Revision 6f1b228529ae49b0f85ab89bcdb6c365df401558 authored by Gautham Ananthakrishna on 28 October 2021, 21:36:17 UTC, committed by Linus Torvalds on 29 October 2021, 00:18:55 UTC
Encountered a race between ocfs2_test_bg_bit_allocatable() and jbd2_journal_put_journal_head() resulting in the below vmcore. PID: 106879 TASK: ffff880244ba9c00 CPU: 2 COMMAND: "loop3" Call trace: panic oops_end no_context __bad_area_nosemaphore bad_area_nosemaphore __do_page_fault do_page_fault page_fault [exception RIP: ocfs2_block_group_find_clear_bits+316] ocfs2_block_group_find_clear_bits [ocfs2] ocfs2_cluster_group_search [ocfs2] ocfs2_search_chain [ocfs2] ocfs2_claim_suballoc_bits [ocfs2] __ocfs2_claim_clusters [ocfs2] ocfs2_claim_clusters [ocfs2] ocfs2_local_alloc_slide_window [ocfs2] ocfs2_reserve_local_alloc_bits [ocfs2] ocfs2_reserve_clusters_with_limit [ocfs2] ocfs2_reserve_clusters [ocfs2] ocfs2_lock_refcount_allocators [ocfs2] ocfs2_make_clusters_writable [ocfs2] ocfs2_replace_cow [ocfs2] ocfs2_refcount_cow [ocfs2] ocfs2_file_write_iter [ocfs2] lo_rw_aio loop_queue_work kthread_worker_fn kthread ret_from_fork When ocfs2_test_bg_bit_allocatable() called bh2jh(bg_bh), the bg_bh->b_private NULL as jbd2_journal_put_journal_head() raced and released the jounal head from the buffer head. Needed to take bit lock for the bit 'BH_JournalHead' to fix this race. Link: https://lkml.kernel.org/r/1634820718-6043-1-git-send-email-gautham.ananthakrishna@oracle.com Signed-off-by: Gautham Ananthakrishna <gautham.ananthakrishna@oracle.com> Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com> Cc: <rajesh.sivaramasubramaniom@oracle.com> Cc: Mark Fasheh <mark@fasheh.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Junxiao Bi <junxiao.bi@oracle.com> Cc: Changwei Ge <gechangwei@live.cn> Cc: Gang He <ghe@suse.com> Cc: Jun Piao <piaojun@huawei.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent 337546e
blk-ioc.c
// SPDX-License-Identifier: GPL-2.0
/*
* Functions related to io context handling
*/
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/bio.h>
#include <linux/blkdev.h>
#include <linux/slab.h>
#include <linux/sched/task.h>
#include "blk.h"
/*
* For io context allocations
*/
static struct kmem_cache *iocontext_cachep;
/**
* get_io_context - increment reference count to io_context
* @ioc: io_context to get
*
* Increment reference count to @ioc.
*/
void get_io_context(struct io_context *ioc)
{
BUG_ON(atomic_long_read(&ioc->refcount) <= 0);
atomic_long_inc(&ioc->refcount);
}
static void icq_free_icq_rcu(struct rcu_head *head)
{
struct io_cq *icq = container_of(head, struct io_cq, __rcu_head);
kmem_cache_free(icq->__rcu_icq_cache, icq);
}
/*
* Exit an icq. Called with ioc locked for blk-mq, and with both ioc
* and queue locked for legacy.
*/
static void ioc_exit_icq(struct io_cq *icq)
{
struct elevator_type *et = icq->q->elevator->type;
if (icq->flags & ICQ_EXITED)
return;
if (et->ops.exit_icq)
et->ops.exit_icq(icq);
icq->flags |= ICQ_EXITED;
}
/*
* Release an icq. Called with ioc locked for blk-mq, and with both ioc
* and queue locked for legacy.
*/
static void ioc_destroy_icq(struct io_cq *icq)
{
struct io_context *ioc = icq->ioc;
struct request_queue *q = icq->q;
struct elevator_type *et = q->elevator->type;
lockdep_assert_held(&ioc->lock);
radix_tree_delete(&ioc->icq_tree, icq->q->id);
hlist_del_init(&icq->ioc_node);
list_del_init(&icq->q_node);
/*
* Both setting lookup hint to and clearing it from @icq are done
* under queue_lock. If it's not pointing to @icq now, it never
* will. Hint assignment itself can race safely.
*/
if (rcu_access_pointer(ioc->icq_hint) == icq)
rcu_assign_pointer(ioc->icq_hint, NULL);
ioc_exit_icq(icq);
/*
* @icq->q might have gone away by the time RCU callback runs
* making it impossible to determine icq_cache. Record it in @icq.
*/
icq->__rcu_icq_cache = et->icq_cache;
icq->flags |= ICQ_DESTROYED;
call_rcu(&icq->__rcu_head, icq_free_icq_rcu);
}
/*
* Slow path for ioc release in put_io_context(). Performs double-lock
* dancing to unlink all icq's and then frees ioc.
