linux/fs/btrfs/async-thread.c
Zhao Lei 20b2e3029e btrfs: Fix lockdep warning of wr_ctx->wr_lock in scrub_free_wr_ctx()
lockdep report following warning in test:
 [25176.843958] =================================
 [25176.844519] [ INFO: inconsistent lock state ]
 [25176.845047] 4.1.0-rc3 #22 Tainted: G        W
 [25176.845591] ---------------------------------
 [25176.846153] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
 [25176.846713] fsstress/26661 [HC0[0]:SC1[1]:HE1:SE0] takes:
 [25176.847246]  (&wr_ctx->wr_lock){+.?...}, at: [<ffffffffa04cdc6d>] scrub_free_ctx+0x2d/0xf0 [btrfs]
 [25176.847838] {SOFTIRQ-ON-W} state was registered at:
 [25176.848396]   [<ffffffff810bf460>] __lock_acquire+0x6a0/0xe10
 [25176.848955]   [<ffffffff810bfd1e>] lock_acquire+0xce/0x2c0
 [25176.849491]   [<ffffffff816489af>] mutex_lock_nested+0x7f/0x410
 [25176.850029]   [<ffffffffa04d04ff>] scrub_stripe+0x4df/0x1080 [btrfs]
 [25176.850575]   [<ffffffffa04d11b1>] scrub_chunk.isra.19+0x111/0x130 [btrfs]
 [25176.851110]   [<ffffffffa04d144c>] scrub_enumerate_chunks+0x27c/0x510 [btrfs]
 [25176.851660]   [<ffffffffa04d3b87>] btrfs_scrub_dev+0x1c7/0x6c0 [btrfs]
 [25176.852189]   [<ffffffffa04e918e>] btrfs_dev_replace_start+0x36e/0x450 [btrfs]
 [25176.852771]   [<ffffffffa04a98e0>] btrfs_ioctl+0x1e10/0x2d20 [btrfs]
 [25176.853315]   [<ffffffff8121c5b8>] do_vfs_ioctl+0x318/0x570
 [25176.853868]   [<ffffffff8121c851>] SyS_ioctl+0x41/0x80
 [25176.854406]   [<ffffffff8164da17>] system_call_fastpath+0x12/0x6f
 [25176.854935] irq event stamp: 51506
 [25176.855511] hardirqs last  enabled at (51506): [<ffffffff810d4ce5>] vprintk_emit+0x225/0x5e0
 [25176.856059] hardirqs last disabled at (51505): [<ffffffff810d4b77>] vprintk_emit+0xb7/0x5e0
 [25176.856642] softirqs last  enabled at (50886): [<ffffffff81067a23>] __do_softirq+0x363/0x640
 [25176.857184] softirqs last disabled at (50949): [<ffffffff8106804d>] irq_exit+0x10d/0x120
 [25176.857746]
 other info that might help us debug this:
 [25176.858845]  Possible unsafe locking scenario:
 [25176.859981]        CPU0
 [25176.860537]        ----
 [25176.861059]   lock(&wr_ctx->wr_lock);
 [25176.861705]   <Interrupt>
 [25176.862272]     lock(&wr_ctx->wr_lock);
 [25176.862881]
  *** DEADLOCK ***

Reason:
 Above warning is caused by:
 Interrupt
 -> bio_endio()
 -> ...
 -> scrub_put_ctx()
 -> scrub_free_ctx() *1
 -> ...
 -> mutex_lock(&wr_ctx->wr_lock);

 scrub_put_ctx() is allowed to be called in end_bio interrupt, but
 in code design, it will never call scrub_free_ctx(sctx) in interrupe
 context(above *1), because btrfs_scrub_dev() get one additional
 reference of sctx->refs, which makes scrub_free_ctx() only called
 withine btrfs_scrub_dev().

 Now the code runs out of our wish, because free sequence in
 scrub_pending_bio_dec() have a gap.

