1
linux/io_uring/io-wq.c

1384 lines
33 KiB
C
Raw Normal View History

2019-10-22 09:25:58 -07:00
// SPDX-License-Identifier: GPL-2.0
/*
* Basic worker thread pool for io_uring
*
* Copyright (C) 2019 Jens Axboe
*
*/
#include <linux/kernel.h>
#include <linux/init.h>
#include <linux/errno.h>
#include <linux/sched/signal.h>
#include <linux/percpu.h>
#include <linux/slab.h>
#include <linux/rculist_nulls.h>
#include <linux/cpu.h>
#include <linux/task_work.h>
#include <linux/audit.h>
#include <linux/mmu_context.h>
#include <uapi/linux/io_uring.h>
2019-10-22 09:25:58 -07:00
#include "io-wq.h"
#include "slist.h"
#include "io_uring.h"
2019-10-22 09:25:58 -07:00
#define WORKER_IDLE_TIMEOUT (5 * HZ)
enum {
IO_WORKER_F_UP = 1, /* up and active */
IO_WORKER_F_RUNNING = 2, /* account as running */
IO_WORKER_F_FREE = 4, /* worker on free list */
IO_WORKER_F_BOUND = 8, /* is doing bounded work */
2019-10-22 09:25:58 -07:00
};
enum {
IO_WQ_BIT_EXIT = 0, /* wq exiting */
};
enum {
IO_ACCT_STALLED_BIT = 0, /* stalled on hash */
2019-10-22 09:25:58 -07:00
};
/*
* One for each thread in a wq pool
2019-10-22 09:25:58 -07:00
*/
struct io_worker {
refcount_t ref;
unsigned flags;
struct hlist_nulls_node nulls_node;
struct list_head all_list;
2019-10-22 09:25:58 -07:00
struct task_struct *task;
struct io_wq *wq;
2019-10-22 09:25:58 -07:00
struct io_wq_work *cur_work;
struct io_wq_work *next_work;
raw_spinlock_t lock;
2019-10-22 09:25:58 -07:00
struct completion ref_done;
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
unsigned long create_state;
struct callback_head create_work;
int create_index;
union {
struct rcu_head rcu;
struct work_struct work;
};
2019-10-22 09:25:58 -07:00
};
#if BITS_PER_LONG == 64
#define IO_WQ_HASH_ORDER 6
#else
#define IO_WQ_HASH_ORDER 5
#endif
#define IO_WQ_NR_HASH_BUCKETS (1u << IO_WQ_HASH_ORDER)
struct io_wq_acct {
unsigned nr_workers;
unsigned max_workers;
int index;
atomic_t nr_running;
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spinlock_t lock;
struct io_wq_work_list work_list;
unsigned long flags;
};
enum {
IO_WQ_ACCT_BOUND,
IO_WQ_ACCT_UNBOUND,
IO_WQ_ACCT_NR,
};
2019-10-22 09:25:58 -07:00
/*
* Per io_wq state
*/
struct io_wq {
unsigned long state;
free_work_fn *free_work;
io_wq_work_fn *do_work;
struct io_wq_hash *hash;
atomic_t worker_refs;
struct completion worker_done;
struct hlist_node cpuhp_node;
struct task_struct *task;
struct io_wq_acct acct[IO_WQ_ACCT_NR];
/* lock protects access to elements below */
raw_spinlock_t lock;
struct hlist_nulls_head free_list;
struct list_head all_list;
struct wait_queue_entry wait;
struct io_wq_work *hash_tail[IO_WQ_NR_HASH_BUCKETS];
cpumask_var_t cpu_mask;
2019-10-22 09:25:58 -07:00
};
static enum cpuhp_state io_wq_online;
struct io_cb_cancel_data {
work_cancel_fn *fn;
void *data;
int nr_running;
int nr_pending;
bool cancel_all;
};
static bool create_io_worker(struct io_wq *wq, int index);
static void io_wq_dec_running(struct io_worker *worker);
static bool io_acct_cancel_pending_work(struct io_wq *wq,
struct io_wq_acct *acct,
struct io_cb_cancel_data *match);
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
static void create_worker_cb(struct callback_head *cb);
static void io_wq_cancel_tw_create(struct io_wq *wq);
2019-10-22 09:25:58 -07:00
static bool io_worker_get(struct io_worker *worker)
{
return refcount_inc_not_zero(&worker->ref);
}
static void io_worker_release(struct io_worker *worker)
{
if (refcount_dec_and_test(&worker->ref))
complete(&worker->ref_done);
2019-10-22 09:25:58 -07:00
}
static inline struct io_wq_acct *io_get_acct(struct io_wq *wq, bool bound)
{
return &wq->acct[bound ? IO_WQ_ACCT_BOUND : IO_WQ_ACCT_UNBOUND];
}
static inline struct io_wq_acct *io_work_get_acct(struct io_wq *wq,
struct io_wq_work *work)
{
return io_get_acct(wq, !(work->flags & IO_WQ_WORK_UNBOUND));
}
static inline struct io_wq_acct *io_wq_get_acct(struct io_worker *worker)
{
return io_get_acct(worker->wq, worker->flags & IO_WORKER_F_BOUND);
}
static void io_worker_ref_put(struct io_wq *wq)
{
if (atomic_dec_and_test(&wq->worker_refs))
complete(&wq->worker_done);
}
bool io_wq_worker_stopped(void)
{
struct io_worker *worker = current->worker_private;
if (WARN_ON_ONCE(!io_wq_current_is_worker()))
return true;
return test_bit(IO_WQ_BIT_EXIT, &worker->wq->state);
}
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
static void io_worker_cancel_cb(struct io_worker *worker)
{
struct io_wq_acct *acct = io_wq_get_acct(worker);
struct io_wq *wq = worker->wq;
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
atomic_dec(&acct->nr_running);
raw_spin_lock(&wq->lock);
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
acct->nr_workers--;
raw_spin_unlock(&wq->lock);
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
io_worker_ref_put(wq);
clear_bit_unlock(0, &worker->create_state);
io_worker_release(worker);
}
static bool io_task_worker_match(struct callback_head *cb, void *data)
{
struct io_worker *worker;
if (cb->func != create_worker_cb)
return false;
worker = container_of(cb, struct io_worker, create_work);
return worker == data;
}
2019-10-22 09:25:58 -07:00
static void io_worker_exit(struct io_worker *worker)
{
struct io_wq *wq = worker->wq;
2019-10-22 09:25:58 -07:00
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
while (1) {
struct callback_head *cb = task_work_cancel_match(wq->task,
io_task_worker_match, worker);
if (!cb)
break;
io_worker_cancel_cb(worker);
}
2019-10-22 09:25:58 -07:00
io_worker_release(worker);
wait_for_completion(&worker->ref_done);
2019-10-22 09:25:58 -07:00
raw_spin_lock(&wq->lock);
if (worker->flags & IO_WORKER_F_FREE)
hlist_nulls_del_rcu(&worker->nulls_node);
list_del_rcu(&worker->all_list);
raw_spin_unlock(&wq->lock);
io_wq_dec_running(worker);
/*
* this worker is a goner, clear ->worker_private to avoid any
* inc/dec running calls that could happen as part of exit from
* touching 'worker'.
*/
current->worker_private = NULL;
2019-10-22 09:25:58 -07:00
kfree_rcu(worker, rcu);
io_worker_ref_put(wq);
do_exit(0);
2019-10-22 09:25:58 -07:00
}
static inline bool __io_acct_run_queue(struct io_wq_acct *acct)
{
return !test_bit(IO_ACCT_STALLED_BIT, &acct->flags) &&
!wq_list_empty(&acct->work_list);
}
/*
* If there's work to do, returns true with acct->lock acquired. If not,
* returns false with no lock held.
*/
static inline bool io_acct_run_queue(struct io_wq_acct *acct)
__acquires(&acct->lock)
{
raw_spin_lock(&acct->lock);
if (__io_acct_run_queue(acct))
return true;
raw_spin_unlock(&acct->lock);
return false;
}
/*
* Check head of free list for an available worker. If one isn't available,
* caller must create one.
*/
static bool io_wq_activate_free_worker(struct io_wq *wq,
struct io_wq_acct *acct)
__must_hold(RCU)
{
struct hlist_nulls_node *n;
struct io_worker *worker;
/*
* Iterate free_list and see if we can find an idle worker to
* activate. If a given worker is on the free_list but in the process
* of exiting, keep trying.
*/
hlist_nulls_for_each_entry_rcu(worker, n, &wq->free_list, nulls_node) {
if (!io_worker_get(worker))
continue;
if (io_wq_get_acct(worker) != acct) {
io_worker_release(worker);
continue;
}
/*
* If the worker is already running, it's either already
* starting work or finishing work. In either case, if it does
* to go sleep, we'll kick off a new task for this work anyway.
*/
wake_up_process(worker->task);
io_worker_release(worker);
return true;
}
return false;
}
/*
* We need a worker. If we find a free one, we're good. If not, and we're
* below the max number of workers, create one.
*/
static bool io_wq_create_worker(struct io_wq *wq, struct io_wq_acct *acct)
{
/*
* Most likely an attempt to queue unbounded work on an io_wq that
* wasn't setup with any unbounded workers.
*/
if (unlikely(!acct->max_workers))
pr_warn_once("io-wq is not configured for unbound workers");
raw_spin_lock(&wq->lock);
if (acct->nr_workers >= acct->max_workers) {
raw_spin_unlock(&wq->lock);
return true;
}
acct->nr_workers++;
raw_spin_unlock(&wq->lock);
atomic_inc(&acct->nr_running);
atomic_inc(&wq->worker_refs);
return create_io_worker(wq, acct->index);
}
static void io_wq_inc_running(struct io_worker *worker)
{
struct io_wq_acct *acct = io_wq_get_acct(worker);
atomic_inc(&acct->nr_running);
}
static void create_worker_cb(struct callback_head *cb)
{
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
struct io_worker *worker;
struct io_wq *wq;
struct io_wq_acct *acct;
bool do_create = false;
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
worker = container_of(cb, struct io_worker, create_work);
wq = worker->wq;
acct = &wq->acct[worker->create_index];
raw_spin_lock(&wq->lock);
if (acct->nr_workers < acct->max_workers) {
acct->nr_workers++;
do_create = true;
}
raw_spin_unlock(&wq->lock);
if (do_create) {
create_io_worker(wq, worker->create_index);
} else {
atomic_dec(&acct->nr_running);
io_worker_ref_put(wq);
}
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
clear_bit_unlock(0, &worker->create_state);
io_worker_release(worker);
}
static bool io_queue_worker_create(struct io_worker *worker,
struct io_wq_acct *acct,
task_work_func_t func)
{
struct io_wq *wq = worker->wq;
/* raced with exit, just ignore create call */
if (test_bit(IO_WQ_BIT_EXIT, &wq->state))
goto fail;
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
if (!io_worker_get(worker))
goto fail;
/*
* create_state manages ownership of create_work/index. We should
* only need one entry per worker, as the worker going to sleep
* will trigger the condition, and waking will clear it once it
* runs the task_work.