*/
static void ioc_release_fn(struct work_struct *work)
{
struct io_context *ioc = container_of(work, struct io_context,
release_work);
spin_lock_irq(&ioc->lock);
while (!hlist_empty(&ioc->icq_list)) {
struct io_cq *icq = hlist_entry(ioc->icq_list.first,
struct io_cq, ioc_node);
struct request_queue *q = icq->q;
if (spin_trylock(&q->queue_lock)) {
ioc_destroy_icq(icq);
spin_unlock(&q->queue_lock);
} else {
/* Make sure q and icq cannot be freed. */
rcu_read_lock();
/* Re-acquire the locks in the correct order. */
spin_unlock(&ioc->lock);
spin_lock(&q->queue_lock);
spin_lock(&ioc->lock);
/*
* The icq may have been destroyed when the ioc lock
* was released.
*/
if (!(icq->flags & ICQ_DESTROYED))
ioc_destroy_icq(icq);
spin_unlock(&q->queue_lock);
rcu_read_unlock();
}
}
spin_unlock_irq(&ioc->lock);
kmem_cache_free(iocontext_cachep, ioc);
}
/**
* put_io_context - put a reference of io_context
* @ioc: io_context to put
*
* Decrement reference count of @ioc and release it if the count reaches
* zero.
*/
void put_io_context(struct io_context *ioc)
{
unsigned long flags;
bool free_ioc = false;
if (ioc == NULL)
return;
BUG_ON(atomic_long_read(&ioc->refcount) <= 0);
/*
* Releasing ioc requires reverse order double locking and we may
* already be holding a queue_lock. Do it asynchronously from wq.
*/
if (atomic_long_dec_and_test(&ioc->refcount)) {
spin_lock_irqsave(&ioc->lock, flags);
if (!hlist_empty(&ioc->icq_list))
queue_work(system_power_efficient_wq,
&ioc->release_work);
else
free_ioc = true;
spin_unlock_irqrestore(&ioc->lock, flags);
}
if (free_ioc)
kmem_cache_free(iocontext_cachep, ioc);
}
/**
* put_io_context_active - put active reference on ioc
* @ioc: ioc of interest
*
* Undo get_io_context_active(). If active reference reaches zero after
* put, @ioc can never issue further IOs and ioscheds are notified.
*/
void put_io_context_active(struct io_context *ioc)
{
struct io_cq *icq;
if (!atomic_dec_and_test(&ioc->active_ref)) {
put_io_context(ioc);
return;
}
spin_lock_irq(&ioc->lock);
hlist_for_each_entry(icq, &ioc->icq_list, ioc_node) {
if (icq->flags & ICQ_EXITED)
continue;
ioc_exit_icq(icq);
}
spin_unlock_irq(&ioc->lock);
put_io_context(ioc);
}
/* Called by the exiting task */
void exit_io_context(struct task_struct *task)
{
struct io_context *ioc;
task_lock(task);
ioc = task->io_context;
task->io_context = NULL;
task_unlock(task);
atomic_dec(&ioc->nr_tasks);
put_io_context_active(ioc);
}
static void __ioc_clear_queue(struct list_head *icq_list)
{
unsigned long flags;
rcu_read_lock();
while (!list_empty(icq_list)) {
struct io_cq *icq = list_entry(icq_list->next,
struct io_cq, q_node);
struct io_context *ioc = icq->ioc;
spin_lock_irqsave(&ioc->lock, flags);
if (icq->flags & ICQ_DESTROYED) {
spin_unlock_irqrestore(&ioc->lock, flags);
continue;
}
ioc_destroy_icq(icq);
spin_unlock_irqrestore(&ioc->lock, flags);
}
rcu_read_unlock();
}
/**
* ioc_clear_queue - break any ioc association with the specified queue
* @q: request_queue being cleared
*
* Walk @q->icq_list and exit all io_cq's.
*/
void ioc_clear_queue(struct request_queue *q)
{
LIST_HEAD(icq_list);
spin_lock_irq(&q->queue_lock);
list_splice_init(&q->icq_list, &icq_list);
spin_unlock_irq(&q->queue_lock);
__ioc_clear_queue(&icq_list);
}
int create_task_io_context(struct task_struct *task, gfp_t gfp_flags, int node)
{
struct io_context *ioc;
int ret;
ioc = kmem_cache_alloc_node(iocontext_cachep, gfp_flags | __GFP_ZERO,
node);
if (unlikely(!ioc))
return -ENOMEM;
/* initialize */
atomic_long_set(&ioc->refcount, 1);
atomic_set(&ioc->nr_tasks, 1);
atomic_set(&ioc->active_ref, 1);
spin_lock_init(&ioc->lock);
INIT_RADIX_TREE(&ioc->icq_tree, GFP_ATOMIC);
INIT_HLIST_HEAD(&ioc->icq_list);
INIT_WORK(&ioc->release_work, ioc_release_fn);
/*
* Try to install. ioc shouldn't be installed if someone else
* already did or @task, which isn't %current, is exiting. Note
* that we need to allow ioc creation on exiting %current as exit
* path may issue IOs from e.g. exit_files(). The exit path is
* responsible for not issuing IO after exit_io_context().