 Current code:
 -----------------------------------+-----------------------------------
 scrub_pending_bio_dec()            |  btrfs_scrub_dev
 -----------------------------------+-----------------------------------
 atomic_dec(&sctx->bios_in_flight); |
 wake_up(&sctx->list_wait);         |
                                    | scrub_put_ctx()
                                    | -> atomic_dec_and_test(&sctx->refs)
 scrub_put_ctx(sctx);               |
 -> atomic_dec_and_test(&sctx->refs)|
 -> scrub_free_ctx()                |
 -----------------------------------+-----------------------------------

 We expected:
 -----------------------------------+-----------------------------------
 scrub_pending_bio_dec()            |  btrfs_scrub_dev
 -----------------------------------+-----------------------------------
 atomic_dec(&sctx->bios_in_flight); |
 wake_up(&sctx->list_wait);         |
 scrub_put_ctx(sctx);               |
 -> atomic_dec_and_test(&sctx->refs)|
                                    | scrub_put_ctx()
                                    | -> atomic_dec_and_test(&sctx->refs)
                                    | -> scrub_free_ctx()
 -----------------------------------+-----------------------------------

Fix:
 Move scrub_pending_bio_dec() to a workqueue, to avoid this function run
 in interrupt context.
 Tested by check tracelog in debug.

Changelog v1->v2:
 Use workqueue instead of adjust function call sequence in v1,
 because v1 will introduce a bug pointed out by:
 Filipe David Manana <fdmanana@gmail.com>

Reported-by: Qu Wenruo <quwenruo@cn.fujitsu.com>
Signed-off-by: Zhao Lei <zhaolei@cn.fujitsu.com>
Reviewed-by: Filipe Manana <fdmanana@suse.com>
Signed-off-by: Chris Mason <clm@fb.com>
2015-06-10 07:04:52 -07:00