*/
if (test_bit(0, &worker->create_state) ||
test_and_set_bit_lock(0, &worker->create_state))
goto fail_release;
atomic_inc(&wq->worker_refs);
init_task_work(&worker->create_work, func);
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
worker->create_index = acct->index;
if (!task_work_add(wq->task, &worker->create_work, TWA_SIGNAL)) {
/*
* EXIT may have been set after checking it above, check after
* adding the task_work and remove any creation item if it is
* now set. wq exit does that too, but we can have added this
* work item after we canceled in io_wq_exit_workers().
*/
if (test_bit(IO_WQ_BIT_EXIT, &wq->state))
io_wq_cancel_tw_create(wq);
io_worker_ref_put(wq);
return true;
}
io_worker_ref_put(wq);
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
clear_bit_unlock(0, &worker->create_state);
fail_release:
io_worker_release(worker);
fail:
atomic_dec(&acct->nr_running);
io_worker_ref_put(wq);
return false;
}
static void io_wq_dec_running(struct io_worker *worker)
{
struct io_wq_acct *acct = io_wq_get_acct(worker);
struct io_wq *wq = worker->wq;
if (!(worker->flags & IO_WORKER_F_UP))
return;
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
if (!atomic_dec_and_test(&acct->nr_running))
return;
if (!io_acct_run_queue(acct))
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
return;
raw_spin_unlock(&acct->lock);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
atomic_inc(&acct->nr_running);
atomic_inc(&wq->worker_refs);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
io_queue_worker_create(worker, acct, create_worker_cb);
}
2019-10-22 09:25:58 -07:00
/*
* Worker will start processing some work. Move it to the busy list, if
* it's currently on the freelist
*/
static void __io_worker_busy(struct io_wq *wq, struct io_worker *worker)
2019-10-22 09:25:58 -07:00
{
if (worker->flags & IO_WORKER_F_FREE) {
worker->flags &= ~IO_WORKER_F_FREE;
raw_spin_lock(&wq->lock);
2019-10-22 09:25:58 -07:00
hlist_nulls_del_init_rcu(&worker->nulls_node);
raw_spin_unlock(&wq->lock);
2019-10-22 09:25:58 -07:00
}
}
/*
* No work, worker going to sleep. Move to freelist.
2019-10-22 09:25:58 -07:00
*/
static void __io_worker_idle(struct io_wq *wq, struct io_worker *worker)
__must_hold(wq->lock)
2019-10-22 09:25:58 -07:00
{
if (!(worker->flags & IO_WORKER_F_FREE)) {
worker->flags |= IO_WORKER_F_FREE;
hlist_nulls_add_head_rcu(&worker->nulls_node, &wq->free_list);
2019-10-22 09:25:58 -07:00
}
}
static inline unsigned int io_get_work_hash(struct io_wq_work *work)
{
return work->flags >> IO_WQ_HASH_SHIFT;
}
static bool io_wait_on_hash(struct io_wq *wq, unsigned int hash)
{
bool ret = false;
spin_lock_irq(&wq->hash->wait.lock);
if (list_empty(&wq->wait.entry)) {
__add_wait_queue(&wq->hash->wait, &wq->wait);
if (!test_bit(hash, &wq->hash->map)) {
__set_current_state(TASK_RUNNING);
list_del_init(&wq->wait.entry);
ret = true;
}
}
spin_unlock_irq(&wq->hash->wait.lock);
return ret;
}
static struct io_wq_work *io_get_next_work(struct io_wq_acct *acct,
struct io_worker *worker)
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
__must_hold(acct->lock)
2019-10-22 09:25:58 -07:00
{
struct io_wq_work_node *node, *prev;
struct io_wq_work *work, *tail;
unsigned int stall_hash = -1U;
struct io_wq *wq = worker->wq;
2019-10-22 09:25:58 -07:00
wq_list_for_each(node, prev, &acct->work_list) {
unsigned int hash;
work = container_of(node, struct io_wq_work, list);
2019-10-22 09:25:58 -07:00
/* not hashed, can run anytime */
if (!io_wq_is_hashed(work)) {
wq_list_del(&acct->work_list, node, prev);
2019-10-22 09:25:58 -07:00
return work;
}
hash = io_get_work_hash(work);
/* all items with this hash lie in [work, tail] */
tail = wq->hash_tail[hash];
/* hashed, can run if not already running */
if (!test_and_set_bit(hash, &wq->hash->map)) {
wq->hash_tail[hash] = NULL;
wq_list_cut(&acct->work_list, &tail->list, prev);
2019-10-22 09:25:58 -07:00
return work;
}
if (stall_hash == -1U)
stall_hash = hash;
/* fast forward to a next hash, for-each will fix up @prev */
node = &tail->list;
}
if (stall_hash != -1U) {
bool unstalled;
/*
* Set this before dropping the lock to avoid racing with new
* work being added and clearing the stalled bit.
*/
set_bit(IO_ACCT_STALLED_BIT, &acct->flags);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_unlock(&acct->lock);
unstalled = io_wait_on_hash(wq, stall_hash);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_lock(&acct->lock);
if (unstalled) {
clear_bit(IO_ACCT_STALLED_BIT, &acct->flags);
if (wq_has_sleeper(&wq->hash->wait))
wake_up(&wq->hash->wait);
}
2019-10-22 09:25:58 -07:00
}
return NULL;
}
static void io_assign_current_work(struct io_worker *worker,
struct io_wq_work *work)
{
if (work) {
io_run_task_work();
cond_resched();
}
raw_spin_lock(&worker->lock);
worker->cur_work = work;
worker->next_work = NULL;
raw_spin_unlock(&worker->lock);
}
/*
* Called with acct->lock held, drops it before returning
*/
static void io_worker_handle_work(struct io_wq_acct *acct,
struct io_worker *worker)
__releases(&acct->lock)
2019-10-22 09:25:58 -07:00
{
struct io_wq *wq = worker->wq;
bool do_kill = test_bit(IO_WQ_BIT_EXIT, &wq->state);
2019-10-22 09:25:58 -07:00
do {
struct io_wq_work *work;
2019-10-22 09:25:58 -07:00
/*
* If we got some work, mark us as busy. If we didn't, but
* the list isn't empty, it means we stalled on hashed work.
* Mark us stalled so we don't keep looking for work when we
* can't make progress, any work completion or insertion will
* clear the stalled flag.
*/
work = io_get_next_work(acct, worker);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_unlock(&acct->lock);
if (work) {
__io_worker_busy(wq, worker);
2019-10-22 09:25:58 -07:00
/*
* Make sure cancelation can find this, even before
* it becomes the active work. That avoids a window
* where the work has been removed from our general
* work list, but isn't yet discoverable as the
* current work item for this worker.
*/
raw_spin_lock(&worker->lock);
worker->next_work = work;
raw_spin_unlock(&worker->lock);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
} else {
2019-10-22 09:25:58 -07:00
break;
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
}
io_assign_current_work(worker, work);
__set_current_state(TASK_RUNNING);
/* handle a whole dependent link */
do {
struct io_wq_work *next_hashed, *linked;
unsigned int hash = io_get_work_hash(work);
next_hashed = wq_next_work(work);
if (unlikely(do_kill) && (work->flags & IO_WQ_WORK_UNBOUND))
work->flags |= IO_WQ_WORK_CANCEL;
wq->do_work(work);
io_assign_current_work(worker, NULL);
linked = wq->free_work(work);
work = next_hashed;
if (!work && linked && !io_wq_is_hashed(linked)) {
work = linked;
linked = NULL;
}
io_assign_current_work(worker, work);
if (linked)
io_wq_enqueue(wq, linked);
if (hash != -1U && !next_hashed) {
/* serialize hash clear with wake_up() */
spin_lock_irq(&wq->hash->wait.lock);
clear_bit(hash, &wq->hash->map);
clear_bit(IO_ACCT_STALLED_BIT, &acct->flags);
spin_unlock_irq(&wq->hash->wait.lock);
if (wq_has_sleeper(&wq->hash->wait))
wake_up(&wq->hash->wait);
}
} while (work);
if (!__io_acct_run_queue(acct))
break;
raw_spin_lock(&acct->lock);
2019-10-22 09:25:58 -07:00
} while (1);
}
static int io_wq_worker(void *data)
2019-10-22 09:25:58 -07:00
{
struct io_worker *worker = data;
struct io_wq_acct *acct = io_wq_get_acct(worker);
struct io_wq *wq = worker->wq;
bool exit_mask = false, last_timeout = false;
char buf[TASK_COMM_LEN];
2019-10-22 09:25:58 -07:00
worker->flags |= (IO_WORKER_F_UP | IO_WORKER_F_RUNNING);
snprintf(buf, sizeof(buf), "iou-wrk-%d", wq->task->pid);
set_task_comm(current, buf);
2019-10-22 09:25:58 -07:00
while (!test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
long ret;
set_current_state(TASK_INTERRUPTIBLE);
/*
* If we have work to do, io_acct_run_queue() returns with
* the acct->lock held. If not, it will drop it.
*/
while (io_acct_run_queue(acct))
io_worker_handle_work(acct, worker);
raw_spin_lock(&wq->lock);
/*
* Last sleep timed out. Exit if we're not the last worker,
* or if someone modified our affinity.
*/
if (last_timeout && (exit_mask || acct->nr_workers > 1)) {
acct->nr_workers--;
raw_spin_unlock(&wq->lock);
__set_current_state(TASK_RUNNING);
break;
}
last_timeout = false;
__io_worker_idle(wq, worker);
raw_spin_unlock(&wq->lock);
if (io_run_task_work())
continue;
ret = schedule_timeout(WORKER_IDLE_TIMEOUT);
if (signal_pending(current)) {
struct ksignal ksig;
if (!get_signal(&ksig))
continue;
break;
}
if (!ret) {
last_timeout = true;
exit_mask = !cpumask_test_cpu(raw_smp_processor_id(),
wq->cpu_mask);
}
2019-10-22 09:25:58 -07:00
}
if (test_bit(IO_WQ_BIT_EXIT, &wq->state) && io_acct_run_queue(acct))
io_worker_handle_work(acct, worker);
2019-10-22 09:25:58 -07:00
io_worker_exit(worker);
return 0;
}
/*
* Called when a worker is scheduled in. Mark us as currently running.
*/
void io_wq_worker_running(struct task_struct *tsk)
{
struct io_worker *worker = tsk->worker_private;
2019-10-22 09:25:58 -07:00
if (!worker)
return;
2019-10-22 09:25:58 -07:00
if (!(worker->flags & IO_WORKER_F_UP))
return;
if (worker->flags & IO_WORKER_F_RUNNING)
return;
worker->flags |= IO_WORKER_F_RUNNING;
io_wq_inc_running(worker);
2019-10-22 09:25:58 -07:00
}
/*
* Called when worker is going to sleep. If there are no workers currently
* running and we have work pending, wake up a free one or create a new one.
