700, Roman Gushchin wrote:
> Hi Dan,
>
> JFYI: I need a similar patch for the bpf memory accounting rework,
> so I ended up sending it separately (with some modifications including
> different naming): https://lkml.org/lkml/2020/8/21/1464 .
>
> Can you please, rebase your patchset using this patch?
>
> I hope Andrew can pull this standalone patch into 5.9-rc*,
> as Shakeel suggested. It will help us to avoid merge conflicts
> during the 5.10 merge window.
>
> Thanks!
Yeah I mentioned it in the cover letter and linked to your patch. I
had not realized the naming change, so I can rebase on top of that -
I'll wait to see if there's other feedback first.