367 lines
9.2 KiB
C

/*
* Copyright (C) 2007 Oracle. All rights reserved.
* Copyright (C) 2014 Fujitsu. All rights reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public
* License v2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public
* License along with this program; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*/
#include <linux/kthread.h>
#include <linux/slab.h>
#include <linux/list.h>
#include <linux/spinlock.h>
#include <linux/freezer.h>
#include "async-thread.h"
#include "ctree.h"
#define WORK_DONE_BIT 0
#define WORK_ORDER_DONE_BIT 1
#define WORK_HIGH_PRIO_BIT 2
#define NO_THRESHOLD (-1)
#define DFT_THRESHOLD (32)
struct __btrfs_workqueue {
struct workqueue_struct *normal_wq;
/* List head pointing to ordered work list */
struct list_head ordered_list;
/* Spinlock for ordered_list */
spinlock_t list_lock;
/* Thresholding related variants */
atomic_t pending;
int max_active;
int current_max;
int thresh;
unsigned int count;
spinlock_t thres_lock;
};
struct btrfs_workqueue {
struct __btrfs_workqueue *normal;
struct __btrfs_workqueue *high;
};
static void normal_work_helper(struct btrfs_work *work);
#define BTRFS_WORK_HELPER(name) \
void btrfs_##name(struct work_struct *arg) \
{ \
struct btrfs_work *work = container_of(arg, struct btrfs_work, \
normal_work); \
normal_work_helper(work); \
}
BTRFS_WORK_HELPER(worker_helper);
BTRFS_WORK_HELPER(delalloc_helper);
BTRFS_WORK_HELPER(flush_delalloc_helper);
BTRFS_WORK_HELPER(cache_helper);
BTRFS_WORK_HELPER(submit_helper);
BTRFS_WORK_HELPER(fixup_helper);
BTRFS_WORK_HELPER(endio_helper);
BTRFS_WORK_HELPER(endio_meta_helper);
BTRFS_WORK_HELPER(endio_meta_write_helper);
BTRFS_WORK_HELPER(endio_raid56_helper);
BTRFS_WORK_HELPER(endio_repair_helper);
BTRFS_WORK_HELPER(rmw_helper);
BTRFS_WORK_HELPER(endio_write_helper);
BTRFS_WORK_HELPER(freespace_write_helper);
BTRFS_WORK_HELPER(delayed_meta_helper);
BTRFS_WORK_HELPER(readahead_helper);
BTRFS_WORK_HELPER(qgroup_rescan_helper);
BTRFS_WORK_HELPER(extent_refs_helper);
BTRFS_WORK_HELPER(scrub_helper);
BTRFS_WORK_HELPER(scrubwrc_helper);
BTRFS_WORK_HELPER(scrubnc_helper);
BTRFS_WORK_HELPER(scrubparity_helper);
static struct __btrfs_workqueue *
__btrfs_alloc_workqueue(const char *name, unsigned int flags, int max_active,
int thresh)
{
struct __btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_NOFS);
if (!ret)
return NULL;
ret->max_active = max_active;
atomic_set(&ret->pending, 0);
if (thresh == 0)
thresh = DFT_THRESHOLD;
/* For low threshold, disabling threshold is a better choice */
if (thresh < DFT_THRESHOLD) {
ret->current_max = max_active;
ret->thresh = NO_THRESHOLD;
} else {
ret->current_max = 1;
ret->thresh = thresh;
}
if (flags & WQ_HIGHPRI)
ret->normal_wq = alloc_workqueue("%s-%s-high", flags,
ret->max_active,
"btrfs", name);
else
ret->normal_wq = alloc_workqueue("%s-%s", flags,
ret->max_active, "btrfs",
name);
if (!ret->normal_wq) {
kfree(ret);
return NULL;
}
INIT_LIST_HEAD(&ret->ordered_list);
spin_lock_init(&ret->list_lock);
spin_lock_init(&ret->thres_lock);
trace_btrfs_workqueue_alloc(ret, name, flags & WQ_HIGHPRI);
return ret;
}
static inline void
__btrfs_destroy_workqueue(struct __btrfs_workqueue *wq);
struct btrfs_workqueue *btrfs_alloc_workqueue(const char *name,
unsigned int flags,
int max_active,
int thresh)
{
struct btrfs_workqueue *ret = kzalloc(sizeof(*ret), GFP_NOFS);
if (!ret)
return NULL;
ret->normal = __btrfs_alloc_workqueue(name, flags & ~WQ_HIGHPRI,
max_active, thresh);
if (!ret->normal) {
kfree(ret);
return NULL;
}
if (flags & WQ_HIGHPRI) {
ret->high = __btrfs_alloc_workqueue(name, flags, max_active,
thresh);
if (!ret->high) {
__btrfs_destroy_workqueue(ret->normal);
kfree(ret);
return NULL;
}
}
return ret;
}
/*
* Hook for threshold which will be called in btrfs_queue_work.
* This hook WILL be called in IRQ handler context,
* so workqueue_set_max_active MUST NOT be called in this hook
*/
static inline void thresh_queue_hook(struct __btrfs_workqueue *wq)
{
if (wq->thresh == NO_THRESHOLD)
return;
atomic_inc(&wq->pending);
}
/*
* Hook for threshold which will be called before executing the work,
* This hook is called in kthread content.
* So workqueue_set_max_active is called here.
*/
static inline void thresh_exec_hook(struct __btrfs_workqueue *wq)