2019-10-22 09:25:58 -07:00
*/
void io_wq_worker_sleeping(struct task_struct *tsk)
{
struct io_worker *worker = tsk->worker_private;
2019-10-22 09:25:58 -07:00
if (!worker)
return;
2019-10-22 09:25:58 -07:00
if (!(worker->flags & IO_WORKER_F_UP))
return;
if (!(worker->flags & IO_WORKER_F_RUNNING))
return;
worker->flags &= ~IO_WORKER_F_RUNNING;
io_wq_dec_running(worker);
2019-10-22 09:25:58 -07:00
}
static void io_init_new_worker(struct io_wq *wq, struct io_worker *worker,
struct task_struct *tsk)
{
tsk->worker_private = worker;
worker->task = tsk;
set_cpus_allowed_ptr(tsk, wq->cpu_mask);
raw_spin_lock(&wq->lock);
hlist_nulls_add_head_rcu(&worker->nulls_node, &wq->free_list);
list_add_tail_rcu(&worker->all_list, &wq->all_list);
worker->flags |= IO_WORKER_F_FREE;
raw_spin_unlock(&wq->lock);
wake_up_new_task(tsk);
}
static bool io_wq_work_match_all(struct io_wq_work *work, void *data)
{
return true;
}
static inline bool io_should_retry_thread(long err)
{
/*
* Prevent perpetual task_work retry, if the task (or its group) is
* exiting.
*/
if (fatal_signal_pending(current))
return false;
switch (err) {
case -EAGAIN:
case -ERESTARTSYS:
case -ERESTARTNOINTR:
case -ERESTARTNOHAND:
return true;
default:
return false;
}
}
static void create_worker_cont(struct callback_head *cb)
{
struct io_worker *worker;
struct task_struct *tsk;
struct io_wq *wq;
worker = container_of(cb, struct io_worker, create_work);
clear_bit_unlock(0, &worker->create_state);
wq = worker->wq;
tsk = create_io_thread(io_wq_worker, worker, NUMA_NO_NODE);
if (!IS_ERR(tsk)) {
io_init_new_worker(wq, worker, tsk);
io_worker_release(worker);
return;
} else if (!io_should_retry_thread(PTR_ERR(tsk))) {
struct io_wq_acct *acct = io_wq_get_acct(worker);
atomic_dec(&acct->nr_running);
raw_spin_lock(&wq->lock);
acct->nr_workers--;
if (!acct->nr_workers) {
struct io_cb_cancel_data match = {
.fn = io_wq_work_match_all,
.cancel_all = true,
};
raw_spin_unlock(&wq->lock);
while (io_acct_cancel_pending_work(wq, acct, &match))
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
;
} else {
raw_spin_unlock(&wq->lock);
}
io_worker_ref_put(wq);
io-wq: fix memory leak in create_io_worker() BUG: memory leak unreferenced object 0xffff888126fcd6c0 (size 192): comm "syz-executor.1", pid 11934, jiffies 4294983026 (age 15.690s) backtrace: [<ffffffff81632c91>] kmalloc_node include/linux/slab.h:609 [inline] [<ffffffff81632c91>] kzalloc_node include/linux/slab.h:732 [inline] [<ffffffff81632c91>] create_io_worker+0x41/0x1e0 fs/io-wq.c:739 [<ffffffff8163311e>] io_wqe_create_worker fs/io-wq.c:267 [inline] [<ffffffff8163311e>] io_wqe_enqueue+0x1fe/0x330 fs/io-wq.c:866 [<ffffffff81620b64>] io_queue_async_work+0xc4/0x200 fs/io_uring.c:1473 [<ffffffff8162c59c>] __io_queue_sqe+0x34c/0x510 fs/io_uring.c:6933 [<ffffffff8162c7ab>] io_req_task_submit+0x4b/0xa0 fs/io_uring.c:2233 [<ffffffff8162cb48>] io_async_task_func+0x108/0x1c0 fs/io_uring.c:5462 [<ffffffff816259e3>] tctx_task_work+0x1b3/0x3a0 fs/io_uring.c:2158 [<ffffffff81269b43>] task_work_run+0x73/0xb0 kernel/task_work.c:164 [<ffffffff812dcdd1>] tracehook_notify_signal include/linux/tracehook.h:212 [inline] [<ffffffff812dcdd1>] handle_signal_work kernel/entry/common.c:146 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_prepare+0x151/0x180 kernel/entry/common.c:209 [<ffffffff843ff25d>] __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] [<ffffffff843ff25d>] syscall_exit_to_user_mode+0x1d/0x40 kernel/entry/common.c:302 [<ffffffff843fa4a2>] do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 [<ffffffff84600068>] entry_SYSCALL_64_after_hwframe+0x44/0xae when create_io_thread() return error, and not retry, the worker object need to be freed. Reported-by: syzbot+65454c239241d3d647da@syzkaller.appspotmail.com Signed-off-by: Qiang.zhang <qiang.zhang@windriver.com> Link: https://lore.kernel.org/r/20210909115822.181188-1-qiang.zhang@windriver.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-09 04:58:22 -07:00
kfree(worker);
return;
}
/* re-create attempts grab a new worker ref, drop the existing one */
io_worker_release(worker);
schedule_work(&worker->work);
}
static void io_workqueue_create(struct work_struct *work)
{
struct io_worker *worker = container_of(work, struct io_worker, work);
struct io_wq_acct *acct = io_wq_get_acct(worker);
if (!io_queue_worker_create(worker, acct, create_worker_cont))
io-wq: fix memory leak in create_io_worker() BUG: memory leak unreferenced object 0xffff888126fcd6c0 (size 192): comm "syz-executor.1", pid 11934, jiffies 4294983026 (age 15.690s) backtrace: [<ffffffff81632c91>] kmalloc_node include/linux/slab.h:609 [inline] [<ffffffff81632c91>] kzalloc_node include/linux/slab.h:732 [inline] [<ffffffff81632c91>] create_io_worker+0x41/0x1e0 fs/io-wq.c:739 [<ffffffff8163311e>] io_wqe_create_worker fs/io-wq.c:267 [inline] [<ffffffff8163311e>] io_wqe_enqueue+0x1fe/0x330 fs/io-wq.c:866 [<ffffffff81620b64>] io_queue_async_work+0xc4/0x200 fs/io_uring.c:1473 [<ffffffff8162c59c>] __io_queue_sqe+0x34c/0x510 fs/io_uring.c:6933 [<ffffffff8162c7ab>] io_req_task_submit+0x4b/0xa0 fs/io_uring.c:2233 [<ffffffff8162cb48>] io_async_task_func+0x108/0x1c0 fs/io_uring.c:5462 [<ffffffff816259e3>] tctx_task_work+0x1b3/0x3a0 fs/io_uring.c:2158 [<ffffffff81269b43>] task_work_run+0x73/0xb0 kernel/task_work.c:164 [<ffffffff812dcdd1>] tracehook_notify_signal include/linux/tracehook.h:212 [inline] [<ffffffff812dcdd1>] handle_signal_work kernel/entry/common.c:146 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_prepare+0x151/0x180 kernel/entry/common.c:209 [<ffffffff843ff25d>] __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] [<ffffffff843ff25d>] syscall_exit_to_user_mode+0x1d/0x40 kernel/entry/common.c:302 [<ffffffff843fa4a2>] do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 [<ffffffff84600068>] entry_SYSCALL_64_after_hwframe+0x44/0xae when create_io_thread() return error, and not retry, the worker object need to be freed. Reported-by: syzbot+65454c239241d3d647da@syzkaller.appspotmail.com Signed-off-by: Qiang.zhang <qiang.zhang@windriver.com> Link: https://lore.kernel.org/r/20210909115822.181188-1-qiang.zhang@windriver.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-09 04:58:22 -07:00
kfree(worker);
}
static bool create_io_worker(struct io_wq *wq, int index)
{
struct io_wq_acct *acct = &wq->acct[index];
struct io_worker *worker;
struct task_struct *tsk;
__set_current_state(TASK_RUNNING);
worker = kzalloc(sizeof(*worker), GFP_KERNEL);
if (!worker) {
fail:
atomic_dec(&acct->nr_running);
raw_spin_lock(&wq->lock);
acct->nr_workers--;
raw_spin_unlock(&wq->lock);
io_worker_ref_put(wq);
return false;
}
refcount_set(&worker->ref, 1);
worker->wq = wq;
raw_spin_lock_init(&worker->lock);
init_completion(&worker->ref_done);
if (index == IO_WQ_ACCT_BOUND)
worker->flags |= IO_WORKER_F_BOUND;
tsk = create_io_thread(io_wq_worker, worker, NUMA_NO_NODE);
if (!IS_ERR(tsk)) {
io_init_new_worker(wq, worker, tsk);
} else if (!io_should_retry_thread(PTR_ERR(tsk))) {
io-wq: fix memory leak in create_io_worker() BUG: memory leak unreferenced object 0xffff888126fcd6c0 (size 192): comm "syz-executor.1", pid 11934, jiffies 4294983026 (age 15.690s) backtrace: [<ffffffff81632c91>] kmalloc_node include/linux/slab.h:609 [inline] [<ffffffff81632c91>] kzalloc_node include/linux/slab.h:732 [inline] [<ffffffff81632c91>] create_io_worker+0x41/0x1e0 fs/io-wq.c:739 [<ffffffff8163311e>] io_wqe_create_worker fs/io-wq.c:267 [inline] [<ffffffff8163311e>] io_wqe_enqueue+0x1fe/0x330 fs/io-wq.c:866 [<ffffffff81620b64>] io_queue_async_work+0xc4/0x200 fs/io_uring.c:1473 [<ffffffff8162c59c>] __io_queue_sqe+0x34c/0x510 fs/io_uring.c:6933 [<ffffffff8162c7ab>] io_req_task_submit+0x4b/0xa0 fs/io_uring.c:2233 [<ffffffff8162cb48>] io_async_task_func+0x108/0x1c0 fs/io_uring.c:5462 [<ffffffff816259e3>] tctx_task_work+0x1b3/0x3a0 fs/io_uring.c:2158 [<ffffffff81269b43>] task_work_run+0x73/0xb0 kernel/task_work.c:164 [<ffffffff812dcdd1>] tracehook_notify_signal include/linux/tracehook.h:212 [inline] [<ffffffff812dcdd1>] handle_signal_work kernel/entry/common.c:146 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] [<ffffffff812dcdd1>] exit_to_user_mode_prepare+0x151/0x180 kernel/entry/common.c:209 [<ffffffff843ff25d>] __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] [<ffffffff843ff25d>] syscall_exit_to_user_mode+0x1d/0x40 kernel/entry/common.c:302 [<ffffffff843fa4a2>] do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 [<ffffffff84600068>] entry_SYSCALL_64_after_hwframe+0x44/0xae when create_io_thread() return error, and not retry, the worker object need to be freed. Reported-by: syzbot+65454c239241d3d647da@syzkaller.appspotmail.com Signed-off-by: Qiang.zhang <qiang.zhang@windriver.com> Link: https://lore.kernel.org/r/20210909115822.181188-1-qiang.zhang@windriver.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-09 04:58:22 -07:00
kfree(worker);
goto fail;
} else {
INIT_WORK(&worker->work, io_workqueue_create);
schedule_work(&worker->work);
}
return true;
2019-10-22 09:25:58 -07:00
}
