mirror of
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
synced 2026-09-10 22:50:07 -04:00
0ba6912f7e
This reverts commite8eef69a99. While DO_ONCE_SLEEPABLE() is used from sleepable/process context, callers may still be holding arbitrary subsystem locks. For instance, __inet_hash_connect() uses get_random_sleepable_once() which invokes DO_ONCE_SLEEPABLE() while holding the socket lock (sk_lock): lock_sock(sk) __inet_hash_connect() get_random_sleepable_once() DO_ONCE_SLEEPABLE() __do_once_sleepable_done() static_branch_disable() static_key_disable() cpus_read_lock() Calling static_branch_disable() directly from __do_once_sleepable_done() causes static_key_disable() to synchronously acquire cpus_read_lock() (cpu_hotplug_lock) and jump_label_mutex inside the caller's lock context. This introduces an unwanted lockdep dependency: sk_lock -> cpu_hotplug_lock Because cpu_hotplug_lock depends on fs_reclaim (via workqueue CPU bringup allocating memory with GFP_KERNEL), and storage/block layers (such as NVMe-TCP) acquire sk_lock during I/O dispatch, lockdep reports circular locking dependencies: set->srcu -> sk_lock -> cpu_hotplug_lock -> fs_reclaim -> q_usage_counter -> elevator_lock -> set->srcu This false positive previously prompted commit19bdb70c77("nvme-tcp: lockdep: use dynamic lockdep keys per socket instance") to work around the warning using per-socket dynamic keys in NVMe-TCP. That in turn broke asynchronous socket teardown and caused syzbot warnings in tcp_tsq_handler(). Restoring once_disable_jump() in __do_once_sleepable_done() ensures that static_branch_disable() is executed asynchronously from a system workqueue without holding the caller's locks. Link: https://lore.kernel.org/20260825142515.1965654-1-edumazet@google.com Fixes:e8eef69a99("once: don't use a work queue to reset sleepable static key") Signed-off-by: Eric Dumazet <edumazet@google.com> Closes: https://lore.kernel.org/lkml/ao0mwtt8ePAINFni@shinhome/ Reported-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com> Cc: Tony Luck <tony.luck@intel.com> Cc: Reinette Chatre <reinette.chatre@intel.com> Cc: Keith Busch <kbusch@kernel.org> Cc: Nilay Shroff <nilay@linux.ibm.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
99 lines
2.2 KiB
C
99 lines
2.2 KiB
C
// SPDX-License-Identifier: GPL-2.0
|
|
#include <linux/slab.h>
|
|
#include <linux/spinlock.h>
|
|
#include <linux/once.h>
|
|
#include <linux/random.h>
|
|
#include <linux/module.h>
|
|
|
|
struct once_work {
|
|
struct work_struct work;
|
|
struct static_key_true *key;
|
|
struct module *module;
|
|
};
|
|
|
|
static void once_deferred(struct work_struct *w)
|
|
{
|
|
struct once_work *work;
|
|
|
|
work = container_of(w, struct once_work, work);
|
|
BUG_ON(!static_key_enabled(work->key));
|
|
static_branch_disable(work->key);
|
|
module_put(work->module);
|
|
kfree(work);
|
|
}
|
|
|
|
static void once_disable_jump(struct static_key_true *key, struct module *mod)
|
|
{
|
|
struct once_work *w;
|
|
|
|
w = kmalloc_obj(*w, GFP_ATOMIC);
|
|
if (!w)
|
|
return;
|
|
|
|
INIT_WORK(&w->work, once_deferred);
|
|
w->key = key;
|
|
w->module = mod;
|
|
__module_get(mod);
|
|
schedule_work(&w->work);
|
|
}
|
|
|
|
static DEFINE_SPINLOCK(once_lock);
|
|
|
|
bool __do_once_start(bool *done, unsigned long *flags)
|
|
__acquires(once_lock)
|
|
{
|
|
spin_lock_irqsave(&once_lock, *flags);
|
|
if (*done) {
|
|
spin_unlock_irqrestore(&once_lock, *flags);
|
|
/* Keep sparse happy by restoring an even lock count on
|
|
* this lock. In case we return here, we don't call into
|
|
* __do_once_done but return early in the DO_ONCE() macro.
|
|
*/
|
|
__acquire(once_lock);
|
|
return false;
|
|
}
|
|
|
|
return true;
|
|
}
|
|
EXPORT_SYMBOL(__do_once_start);
|
|
|
|
void __do_once_done(bool *done, struct static_key_true *once_key,
|
|
unsigned long *flags, struct module *mod)
|
|
__releases(once_lock)
|
|
{
|
|
*done = true;
|
|
spin_unlock_irqrestore(&once_lock, *flags);
|
|
once_disable_jump(once_key, mod);
|
|
}
|
|
EXPORT_SYMBOL(__do_once_done);
|
|
|
|
static DEFINE_MUTEX(once_mutex);
|
|
|
|
bool __do_once_sleepable_start(bool *done)
|
|
__acquires(once_mutex)
|
|
{
|
|
mutex_lock(&once_mutex);
|
|
if (*done) {
|
|
mutex_unlock(&once_mutex);
|
|
/* Keep sparse happy by restoring an even lock count on
|
|
* this mutex. In case we return here, we don't call into
|
|
* __do_once_done but return early in the DO_ONCE_SLEEPABLE() macro.
|
|
*/
|
|
__acquire(once_mutex);
|
|
return false;
|
|
}
|
|
|
|
return true;
|
|
}
|
|
EXPORT_SYMBOL(__do_once_sleepable_start);
|
|
|
|
void __do_once_sleepable_done(bool *done, struct static_key_true *once_key,
|
|
struct module *mod)
|
|
__releases(once_mutex)
|
|
{
|
|
*done = true;
|
|
mutex_unlock(&once_mutex);
|
|
once_disable_jump(once_key, mod);
|
|
}
|
|
EXPORT_SYMBOL(__do_once_sleepable_done);
|