*/
task_lock(task);
if (!task->io_context &&
(task == current || !(task->flags & PF_EXITING)))
task->io_context = ioc;
else
kmem_cache_free(iocontext_cachep, ioc);
ret = task->io_context ? 0 : -EBUSY;
task_unlock(task);
return ret;
}
/**
* get_task_io_context - get io_context of a task
* @task: task of interest
* @gfp_flags: allocation flags, used if allocation is necessary
* @node: allocation node, used if allocation is necessary
*
* Return io_context of @task. If it doesn't exist, it is created with
* @gfp_flags and @node. The returned io_context has its reference count
* incremented.
*
* This function always goes through task_lock() and it's better to use
* %current->io_context + get_io_context() for %current.
*/
struct io_context *get_task_io_context(struct task_struct *task,
gfp_t gfp_flags, int node)
{
struct io_context *ioc;
might_sleep_if(gfpflags_allow_blocking(gfp_flags));
do {
task_lock(task);
ioc = task->io_context;
if (likely(ioc)) {
get_io_context(ioc);
task_unlock(task);
return ioc;
}
task_unlock(task);
} while (!create_task_io_context(task, gfp_flags, node));
return NULL;
}
/**
* ioc_lookup_icq - lookup io_cq from ioc
* @ioc: the associated io_context
* @q: the associated request_queue
*
* Look up io_cq associated with @ioc - @q pair from @ioc. Must be called
* with @q->queue_lock held.
*/
struct io_cq *ioc_lookup_icq(struct io_context *ioc, struct request_queue *q)
{
struct io_cq *icq;
lockdep_assert_held(&q->queue_lock);
/*
* icq's are indexed from @ioc using radix tree and hint pointer,
* both of which are protected with RCU. All removals are done
* holding both q and ioc locks, and we're holding q lock - if we
* find a icq which points to us, it's guaranteed to be valid.
*/
rcu_read_lock();
icq = rcu_dereference(ioc->icq_hint);
if (icq && icq->q == q)
goto out;
icq = radix_tree_lookup(&ioc->icq_tree, q->id);
if (icq && icq->q == q)
rcu_assign_pointer(ioc->icq_hint, icq); /* allowed to race */
else
icq = NULL;
out:
rcu_read_unlock();
return icq;
}
EXPORT_SYMBOL(ioc_lookup_icq);
/**
* ioc_create_icq - create and link io_cq
* @ioc: io_context of interest
* @q: request_queue of interest
* @gfp_mask: allocation mask
*
* Make sure io_cq linking @ioc and @q exists. If icq doesn't exist, they
* will be created using @gfp_mask.
*
* The caller is responsible for ensuring @ioc won't go away and @q is
* alive and will stay alive until this function returns.
*/
struct io_cq *ioc_create_icq(struct io_context *ioc, struct request_queue *q,
gfp_t gfp_mask)
{
struct elevator_type *et = q->elevator->type;
struct io_cq *icq;
/* allocate stuff */
icq = kmem_cache_alloc_node(et->icq_cache, gfp_mask | __GFP_ZERO,
q->node);
if (!icq)
return NULL;
if (radix_tree_maybe_preload(gfp_mask) < 0) {
kmem_cache_free(et->icq_cache, icq);
return NULL;
}
icq->ioc = ioc;
icq->q = q;
INIT_LIST_HEAD(&icq->q_node);
INIT_HLIST_NODE(&icq->ioc_node);
/* lock both q and ioc and try to link @icq */
spin_lock_irq(&q->queue_lock);
spin_lock(&ioc->lock);
if (likely(!radix_tree_insert(&ioc->icq_tree, q->id, icq))) {
hlist_add_head(&icq->ioc_node, &ioc->icq_list);
list_add(&icq->q_node, &q->icq_list);
if (et->ops.init_icq)
et->ops.init_icq(icq);
} else {
kmem_cache_free(et->icq_cache, icq);
icq = ioc_lookup_icq(ioc, q);
if (!icq)
printk(KERN_ERR "cfq: icq link failed!\n");
}
spin_unlock(&ioc->lock);
spin_unlock_irq(&q->queue_lock);
radix_tree_preload_end();
return icq;
}
static int __init blk_ioc_init(void)
{
iocontext_cachep = kmem_cache_create("blkdev_ioc",
sizeof(struct io_context), 0, SLAB_PANIC, NULL);
return 0;
}
subsys_initcall(blk_ioc_init);
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...