io-wq: fix use-after-free in io_wq_worker_running The smart syzbot has found a reproducer for the following issue: ================================================================== BUG: KASAN: use-after-free in instrument_atomic_write include/linux/instrumented.h:71 [inline] BUG: KASAN: use-after-free in atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] BUG: KASAN: use-after-free in io_wqe_inc_running fs/io-wq.c:301 [inline] BUG: KASAN: use-after-free in io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 Write of size 4 at addr ffff8882183db08c by task io_wqe_worker-0/7771 CPU: 0 PID: 7771 Comm: io_wqe_worker-0 Not tainted 5.9.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x198/0x1fd lib/dump_stack.c:118 print_address_description.constprop.0.cold+0xae/0x497 mm/kasan/report.c:383 __kasan_report mm/kasan/report.c:513 [inline] kasan_report.cold+0x1f/0x37 mm/kasan/report.c:530 check_memory_region_inline mm/kasan/generic.c:186 [inline] check_memory_region+0x13d/0x180 mm/kasan/generic.c:192 instrument_atomic_write include/linux/instrumented.h:71 [inline] atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] io_wqe_inc_running fs/io-wq.c:301 [inline] io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 schedule_timeout+0x148/0x250 kernel/time/timer.c:1879 io_wqe_worker+0x517/0x10e0 fs/io-wq.c:580 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 Allocated by task 7768: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track mm/kasan/common.c:56 [inline] __kasan_kmalloc.constprop.0+0xbf/0xd0 mm/kasan/common.c:461 kmem_cache_alloc_node_trace+0x17b/0x3f0 mm/slab.c:3594 kmalloc_node include/linux/slab.h:572 [inline] kzalloc_node include/linux/slab.h:677 [inline] io_wq_create+0x57b/0xa10 fs/io-wq.c:1064 io_init_wq_offload fs/io_uring.c:7432 [inline] io_sq_offload_start fs/io_uring.c:7504 [inline] io_uring_create fs/io_uring.c:8625 [inline] io_uring_setup+0x1836/0x28e0 fs/io_uring.c:8694 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Freed by task 21: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track+0x1c/0x30 mm/kasan/common.c:56 kasan_set_free_info+0x1b/0x30 mm/kasan/generic.c:355 __kasan_slab_free+0xd8/0x120 mm/kasan/common.c:422 __cache_free mm/slab.c:3418 [inline] kfree+0x10e/0x2b0 mm/slab.c:3756 __io_wq_destroy fs/io-wq.c:1138 [inline] io_wq_destroy+0x2af/0x460 fs/io-wq.c:1146 io_finish_async fs/io_uring.c:6836 [inline] io_ring_ctx_free fs/io_uring.c:7870 [inline] io_ring_exit_work+0x1e4/0x6d0 fs/io_uring.c:7954 process_one_work+0x94c/0x1670 kernel/workqueue.c:2269 worker_thread+0x64c/0x1120 kernel/workqueue.c:2415 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 The buggy address belongs to the object at ffff8882183db000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 140 bytes inside of 1024-byte region [ffff8882183db000, ffff8882183db400) The buggy address belongs to the page: page:000000009bada22b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x2183db flags: 0x57ffe0000000200(slab) raw: 057ffe0000000200 ffffea0008604c48 ffffea00086a8648 ffff8880aa040700 raw: 0000000000000000 ffff8882183db000 0000000100000002 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff8882183daf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff8882183db000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff8882183db080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff8882183db100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8882183db180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== which is down to the comment below, /* all workers gone, wq exit can proceed */ if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs)) complete(&wqe->wq->done); because there might be multiple cases of wqe in a wq and we would wait for every worker in every wqe to go home before releasing wq's resources on destroying. To that end, rework wq's refcount by making it independent of the tracking of workers because after all they are two different things, and keeping it balanced when workers come and go. Note the manager kthread, like other workers, now holds a grab to wq during its lifetime. Finally to help destroy wq, check IO_WQ_BIT_EXIT upon creating worker and do nothing for exiting wq. Cc: stable@vger.kernel.org # v5.5+ Reported-by: syzbot+45fa0a195b941764e0f0@syzkaller.appspotmail.com Reported-by: syzbot+9af99580130003da82b1@syzkaller.appspotmail.com Cc: Pavel Begunkov <asml.silence@gmail.com> Signed-off-by: Hillf Danton <hdanton@sina.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2020-09-26 06:26:55 -07:00
/*
* Iterate the passed in list and call the specific function for each
* worker that isn't exiting
*/
static bool io_wq_for_each_worker(struct io_wq *wq,
io-wq: fix use-after-free in io_wq_worker_running The smart syzbot has found a reproducer for the following issue: ================================================================== BUG: KASAN: use-after-free in instrument_atomic_write include/linux/instrumented.h:71 [inline] BUG: KASAN: use-after-free in atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] BUG: KASAN: use-after-free in io_wqe_inc_running fs/io-wq.c:301 [inline] BUG: KASAN: use-after-free in io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 Write of size 4 at addr ffff8882183db08c by task io_wqe_worker-0/7771 CPU: 0 PID: 7771 Comm: io_wqe_worker-0 Not tainted 5.9.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x198/0x1fd lib/dump_stack.c:118 print_address_description.constprop.0.cold+0xae/0x497 mm/kasan/report.c:383 __kasan_report mm/kasan/report.c:513 [inline] kasan_report.cold+0x1f/0x37 mm/kasan/report.c:530 check_memory_region_inline mm/kasan/generic.c:186 [inline] check_memory_region+0x13d/0x180 mm/kasan/generic.c:192 instrument_atomic_write include/linux/instrumented.h:71 [inline] atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] io_wqe_inc_running fs/io-wq.c:301 [inline] io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 schedule_timeout+0x148/0x250 kernel/time/timer.c:1879 io_wqe_worker+0x517/0x10e0 fs/io-wq.c:580 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 Allocated by task 7768: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track mm/kasan/common.c:56 [inline] __kasan_kmalloc.constprop.0+0xbf/0xd0 mm/kasan/common.c:461 kmem_cache_alloc_node_trace+0x17b/0x3f0 mm/slab.c:3594 kmalloc_node include/linux/slab.h:572 [inline] kzalloc_node include/linux/slab.h:677 [inline] io_wq_create+0x57b/0xa10 fs/io-wq.c:1064 io_init_wq_offload fs/io_uring.c:7432 [inline] io_sq_offload_start fs/io_uring.c:7504 [inline] io_uring_create fs/io_uring.c:8625 [inline] io_uring_setup+0x1836/0x28e0 fs/io_uring.c:8694 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Freed by task 21: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track+0x1c/0x30 mm/kasan/common.c:56 kasan_set_free_info+0x1b/0x30 mm/kasan/generic.c:355 __kasan_slab_free+0xd8/0x120 mm/kasan/common.c:422 __cache_free mm/slab.c:3418 [inline] kfree+0x10e/0x2b0 mm/slab.c:3756 __io_wq_destroy fs/io-wq.c:1138 [inline] io_wq_destroy+0x2af/0x460 fs/io-wq.c:1146 io_finish_async fs/io_uring.c:6836 [inline] io_ring_ctx_free fs/io_uring.c:7870 [inline] io_ring_exit_work+0x1e4/0x6d0 fs/io_uring.c:7954 process_one_work+0x94c/0x1670 kernel/workqueue.c:2269 worker_thread+0x64c/0x1120 kernel/workqueue.c:2415 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 The buggy address belongs to the object at ffff8882183db000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 140 bytes inside of 1024-byte region [ffff8882183db000, ffff8882183db400) The buggy address belongs to the page: page:000000009bada22b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x2183db flags: 0x57ffe0000000200(slab) raw: 057ffe0000000200 ffffea0008604c48 ffffea00086a8648 ffff8880aa040700 raw: 0000000000000000 ffff8882183db000 0000000100000002 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff8882183daf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff8882183db000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff8882183db080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff8882183db100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8882183db180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== which is down to the comment below, /* all workers gone, wq exit can proceed */ if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs)) complete(&wqe->wq->done); because there might be multiple cases of wqe in a wq and we would wait for every worker in every wqe to go home before releasing wq's resources on destroying. To that end, rework wq's refcount by making it independent of the tracking of workers because after all they are two different things, and keeping it balanced when workers come and go. Note the manager kthread, like other workers, now holds a grab to wq during its lifetime. Finally to help destroy wq, check IO_WQ_BIT_EXIT upon creating worker and do nothing for exiting wq. Cc: stable@vger.kernel.org # v5.5+ Reported-by: syzbot+45fa0a195b941764e0f0@syzkaller.appspotmail.com Reported-by: syzbot+9af99580130003da82b1@syzkaller.appspotmail.com Cc: Pavel Begunkov <asml.silence@gmail.com> Signed-off-by: Hillf Danton <hdanton@sina.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2020-09-26 06:26:55 -07:00
bool (*func)(struct io_worker *, void *),
void *data)
{
struct io_worker *worker;
bool ret = false;
list_for_each_entry_rcu(worker, &wq->all_list, all_list) {
io-wq: fix use-after-free in io_wq_worker_running The smart syzbot has found a reproducer for the following issue: ================================================================== BUG: KASAN: use-after-free in instrument_atomic_write include/linux/instrumented.h:71 [inline] BUG: KASAN: use-after-free in atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] BUG: KASAN: use-after-free in io_wqe_inc_running fs/io-wq.c:301 [inline] BUG: KASAN: use-after-free in io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 Write of size 4 at addr ffff8882183db08c by task io_wqe_worker-0/7771 CPU: 0 PID: 7771 Comm: io_wqe_worker-0 Not tainted 5.9.