{
int new_max_active;
long pending;
int need_change = 0;
if (wq->thresh == NO_THRESHOLD)
return;
atomic_dec(&wq->pending);
spin_lock(&wq->thres_lock);
/*
* Use wq->count to limit the calling frequency of
* workqueue_set_max_active.
*/
wq->count++;
wq->count %= (wq->thresh / 4);
if (!wq->count)
goto out;
new_max_active = wq->current_max;
/*
* pending may be changed later, but it's OK since we really
* don't need it so accurate to calculate new_max_active.
*/
pending = atomic_read(&wq->pending);
if (pending > wq->thresh)
new_max_active++;
if (pending < wq->thresh / 2)
new_max_active--;
new_max_active = clamp_val(new_max_active, 1, wq->max_active);
if (new_max_active != wq->current_max) {
need_change = 1;
wq->current_max = new_max_active;
}
out:
spin_unlock(&wq->thres_lock);
if (need_change) {
workqueue_set_max_active(wq->normal_wq, wq->current_max);
}
}
static void run_ordered_work(struct __btrfs_workqueue *wq)
{
struct list_head *list = &wq->ordered_list;
struct btrfs_work *work;
spinlock_t *lock = &wq->list_lock;
unsigned long flags;
while (1) {
spin_lock_irqsave(lock, flags);
if (list_empty(list))
break;
work = list_entry(list->next, struct btrfs_work,
ordered_list);
if (!test_bit(WORK_DONE_BIT, &work->flags))
break;
/*
* we are going to call the ordered done function, but
* we leave the work item on the list as a barrier so
* that later work items that are done don't have their
* functions called before this one returns
*/
if (test_and_set_bit(WORK_ORDER_DONE_BIT, &work->flags))
break;
trace_btrfs_ordered_sched(work);
spin_unlock_irqrestore(lock, flags);
work->ordered_func(work);
/* now take the lock again and drop our item from the list */
spin_lock_irqsave(lock, flags);
list_del(&work->ordered_list);
spin_unlock_irqrestore(lock, flags);
/*
* we don't want to call the ordered free functions
* with the lock held though
*/
work->ordered_free(work);
trace_btrfs_all_work_done(work);
}
spin_unlock_irqrestore(lock, flags);
}
static void normal_work_helper(struct btrfs_work *work)
{
struct __btrfs_workqueue *wq;
int need_order = 0;
/*
* We should not touch things inside work in the following cases:
* 1) after work->func() if it has no ordered_free
* Since the struct is freed in work->func().
* 2) after setting WORK_DONE_BIT
* The work may be freed in other threads almost instantly.
* So we save the needed things here.
*/
if (work->ordered_func)
need_order = 1;
wq = work->wq;
trace_btrfs_work_sched(work);
thresh_exec_hook(wq);
work->func(work);
if (need_order) {
set_bit(WORK_DONE_BIT, &work->flags);
run_ordered_work(wq);
}
if (!need_order)
trace_btrfs_all_work_done(work);
}
void btrfs_init_work(struct btrfs_work *work, btrfs_work_func_t uniq_func,
btrfs_func_t func,
btrfs_func_t ordered_func,
btrfs_func_t ordered_free)
{
work->func = func;
work->ordered_func = ordered_func;
work->ordered_free = ordered_free;
INIT_WORK(&work->normal_work, uniq_func);
INIT_LIST_HEAD(&work->ordered_list);
work->flags = 0;
}
static inline void __btrfs_queue_work(struct __btrfs_workqueue *wq,
struct btrfs_work *work)
{
unsigned long flags;
work->wq = wq;
thresh_queue_hook(wq);
if (work->ordered_func) {
spin_lock_irqsave(&wq->list_lock, flags);
list_add_tail(&work->ordered_list, &wq->ordered_list);
spin_unlock_irqrestore(&wq->list_lock, flags);
}
queue_work(wq->normal_wq, &work->normal_work);
trace_btrfs_work_queued(work);
}
void btrfs_queue_work(struct btrfs_workqueue *wq,
struct btrfs_work *work)
{
struct __btrfs_workqueue *dest_wq;
if (test_bit(WORK_HIGH_PRIO_BIT, &work->flags) && wq->high)
dest_wq = wq->high;
else
dest_wq = wq->normal;
__btrfs_queue_work(dest_wq, work);
}
static inline void
__btrfs_destroy_workqueue(struct __btrfs_workqueue *wq)
{
destroy_workqueue(wq->normal_wq);
trace_btrfs_workqueue_destroy(wq);
kfree(wq);
}
void btrfs_destroy_workqueue(struct btrfs_workqueue *wq)
{
if (!wq)
return;
if (wq->high)
__btrfs_destroy_workqueue(wq->high);
__btrfs_destroy_workqueue(wq->normal);
kfree(wq);
}
void btrfs_workqueue_set_max(struct btrfs_workqueue *wq, int max)
{
if (!wq)
return;
wq->normal->max_active = max;
if (wq->high)
wq->high->max_active = max;
}
void btrfs_set_work_high_priority(struct btrfs_work *work)
{
set_bit(WORK_HIGH_PRIO_BIT, &work->flags);
}