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x198/0x1fd lib/dump_stack.c:118 print_address_description.constprop.0.cold+0xae/0x497 mm/kasan/report.c:383 __kasan_report mm/kasan/report.c:513 [inline] kasan_report.cold+0x1f/0x37 mm/kasan/report.c:530 check_memory_region_inline mm/kasan/generic.c:186 [inline] check_memory_region+0x13d/0x180 mm/kasan/generic.c:192 instrument_atomic_write include/linux/instrumented.h:71 [inline] atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] io_wqe_inc_running fs/io-wq.c:301 [inline] io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 schedule_timeout+0x148/0x250 kernel/time/timer.c:1879 io_wqe_worker+0x517/0x10e0 fs/io-wq.c:580 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 Allocated by task 7768: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track mm/kasan/common.c:56 [inline] __kasan_kmalloc.constprop.0+0xbf/0xd0 mm/kasan/common.c:461 kmem_cache_alloc_node_trace+0x17b/0x3f0 mm/slab.c:3594 kmalloc_node include/linux/slab.h:572 [inline] kzalloc_node include/linux/slab.h:677 [inline] io_wq_create+0x57b/0xa10 fs/io-wq.c:1064 io_init_wq_offload fs/io_uring.c:7432 [inline] io_sq_offload_start fs/io_uring.c:7504 [inline] io_uring_create fs/io_uring.c:8625 [inline] io_uring_setup+0x1836/0x28e0 fs/io_uring.c:8694 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Freed by task 21: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track+0x1c/0x30 mm/kasan/common.c:56 kasan_set_free_info+0x1b/0x30 mm/kasan/generic.c:355 __kasan_slab_free+0xd8/0x120 mm/kasan/common.c:422 __cache_free mm/slab.c:3418 [inline] kfree+0x10e/0x2b0 mm/slab.c:3756 __io_wq_destroy fs/io-wq.c:1138 [inline] io_wq_destroy+0x2af/0x460 fs/io-wq.c:1146 io_finish_async fs/io_uring.c:6836 [inline] io_ring_ctx_free fs/io_uring.c:7870 [inline] io_ring_exit_work+0x1e4/0x6d0 fs/io_uring.c:7954 process_one_work+0x94c/0x1670 kernel/workqueue.c:2269 worker_thread+0x64c/0x1120 kernel/workqueue.c:2415 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 The buggy address belongs to the object at ffff8882183db000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 140 bytes inside of 1024-byte region [ffff8882183db000, ffff8882183db400) The buggy address belongs to the page: page:000000009bada22b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x2183db flags: 0x57ffe0000000200(slab) raw: 057ffe0000000200 ffffea0008604c48 ffffea00086a8648 ffff8880aa040700 raw: 0000000000000000 ffff8882183db000 0000000100000002 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff8882183daf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff8882183db000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff8882183db080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff8882183db100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8882183db180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== which is down to the comment below, /* all workers gone, wq exit can proceed */ if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs)) complete(&wqe->wq->done); because there might be multiple cases of wqe in a wq and we would wait for every worker in every wqe to go home before releasing wq's resources on destroying. To that end, rework wq's refcount by making it independent of the tracking of workers because after all they are two different things, and keeping it balanced when workers come and go. Note the manager kthread, like other workers, now holds a grab to wq during its lifetime. Finally to help destroy wq, check IO_WQ_BIT_EXIT upon creating worker and do nothing for exiting wq. Cc: stable@vger.kernel.org # v5.5+ Reported-by: syzbot+45fa0a195b941764e0f0@syzkaller.appspotmail.com Reported-by: syzbot+9af99580130003da82b1@syzkaller.appspotmail.com Cc: Pavel Begunkov <asml.silence@gmail.com> Signed-off-by: Hillf Danton <hdanton@sina.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2020-09-26 06:26:55 -07:00
if (io_worker_get(worker)) {
/* no task if node is/was offline */
if (worker->task)
ret = func(worker, data);
io_worker_release(worker);
if (ret)
break;
}
}
return ret;
}
static bool io_wq_worker_wake(struct io_worker *worker, void *data)
{
__set_notify_signal(worker->task);
io-wq: fix use-after-free in io_wq_worker_running The smart syzbot has found a reproducer for the following issue: ================================================================== BUG: KASAN: use-after-free in instrument_atomic_write include/linux/instrumented.h:71 [inline] BUG: KASAN: use-after-free in atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] BUG: KASAN: use-after-free in io_wqe_inc_running fs/io-wq.c:301 [inline] BUG: KASAN: use-after-free in io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 Write of size 4 at addr ffff8882183db08c by task io_wqe_worker-0/7771 CPU: 0 PID: 7771 Comm: io_wqe_worker-0 Not tainted 5.9.0-rc4-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:77 [inline] dump_stack+0x198/0x1fd lib/dump_stack.c:118 print_address_description.constprop.0.cold+0xae/0x497 mm/kasan/report.c:383 __kasan_report mm/kasan/report.c:513 [inline] kasan_report.cold+0x1f/0x37 mm/kasan/report.c:530 check_memory_region_inline mm/kasan/generic.c:186 [inline] check_memory_region+0x13d/0x180 mm/kasan/generic.c:192 instrument_atomic_write include/linux/instrumented.h:71 [inline] atomic_inc include/asm-generic/atomic-instrumented.h:240 [inline] io_wqe_inc_running fs/io-wq.c:301 [inline] io_wq_worker_running+0xde/0x110 fs/io-wq.c:613 schedule_timeout+0x148/0x250 kernel/time/timer.c:1879 io_wqe_worker+0x517/0x10e0 fs/io-wq.c:580 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 Allocated by task 7768: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track mm/kasan/common.c:56 [inline] __kasan_kmalloc.constprop.0+0xbf/0xd0 mm/kasan/common.c:461 kmem_cache_alloc_node_trace+0x17b/0x3f0 mm/slab.c:3594 kmalloc_node include/linux/slab.h:572 [inline] kzalloc_node include/linux/slab.h:677 [inline] io_wq_create+0x57b/0xa10 fs/io-wq.c:1064 io_init_wq_offload fs/io_uring.c:7432 [inline] io_sq_offload_start fs/io_uring.c:7504 [inline] io_uring_create fs/io_uring.c:8625 [inline] io_uring_setup+0x1836/0x28e0 fs/io_uring.c:8694 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46 entry_SYSCALL_64_after_hwframe+0x44/0xa9 Freed by task 21: kasan_save_stack+0x1b/0x40 mm/kasan/common.c:48 kasan_set_track+0x1c/0x30 mm/kasan/common.c:56 kasan_set_free_info+0x1b/0x30 mm/kasan/generic.c:355 __kasan_slab_free+0xd8/0x120 mm/kasan/common.c:422 __cache_free mm/slab.c:3418 [inline] kfree+0x10e/0x2b0 mm/slab.c:3756 __io_wq_destroy fs/io-wq.c:1138 [inline] io_wq_destroy+0x2af/0x460 fs/io-wq.c:1146 io_finish_async fs/io_uring.c:6836 [inline] io_ring_ctx_free fs/io_uring.c:7870 [inline] io_ring_exit_work+0x1e4/0x6d0 fs/io_uring.c:7954 process_one_work+0x94c/0x1670 kernel/workqueue.c:2269 worker_thread+0x64c/0x1120 kernel/workqueue.c:2415 kthread+0x3b5/0x4a0 kernel/kthread.c:292 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294 The buggy address belongs to the object at ffff8882183db000 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 140 bytes inside of 1024-byte region [ffff8882183db000, ffff8882183db400) The buggy address belongs to the page: page:000000009bada22b refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x2183db flags: 0x57ffe0000000200(slab) raw: 057ffe0000000200 ffffea0008604c48 ffffea00086a8648 ffff8880aa040700 raw: 0000000000000000 ffff8882183db000 0000000100000002 0000000000000000 page dumped because: kasan: bad access detected Memory state around the buggy address: ffff8882183daf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ffff8882183db000: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb >ffff8882183db080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ^ ffff8882183db100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8882183db180: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ================================================================== which is down to the comment below, /* all workers gone, wq exit can proceed */ if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs)) complete(&wqe->wq->done); because there might be multiple cases of wqe in a wq and we would wait for every worker in every wqe to go home before releasing wq's resources on destroying. To that end, rework wq's refcount by making it independent of the tracking of workers because after all they are two different things, and keeping it balanced when workers come and go. Note the manager kthread, like other workers, now holds a grab to wq during its lifetime. Finally to help destroy wq, check IO_WQ_BIT_EXIT upon creating worker and do nothing for exiting wq. Cc: stable@vger.kernel.org # v5.5+ Reported-by: syzbot+45fa0a195b941764e0f0@syzkaller.appspotmail.com Reported-by: syzbot+9af99580130003da82b1@syzkaller.appspotmail.com Cc: Pavel Begunkov <asml.silence@gmail.com> Signed-off-by: Hillf Danton <hdanton@sina.com> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2020-09-26 06:26:55 -07:00
wake_up_process(worker->task);
return false;
}
static void io_run_cancel(struct io_wq_work *work, struct io_wq *wq)
{
do {
work->flags |= IO_WQ_WORK_CANCEL;
wq->do_work(work);
work = wq->free_work(work);
} while (work);
}
static void io_wq_insert_work(struct io_wq *wq, struct io_wq_work *work)
{
struct io_wq_acct *acct = io_work_get_acct(wq, work);
unsigned int hash;
struct io_wq_work *tail;
if (!io_wq_is_hashed(work)) {
append:
wq_list_add_tail(&work->list, &acct->work_list);
return;
}
hash = io_get_work_hash(work);
tail = wq->hash_tail[hash];
wq->hash_tail[hash] = work;
if (!tail)
goto append;
wq_list_add_after(&work->list, &tail->list, &acct->work_list);
}
io-wq: fix cancellation on create-worker failure WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 req_ref_put_and_test fs/io_uring.c:1151 [inline] WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 req_ref_put_and_test fs/io_uring.c:1146 [inline] WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 io_req_complete_post+0xf5b/0x1190 fs/io_uring.c:1794 Modules linked in: Call Trace: tctx_task_work+0x1e5/0x570 fs/io_uring.c:2158 task_work_run+0xe0/0x1a0 kernel/task_work.c:164 tracehook_notify_signal include/linux/tracehook.h:212 [inline] handle_signal_work kernel/entry/common.c:146 [inline] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] exit_to_user_mode_prepare+0x232/0x2a0 kernel/entry/common.c:209 __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302 do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x44/0xae When io_wqe_enqueue() -> io_wqe_create_worker() fails, we can't just call io_run_cancel() to clean up the request, it's already enqueued via io_wqe_insert_work() and will be executed either by some other worker during cancellation (e.g. in io_wq_put_and_exit()). Reported-by: Hao Sun <sunhao.th@gmail.com> Fixes: 3146cba99aa28 ("io-wq: make worker creation resilient against signals") Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/93b9de0fcf657affab0acfd675d4abcd273ee863.1631092071.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-08 02:09:29 -07:00
static bool io_wq_work_match_item(struct io_wq_work *work, void *data)
{
return work == data;
}
void io_wq_enqueue(struct io_wq *wq, struct io_wq_work *work)
2019-10-22 09:25:58 -07:00
{
struct io_wq_acct *acct = io_work_get_acct(wq, work);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
struct io_cb_cancel_data match;
unsigned work_flags = work->flags;
bool do_create;
2019-10-22 09:25:58 -07:00
/*
* If io-wq is exiting for this task, or if the request has explicitly
* been marked as one that should not get executed, cancel it here.
*/
if (test_bit(IO_WQ_BIT_EXIT, &wq->state) ||
(work->flags & IO_WQ_WORK_CANCEL)) {
io_run_cancel(work, wq);
return;
}
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_lock(&acct->lock);
io_wq_insert_work(wq, work);
clear_bit(IO_ACCT_STALLED_BIT, &acct->flags);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_unlock(&acct->lock);
rcu_read_lock();
do_create = !io_wq_activate_free_worker(wq, acct);
rcu_read_unlock();
if (do_create && ((work_flags & IO_WQ_WORK_CONCURRENT) ||
!atomic_read(&acct->nr_running))) {
bool did_create;
did_create = io_wq_create_worker(wq, acct);
io-wq: fix cancellation on create-worker failure WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 req_ref_put_and_test fs/io_uring.c:1151 [inline] WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 req_ref_put_and_test fs/io_uring.c:1146 [inline] WARNING: CPU: 0 PID: 10392 at fs/io_uring.c:1151 io_req_complete_post+0xf5b/0x1190 fs/io_uring.c:1794 Modules linked in: Call Trace: tctx_task_work+0x1e5/0x570 fs/io_uring.c:2158 task_work_run+0xe0/0x1a0 kernel/task_work.c:164 tracehook_notify_signal include/linux/tracehook.h:212 [inline] handle_signal_work kernel/entry/common.c:146 [inline] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] exit_to_user_mode_prepare+0x232/0x2a0 kernel/entry/common.c:209 __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302 do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x44/0xae When io_wqe_enqueue() -> io_wqe_create_worker() fails, we can't just call io_run_cancel() to clean up the request, it's already enqueued via io_wqe_insert_work() and will be executed either by some other worker during cancellation (e.g. in io_wq_put_and_exit()). Reported-by: Hao Sun <sunhao.th@gmail.com> Fixes: 3146cba99aa28 ("io-wq: make worker creation resilient against signals") Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/93b9de0fcf657affab0acfd675d4abcd273ee863.1631092071.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-08 02:09:29 -07:00
if (likely(did_create))
return;
raw_spin_lock(&wq->lock);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
if (acct->nr_workers) {
raw_spin_unlock(&wq->lock);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
return;
}
raw_spin_unlock(&wq->lock);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
/* fatal condition, failed to create the first worker */
match.fn = io_wq_work_match_item,
match.data = work,
match.cancel_all = false,
io_acct_cancel_pending_work(wq, acct, &match);
}
2019-10-22 09:25:58 -07:00
}
/*
* Work items that hash to the same value will not be done in parallel.
* Used to limit concurrent writes, generally hashed by inode.
2019-10-22 09:25:58 -07:00
*/
void io_wq_hash_work(struct io_wq_work *work, void *val)
2019-10-22 09:25:58 -07:00
{
unsigned int bit;
2019-10-22 09:25:58 -07:00
bit = hash_ptr(val, IO_WQ_HASH_ORDER);
work->flags |= (IO_WQ_WORK_HASHED | (bit << IO_WQ_HASH_SHIFT));
}
static bool __io_wq_worker_cancel(struct io_worker *worker,
struct io_cb_cancel_data *match,
struct io_wq_work *work)
{
if (work && match->fn(work, match->data)) {
work->flags |= IO_WQ_WORK_CANCEL;
__set_notify_signal(worker->task);
return true;
}
return false;
}
static bool io_wq_worker_cancel(struct io_worker *worker, void *data)
{
struct io_cb_cancel_data *match = data;
/*
* Hold the lock to avoid ->cur_work going out of scope, caller
* may dereference the passed in work.
*/
raw_spin_lock(&worker->lock);
if (__io_wq_worker_cancel(worker, match, worker->cur_work) ||
__io_wq_worker_cancel(worker, match, worker->next_work))
match->nr_running++;
raw_spin_unlock(&worker->lock);
2019-10-22 09:25:58 -07:00
return match->nr_running && !match->cancel_all;
2019-10-22 09:25:58 -07:00
}
static inline void io_wq_remove_pending(struct io_wq *wq,
struct io_wq_work *work,
struct io_wq_work_node *prev)
{
struct io_wq_acct *acct = io_work_get_acct(wq, work);
unsigned int hash = io_get_work_hash(work);
struct io_wq_work *prev_work = NULL;
if (io_wq_is_hashed(work) && work == wq->hash_tail[hash]) {
if (prev)
prev_work = container_of(prev, struct io_wq_work, list);
if (prev_work && io_get_work_hash(prev_work) == hash)
wq->hash_tail[hash] = prev_work;
else
wq->hash_tail[hash] = NULL;
}
wq_list_del(&acct->work_list, &work->list, prev);
}
static bool io_acct_cancel_pending_work(struct io_wq *wq,
struct io_wq_acct *acct,
struct io_cb_cancel_data *match)
2019-10-22 09:25:58 -07:00
{
struct io_wq_work_node *node, *prev;
2019-10-22 09:25:58 -07:00
struct io_wq_work *work;
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_lock(&acct->lock);
wq_list_for_each(node, prev, &acct->work_list) {
work = container_of(node, struct io_wq_work, list);
if (!match->fn(work, match->data))
continue;
io_wq_remove_pending(wq, work, prev);
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_unlock(&acct->lock);
io_run_cancel(work, wq);
match->nr_pending++;
/* not safe to continue after unlock */
return true;
}
io-wq: decouple work_list protection from the big wqe->lock wqe->lock is abused, it now protects acct->work_list, hash stuff, nr_workers, wqe->free_list and so on. Lets first get the work_list out of the wqe-lock mess by introduce a specific lock for work list. This is the first step to solve the huge contension between work insertion and work consumption. good thing: - split locking for bound and unbound work list - reduce contension between work_list visit and (worker's)free_list. For the hash stuff, since there won't be a work with same file in both bound and unbound work list, thus they won't visit same hash entry. it works well to use the new lock to protect hash stuff. Results: set max_unbound_worker = 4, test with echo-server: nice -n -15 ./io_uring_echo_server -p 8081 -f -n 1000 -l 16 (-n connection, -l workload) before this patch: Samples: 2M of event 'cycles:ppp', Event count (approx.): 1239982111074 Overhead Command Shared Object Symbol 28.59% iou-wrk-10021 [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 8.89% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpath 6.20% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock 2.45% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 2.36% iou-wrk-10021 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.29% iou-wrk-10021 [kernel.vmlinux] [k] io_worker_handle_work 1.29% io_uring_echo_s [kernel.vmlinux] [k] io_wqe_enqueue 1.06% iou-wrk-10021 [kernel.vmlinux] [k] io_wqe_worker 1.06% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.03% iou-wrk-10021 [kernel.vmlinux] [k] __schedule 0.99% iou-wrk-10021 [kernel.vmlinux] [k] tcp_sendmsg_locked with this patch: Samples: 1M of event 'cycles:ppp', Event count (approx.): 708446691943 Overhead Command Shared Object Symbol 16.86% iou-wrk-10893 [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 9.10% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock 4.53% io_uring_echo_s [kernel.vmlinux] [k] native_queued_spin_lock_slowpat 2.87% iou-wrk-10893 [kernel.vmlinux] [k] io_worker_handle_work 2.57% iou-wrk-10893 [kernel.vmlinux] [k] _raw_spin_lock_irqsave 2.56% io_uring_echo_s [kernel.vmlinux] [k] io_prep_async_work 1.82% io_uring_echo_s [kernel.vmlinux] [k] _raw_spin_lock 1.33% iou-wrk-10893 [kernel.vmlinux] [k] io_wqe_worker 1.26% io_uring_echo_s [kernel.vmlinux] [k] try_to_wake_up spin_lock failure from 25.59% + 8.89% = 34.48% to 16.86% + 4.53% = 21.39% TPS is similar, while cpu usage is from almost 400% to 350% Signed-off-by: Hao Xu <haoxu@linux.alibaba.com> Link: https://lore.kernel.org/r/20220206095241.121485-2-haoxu@linux.alibaba.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2022-02-06 02:52:39 -07:00
raw_spin_unlock(&acct->lock);
return false;
}
static void io_wq_cancel_pending_work(struct io_wq *wq,
struct io_cb_cancel_data *match)
{
int i;
retry:
for (i = 0; i < IO_WQ_ACCT_NR; i++) {
struct io_wq_acct *acct = io_get_acct(wq, i == 0);
if (io_acct_cancel_pending_work(wq, acct, match)) {
if (match->cancel_all)
goto retry;
break;
}
2019-10-22 09:25:58 -07:00
}
}
static void io_wq_cancel_running_work(struct io_wq *wq,
struct io_cb_cancel_data *match)
{
2019-10-22 09:25:58 -07:00
rcu_read_lock();
io_wq_for_each_worker(wq, io_wq_worker_cancel, match);
2019-10-22 09:25:58 -07:00
rcu_read_unlock();
}
enum io_wq_cancel io_wq_cancel_cb(struct io_wq *wq, work_cancel_fn *cancel,
void *data, bool cancel_all)
2019-10-22 09:25:58 -07:00
{
struct io_cb_cancel_data match = {
.fn = cancel,
.data = data,
.cancel_all = cancel_all,
};
2019-10-22 09:25:58 -07:00
/*
* First check pending list, if we're lucky we can just remove it
* from there. CANCEL_OK means that the work is returned as-new,
* no completion will be posted for it.
*
* Then check if a free (going busy) or busy worker has the work
* currently running. If we find it there, we'll return CANCEL_RUNNING
* as an indication that we attempt to signal cancellation. The
* completion will run normally in this case.
*
* Do both of these while holding the wq->lock, to ensure that
* we'll find a work item regardless of state.
*/
io_wq_cancel_pending_work(wq, &match);
if (match.nr_pending && !match.cancel_all)
return IO_WQ_CANCEL_OK;
raw_spin_lock(&wq->lock);
io_wq_cancel_running_work(wq, &match);
raw_spin_unlock(&wq->lock);
if (match.nr_running && !match.cancel_all)
return IO_WQ_CANCEL_RUNNING;
if (match.nr_running)
return IO_WQ_CANCEL_RUNNING;
if (match.nr_pending)
return IO_WQ_CANCEL_OK;
return IO_WQ_CANCEL_NOTFOUND;
2019-10-22 09:25:58 -07:00
}
static int io_wq_hash_wake(struct wait_queue_entry *wait, unsigned mode,
int sync, void *key)
{
struct io_wq *wq = container_of(wait, struct io_wq, wait);
int i;
list_del_init(&wait->entry);
rcu_read_lock();
for (i = 0; i < IO_WQ_ACCT_NR; i++) {
struct io_wq_acct *acct = &wq->acct[i];
if (test_and_clear_bit(IO_ACCT_STALLED_BIT, &acct->flags))
io_wq_activate_free_worker(wq, acct);
}
rcu_read_unlock();
return 1;
}
struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data)
2019-10-22 09:25:58 -07:00
{
int ret, i;
2019-10-22 09:25:58 -07:00
struct io_wq *wq;
if (WARN_ON_ONCE(!data->free_work || !data->do_work))
return ERR_PTR(-EINVAL);
if (WARN_ON_ONCE(!bounded))
return ERR_PTR(-EINVAL);
wq = kzalloc(sizeof(struct io_wq), GFP_KERNEL);
2019-10-22 09:25:58 -07:00
if (!wq)
return ERR_PTR(-ENOMEM);
refcount_inc(&data->hash->refs);
wq->hash = data->hash;
wq->free_work = data->free_work;
wq->do_work = data->do_work;
ret = -ENOMEM;
if (!alloc_cpumask_var(&wq->cpu_mask, GFP_KERNEL))
goto err;
cpumask_copy(wq->cpu_mask, cpu_possible_mask);
wq->acct[IO_WQ_ACCT_BOUND].max_workers = bounded;
wq->acct[IO_WQ_ACCT_UNBOUND].max_workers =
task_rlimit(current, RLIMIT_NPROC);
INIT_LIST_HEAD(&wq->wait.entry);
wq->wait.func = io_wq_hash_wake;
for (i = 0; i < IO_WQ_ACCT_NR; i++) {
struct io_wq_acct *acct = &wq->acct[i];
acct->index = i;
atomic_set(&acct->nr_running, 0);
INIT_WQ_LIST(&acct->work_list);
raw_spin_lock_init(&acct->lock);
2019-10-22 09:25:58 -07:00
}
raw_spin_lock_init(&wq->lock);
INIT_HLIST_NULLS_HEAD(&wq->free_list, 0);
INIT_LIST_HEAD(&wq->all_list);
2019-10-22 09:25:58 -07:00
wq->task = get_task_struct(data->task);
atomic_set(&wq->worker_refs, 1);
init_completion(&wq->worker_done);
io-wq: fully initialize wqe before calling cpuhp_state_add_instance_nocalls() I received a bug report with the following signature: [ 1759.937637] BUG: unable to handle page fault for address: ffffffffffffffe8 [ 1759.944564] #PF: supervisor read access in kernel mode [ 1759.949732] #PF: error_code(0x0000) - not-present page [ 1759.954901] PGD 7ab615067 P4D 7ab615067 PUD 7ab617067 PMD 0 [ 1759.960596] Oops: 0000 1 PREEMPT SMP PTI [ 1759.964804] CPU: 15 PID: 109 Comm: cpuhp/15 Kdump: loaded Tainted: G X ------- — 5.14.0-362.3.1.el9_3.x86_64 #1 [ 1759.976609] Hardware name: HPE ProLiant DL380 Gen10/ProLiant DL380 Gen10, BIOS U30 06/20/2018 [ 1759.985181] RIP: 0010:io_wq_for_each_worker.isra.0+0x24/0xa0 [ 1759.990877] Code: 90 90 90 90 90 90 0f 1f 44 00 00 41 56 41 55 41 54 55 48 8d 6f 78 53 48 8b 47 78 48 39 c5 74 4f 49 89 f5 49 89 d4 48 8d 58 e8 <8b> 13 85 d2 74 32 8d 4a 01 89 d0 f0 0f b1 0b 75 5c 09 ca 78 3d 48 [ 1760.009758] RSP: 0000:ffffb6f403603e20 EFLAGS: 00010286 [ 1760.015013] RAX: 0000000000000000 RBX: ffffffffffffffe8 RCX: 0000000000000000 [ 1760.022188] RDX: ffffb6f403603e50 RSI: ffffffffb11e95b0 RDI: ffff9f73b09e9400 [ 1760.029362] RBP: ffff9f73b09e9478 R08: 000000000000000f R09: 0000000000000000 [ 1760.036536] R10: ffffffffffffff00 R11: ffffb6f403603d80 R12: ffffb6f403603e50 [ 1760.043712] R13: ffffffffb11e95b0 R14: ffffffffb28531e8 R15: ffff9f7a6fbdf548 [ 1760.050887] FS: 0000000000000000(0000) GS:ffff9f7a6fbc0000(0000) knlGS:0000000000000000 [ 1760.059025] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1760.064801] CR2: ffffffffffffffe8 CR3: 00000007ab610002 CR4: 00000000007706e0 [ 1760.071976] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 [ 1760.079150] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 [ 1760.086325] PKRU: 55555554 [ 1760.089044] Call Trace: [ 1760.091501] <TASK> [ 1760.093612] ? show_trace_log_lvl+0x1c4/0x2df [ 1760.097995] ? show_trace_log_lvl+0x1c4/0x2df [ 1760.102377] ? __io_wq_cpu_online+0x54/0xb0 [ 1760.106584] ? __die_body.cold+0x8/0xd [ 1760.110356] ? page_fault_oops+0x134/0x170 [ 1760.114479] ? kernelmode_fixup_or_oops+0x84/0x110 [ 1760.119298] ? exc_page_fault+0xa8/0x150 [ 1760.123247] ? asm_exc_page_fault+0x22/0x30 [ 1760.127458] ? __pfx_io_wq_worker_affinity+0x10/0x10 [ 1760.132453] ? __pfx_io_wq_worker_affinity+0x10/0x10 [ 1760.137446] ? io_wq_for_each_worker.isra.0+0x24/0xa0 [ 1760.142527] __io_wq_cpu_online+0x54/0xb0 [ 1760.146558] cpuhp_invoke_callback+0x109/0x460 [ 1760.151029] ? __pfx_io_wq_cpu_offline+0x10/0x10 [ 1760.155673] ? __pfx_smpboot_thread_fn+0x10/0x10 [ 1760.160320] cpuhp_thread_fun+0x8d/0x140 [ 1760.164266] smpboot_thread_fn+0xd3/0x1a0 [ 1760.168297] kthread+0xdd/0x100 [ 1760.171457] ? __pfx_kthread+0x10/0x10 [ 1760.175225] ret_from_fork+0x29/0x50 [ 1760.178826] </TASK> [ 1760.181022] Modules linked in: rpcsec_gss_krb5 auth_rpcgss nfsv4 dns_resolver nfs lockd grace fscache netfs rfkill sunrpc vfat fat dm_multipath intel_rapl_msr intel_rapl_common isst_if_common ipmi_ssif nfit libnvdimm mgag200 i2c_algo_bit ioatdma drm_shmem_helper drm_kms_helper acpi_ipmi syscopyarea x86_pkg_temp_thermal sysfillrect ipmi_si intel_powerclamp sysimgblt ipmi_devintf coretemp acpi_power_meter ipmi_msghandler rapl pcspkr dca intel_pch_thermal intel_cstate ses lpc_ich intel_uncore enclosure hpilo mei_me mei acpi_tad fuse drm xfs sd_mod sg bnx2x nvme nvme_core crct10dif_pclmul crc32_pclmul nvme_common ghash_clmulni_intel smartpqi tg3 t10_pi mdio uas libcrc32c crc32c_intel scsi_transport_sas usb_storage hpwdt wmi dm_mirror dm_region_hash dm_log dm_mod [ 1760.248623] CR2: ffffffffffffffe8 A cpu hotplug callback was issued before wq->all_list was initialized. This results in a null pointer dereference. The fix is to fully setup the io_wq before calling cpuhp_state_add_instance_nocalls(). Signed-off-by: Jeff Moyer <jmoyer@redhat.com> Link: https://lore.kernel.org/r/x49y1ghnecs.fsf@segfault.boston.devel.redhat.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2023-10-05 10:55:31 -07:00
ret = cpuhp_state_add_instance_nocalls(io_wq_online, &wq->cpuhp_node);
if (ret)
goto err;
return wq;
err:
io_wq_put_hash(data->hash);
free_cpumask_var(wq->cpu_mask);
kfree(wq);
2019-10-22 09:25:58 -07:00
return ERR_PTR(ret);
}
static bool io_task_work_match(struct callback_head *cb, void *data)
{
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
struct io_worker *worker;
io-wq: fix silly logic error in io_task_work_match() We check for the func with an OR condition, which means it always ends up being false and we never match the task_work we want to cancel. In the unexpected case that we do exit with that pending, we can trigger a hang waiting for a worker to exit, but it was never created. syzbot reports that as such: INFO: task syz-executor687:8514 blocked for more than 143 seconds. Not tainted 5.14.0-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:syz-executor687 state:D stack:27296 pid: 8514 ppid: 8479 flags:0x00024004 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0x940/0x26f0 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_wq_exit_workers fs/io-wq.c:1162 [inline] io_wq_put_and_exit+0x40c/0xc70 fs/io-wq.c:1197 io_uring_clean_tctx fs/io_uring.c:9607 [inline] io_uring_cancel_generic+0x5fe/0x740 fs/io_uring.c:9687 io_uring_files_cancel include/linux/io_uring.h:16 [inline] do_exit+0x265/0x2a30 kernel/exit.c:780 do_group_exit+0x125/0x310 kernel/exit.c:922 get_signal+0x47f/0x2160 kernel/signal.c:2868 arch_do_signal_or_restart+0x2a9/0x1c40 arch/x86/kernel/signal.c:865 handle_signal_work kernel/entry/common.c:148 [inline] exit_to_user_mode_loop kernel/entry/common.c:172 [inline] exit_to_user_mode_prepare+0x17d/0x290 kernel/entry/common.c:209 __syscall_exit_to_user_mode_work kernel/entry/common.c:291 [inline] syscall_exit_to_user_mode+0x19/0x60 kernel/entry/common.c:302 do_syscall_64+0x42/0xb0 arch/x86/entry/common.c:86 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x445cd9 RSP: 002b:00007fc657f4b308 EFLAGS: 00000246 ORIG_RAX: 00000000000000ca RAX: 0000000000000001 RBX: 00000000004cb448 RCX: 0000000000445cd9 RDX: 00000000000f4240 RSI: 0000000000000081 RDI: 00000000004cb44c RBP: 00000000004cb440 R08: 000000000000000e R09: 0000000000000000 R10: 0000000000000000 R11: 0000000000000246 R12: 000000000049b154 R13: 0000000000000003 R14: 00007fc657f4b400 R15: 0000000000022000 While in there, also decrement accr->nr_workers. This isn't strictly needed as we're exiting, but let's make sure the accounting matches up. Fixes: 3146cba99aa2 ("io-wq: make worker creation resilient against signals") Reported-by: syzbot+f62d3e0a4ea4f38f5326@syzkaller.appspotmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-09-08 18:57:26 -07:00
if (cb->func != create_worker_cb && cb->func != create_worker_cont)
return false;
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
worker = container_of(cb, struct io_worker, create_work);
return worker->wq == data;
}
void io_wq_exit_start(struct io_wq *wq)
{
set_bit(IO_WQ_BIT_EXIT, &wq->state);
}
static void io_wq_cancel_tw_create(struct io_wq *wq)
{
struct callback_head *cb;
while ((cb = task_work_cancel_match(wq->task, io_task_work_match, wq)) != NULL) {
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
struct io_worker *worker;
io-wq: remove GFP_ATOMIC allocation off schedule out path Daniel reports that the v5.14-rc4-rt4 kernel throws a BUG when running stress-ng: | [ 90.202543] BUG: sleeping function called from invalid context at kernel/locking/spinlock_rt.c:35 | [ 90.202549] in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 2047, name: iou-wrk-2041 | [ 90.202555] CPU: 5 PID: 2047 Comm: iou-wrk-2041 Tainted: G W 5.14.0-rc4-rt4+ #89 | [ 90.202559] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014 | [ 90.202561] Call Trace: | [ 90.202577] dump_stack_lvl+0x34/0x44 | [ 90.202584] ___might_sleep.cold+0x87/0x94 | [ 90.202588] rt_spin_lock+0x19/0x70 | [ 90.202593] ___slab_alloc+0xcb/0x7d0 | [ 90.202598] ? newidle_balance.constprop.0+0xf5/0x3b0 | [ 90.202603] ? dequeue_entity+0xc3/0x290 | [ 90.202605] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202610] ? pick_next_task_fair+0xb9/0x330 | [ 90.202612] ? __schedule+0x670/0x1410 | [ 90.202615] ? io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202618] kmem_cache_alloc_trace+0x79/0x1f0 | [ 90.202621] io_wqe_dec_running.isra.0+0x98/0xe0 | [ 90.202625] io_wq_worker_sleeping+0x37/0x50 | [ 90.202628] schedule+0x30/0xd0 | [ 90.202630] schedule_timeout+0x8f/0x1a0 | [ 90.202634] ? __bpf_trace_tick_stop+0x10/0x10 | [ 90.202637] io_wqe_worker+0xfd/0x320 | [ 90.202641] ? finish_task_switch.isra.0+0xd3/0x290 | [ 90.202644] ? io_worker_handle_work+0x670/0x670 | [ 90.202646] ? io_worker_handle_work+0x670/0x670 | [ 90.202649] ret_from_fork+0x22/0x30 which is due to the RT kernel not liking a GFP_ATOMIC allocation inside a raw spinlock. Besides that not working on RT, doing any kind of allocation from inside schedule() is kind of nasty and should be avoided if at all possible. This particular path happens when an io-wq worker goes to sleep, and we need a new worker to handle pending work. We currently allocate a small data item to hold the information we need to create a new worker, but we can instead include this data in the io_worker struct itself and just protect it with a single bit lock. We only really need one per worker anyway, as we will have run pending work between to sleep cycles. https://lore.kernel.org/lkml/20210804082418.fbibprcwtzyt5qax@beryllium.lan/ Reported-by: Daniel Wagner <dwagner@suse.de> Tested-by: Daniel Wagner <dwagner@suse.de> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-08-04 07:37:25 -07:00
worker = container_of(cb, struct io_worker, create_work);
io-wq: remove worker to owner tw dependency INFO: task iou-wrk-6609:6612 blocked for more than 143 seconds. Not tainted 5.15.0-rc5-syzkaller #0 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. task:iou-wrk-6609 state:D stack:27944 pid: 6612 ppid: 6526 flags:0x00004006 Call Trace: context_switch kernel/sched/core.c:4940 [inline] __schedule+0xb44/0x5960 kernel/sched/core.c:6287 schedule+0xd3/0x270 kernel/sched/core.c:6366 schedule_timeout+0x1db/0x2a0 kernel/time/timer.c:1857 do_wait_for_common kernel/sched/completion.c:85 [inline] __wait_for_common kernel/sched/completion.c:106 [inline] wait_for_common kernel/sched/completion.c:117 [inline] wait_for_completion+0x176/0x280 kernel/sched/completion.c:138 io_worker_exit fs/io-wq.c:183 [inline] io_wqe_worker+0x66d/0xc40 fs/io-wq.c:597 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295 io-wq worker may submit a task_work to the master task and upon io_worker_exit() wait for the tw to get executed. The problem appears when the master task is waiting in coredump.c: 468 freezer_do_not_count(); 469 wait_for_completion(&core_state->startup); 470 freezer_count(); Apparently having some dependency on children threads getting everything stuck. Workaround it by cancelling the taks_work callback that causes it before going into io_worker_exit() waiting. p.s. probably a better option is to not submit tw elevating the refcount in the first place, but let's leave this excercise for the future. Cc: stable@vger.kernel.org Reported-and-tested-by: syzbot+27d62ee6f256b186883e@syzkaller.appspotmail.com Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> Link: https://lore.kernel.org/r/142a716f4ed936feae868959059154362bfa8c19.1635509451.git.asml.silence@gmail.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-10-29 05:11:33 -07:00
io_worker_cancel_cb(worker);
/*
* Only the worker continuation helper has worker allocated and
* hence needs freeing.
*/
if (cb->func == create_worker_cont)
kfree(worker);
}
}
static void io_wq_exit_workers(struct io_wq *wq)
{
if (!wq->task)
return;
io_wq_cancel_tw_create(wq);
rcu_read_lock();
io_wq_for_each_worker(wq, io_wq_worker_wake, NULL);
rcu_read_unlock();
io_worker_ref_put(wq);
wait_for_completion(&wq->worker_done);
io-wq: Fix UAF when wakeup wqe in hash waitqueue BUG: KASAN: use-after-free in __wake_up_common+0x637/0x650 Read of size 8 at addr ffff8880304250d8 by task iou-wrk-28796/28802 Call Trace: __dump_stack [inline] dump_stack+0x141/0x1d7 print_address_description.constprop.0.cold+0x5b/0x2c6 __kasan_report [inline] kasan_report.cold+0x7c/0xd8 __wake_up_common+0x637/0x650 __wake_up_common_lock+0xd0/0x130 io_worker_handle_work+0x9dd/0x1790 io_wqe_worker+0xb2a/0xd40 ret_from_fork+0x1f/0x30 Allocated by task 28798: kzalloc_node [inline] io_wq_create+0x3c4/0xdd0 io_init_wq_offload [inline] io_uring_alloc_task_context+0x1bf/0x6b0 __io_uring_add_task_file+0x29a/0x3c0 io_uring_add_task_file [inline] io_uring_install_fd [inline] io_uring_create [inline] io_uring_setup+0x209a/0x2bd0 do_syscall_64+0x3a/0xb0 entry_SYSCALL_64_after_hwframe+0x44/0xae Freed by task 28798: kfree+0x106/0x2c0 io_wq_destroy+0x182/0x380 io_wq_put [inline] io_wq_put_and_exit+0x7a/0xa0 io_uring_clean_tctx [inline] __io_uring_cancel+0x428/0x530 io_uring_files_cancel do_exit+0x299/0x2a60 do_group_exit+0x125/0x310 get_signal+0x47f/0x2150 arch_do_signal_or_restart+0x2a8/0x1eb0 handle_signal_work[inline] exit_to_user_mode_loop [inline] exit_to_user_mode_prepare+0x171/0x280 __syscall_exit_to_user_mode_work [inline] syscall_exit_to_user_mode+0x19/0x60 do_syscall_64+0x47/0xb0 entry_SYSCALL_64_after_hwframe There are the following scenarios, hash waitqueue is shared by io-wq1 and io-wq2. (note: wqe is worker) io-wq1:worker2 | locks bit1 io-wq2:worker1 | waits bit1 io-wq1:worker3 | waits bit1 io-wq1:worker2 | completes all wqe bit1 work items io-wq1:worker2 | drop bit1, exit io-wq2:worker1 | locks bit1 io-wq1:worker3 | can not locks bit1, waits bit1 and exit io-wq1 | exit and free io-wq1 io-wq2:worker1 | drops bit1 io-wq1:worker3 | be waked up, even though wqe is freed After all iou-wrk belonging to io-wq1 have exited, remove wqe form hash waitqueue, it is guaranteed that there will be no more wqe belonging to io-wq1 in the hash waitqueue. Reported-by: syzbot+6cb11ade52aa17095297@syzkaller.appspotmail.com Signed-off-by: Zqiang <qiang.zhang@windriver.com> Link: https://lore.kernel.org/r/20210526050826.30500-1-qiang.zhang@windriver.com Signed-off-by: Jens Axboe <axboe@kernel.dk>
2021-05-25 22:08:26 -07:00
spin_lock_irq(&wq->hash->wait.lock);
list_del_init(&wq->wait.entry);
spin_unlock_irq(&wq->hash->wait.lock);
put_task_struct(wq->task);
wq->task = NULL;
}
static void io_wq_destroy(struct io_wq *wq)
2019-10-22 09:25:58 -07:00
{
struct io_cb_cancel_data match = {
.fn = io_wq_work_match_all,
.cancel_all = true,
};
2019-10-22 09:25:58 -07:00
cpuhp_state_remove_instance_nocalls(io_wq_online, &wq->cpuhp_node);
io_wq_cancel_pending_work(wq, &match);
free_cpumask_var(wq->cpu_mask);
io_wq_put_hash(wq->hash);
2019-10-22 09:25:58 -07:00
kfree(wq);
}
void io_wq_put_and_exit(struct io_wq *wq)
{
WARN_ON_ONCE(!test_bit(IO_WQ_BIT_EXIT, &wq->state));
io_wq_exit_workers(wq);
io_wq_destroy(wq);
}
struct online_data {
unsigned int cpu;
bool online;
};
static bool io_wq_worker_affinity(struct io_worker *worker, void *data)
{
struct online_data *od = data;
if (od->online)
cpumask_set_cpu(od->cpu, worker->wq->cpu_mask);
else
cpumask_clear_cpu(od->cpu, worker->wq->cpu_mask);
return false;
}
static int __io_wq_cpu_online(struct io_wq *wq, unsigned int cpu, bool online)
{
struct online_data od = {
.cpu = cpu,
.online = online
};
rcu_read_lock();
io_wq_for_each_worker(wq, io_wq_worker_affinity, &od);
rcu_read_unlock();
return 0;
}
static int io_wq_cpu_online(unsigned int cpu, struct hlist_node *node)
{
struct io_wq *wq = hlist_entry_safe(node, struct io_wq, cpuhp_node);
return __io_wq_cpu_online(wq, cpu, true);
}
static int io_wq_cpu_offline(unsigned int cpu, struct hlist_node *node)
{
struct io_wq *wq = hlist_entry_safe(node, struct io_wq, cpuhp_node);
return __io_wq_cpu_online(wq, cpu, false);
}
int io_wq_cpu_affinity(struct io_uring_task *tctx, cpumask_var_t mask)
{
if (!tctx || !tctx->io_wq)
return -EINVAL;
rcu_read_lock();
if (mask)
cpumask_copy(tctx->io_wq->cpu_mask, mask);
else
cpumask_copy(tctx->io_wq->cpu_mask, cpu_possible_mask);
rcu_read_unlock();
return 0;
}
/*
* Set max number of unbounded workers, returns old value. If new_count is 0,
* then just return the old value.
*/
int io_wq_max_workers(struct io_wq *wq, int *new_count)
{
struct io_wq_acct *acct;
int prev[IO_WQ_ACCT_NR];
int i;
BUILD_BUG_ON((int) IO_WQ_ACCT_BOUND != (int) IO_WQ_BOUND);
BUILD_BUG_ON((int) IO_WQ_ACCT_UNBOUND != (int) IO_WQ_UNBOUND);
BUILD_BUG_ON((int) IO_WQ_ACCT_NR != 2);
for (i = 0; i < IO_WQ_ACCT_NR; i++) {
if (new_count[i] > task_rlimit(current, RLIMIT_NPROC))
new_count[i] = task_rlimit(current, RLIMIT_NPROC);
}
for (i = 0; i < IO_WQ_ACCT_NR; i++)
prev[i] = 0;
rcu_read_lock();
raw_spin_lock(&wq->lock);
for (i = 0; i < IO_WQ_ACCT_NR; i++) {
acct = &wq->acct[i];
prev[i] = max_t(int, acct->max_workers, prev[i]);
if (new_count[i])
acct->max_workers = new_count[i];
}
raw_spin_unlock(&wq->lock);
rcu_read_unlock();
for (i = 0; i < IO_WQ_ACCT_NR; i++)
new_count[i] = prev[i];
return 0;
}
static __init int io_wq_init(void)
{
int ret;
ret = cpuhp_setup_state_multi(CPUHP_AP_ONLINE_DYN, "io-wq/online",
io_wq_cpu_online, io_wq_cpu_offline);
if (ret < 0)
return ret;
io_wq_online = ret;
return 0;
}
subsys_initcall(io_wq_init);