summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
Diffstat
-rw-r--r--fs/exec.c6
-rw-r--r--fs/file.c2
-rw-r--r--fs/io-wq.c200
-rw-r--r--fs/io-wq.h4
-rw-r--r--fs/io_uring.c2181
-rw-r--r--include/linux/fs.h46
-rw-r--r--include/linux/io_uring.h58
-rw-r--r--include/linux/sched.h5
-rw-r--r--include/uapi/linux/io_uring.h61
-rw-r--r--init/init_task.c3
-rw-r--r--kernel/fork.c6
-rw-r--r--net/unix/scm.c1
12 files changed, 1662 insertions, 911 deletions
diff --git a/fs/exec.c b/fs/exec.c
index a91003e28eaa..07910f5032e7 100644
--- a/fs/exec.c
+++ b/fs/exec.c
@@ -62,6 +62,7 @@
#include <linux/oom.h>
#include <linux/compat.h>
#include <linux/vmalloc.h>
+#include <linux/io_uring.h>
#include <linux/uaccess.h>
#include <asm/mmu_context.h>
@@ -1895,6 +1896,11 @@ static int bprm_execve(struct linux_binprm *bprm,
struct files_struct *displaced;
int retval;
+ /*
+ * Cancel any io_uring activity across execve
+ */
+ io_uring_task_cancel();
+
retval = unshare_files(&displaced);
if (retval)
return retval;
diff --git a/fs/file.c b/fs/file.c
index 21c0893f2f1d..4559b5fec3bd 100644
--- a/fs/file.c
+++ b/fs/file.c
@@ -21,6 +21,7 @@
#include <linux/rcupdate.h>
#include <linux/close_range.h>
#include <net/sock.h>
+#include <linux/io_uring.h>
unsigned int sysctl_nr_open __read_mostly = 1024*1024;
unsigned int sysctl_nr_open_min = BITS_PER_LONG;
@@ -452,6 +453,7 @@ void exit_files(struct task_struct *tsk)
struct files_struct * files = tsk->files;
if (files) {
+ io_uring_files_cancel(files);
task_lock(tsk);
tsk->files = NULL;
task_unlock(tsk);
diff --git a/fs/io-wq.c b/fs/io-wq.c
index 414beb543883..0a182f1333e8 100644
--- a/fs/io-wq.c
+++ b/fs/io-wq.c
@@ -17,6 +17,7 @@
#include <linux/rculist_nulls.h>
#include <linux/fs_struct.h>
#include <linux/task_work.h>
+#include <linux/blk-cgroup.h>
#include "io-wq.h"
@@ -26,9 +27,8 @@ 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_EXITING = 8, /* worker exiting */
- IO_WORKER_F_FIXED = 16, /* static idle worker */
- IO_WORKER_F_BOUND = 32, /* is doing bounded work */
+ IO_WORKER_F_FIXED = 8, /* static idle worker */
+ IO_WORKER_F_BOUND = 16, /* is doing bounded work */
};
enum {
@@ -57,9 +57,13 @@ struct io_worker {
struct rcu_head rcu;
struct mm_struct *mm;
+#ifdef CONFIG_BLK_CGROUP
+ struct cgroup_subsys_state *blkcg_css;
+#endif
const struct cred *cur_creds;
const struct cred *saved_creds;
struct files_struct *restore_files;
+ struct nsproxy *restore_nsproxy;
struct fs_struct *restore_fs;
};
@@ -87,7 +91,7 @@ enum {
*/
struct io_wqe {
struct {
- spinlock_t lock;
+ raw_spinlock_t lock;
struct io_wq_work_list work_list;
unsigned long hash_map;
unsigned flags;
@@ -148,11 +152,12 @@ static bool __io_worker_unuse(struct io_wqe *wqe, struct io_worker *worker)
if (current->files != worker->restore_files) {
__acquire(&wqe->lock);
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
dropped_lock = true;
task_lock(current);
current->files = worker->restore_files;
+ current->nsproxy = worker->restore_nsproxy;
task_unlock(current);
}
@@ -166,7 +171,7 @@ static bool __io_worker_unuse(struct io_wqe *wqe, struct io_worker *worker)
if (worker->mm) {
if (!dropped_lock) {
__acquire(&wqe->lock);
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
dropped_lock = true;
}
__set_current_state(TASK_RUNNING);
@@ -175,6 +180,13 @@ static bool __io_worker_unuse(struct io_wqe *wqe, struct io_worker *worker)
worker->mm = NULL;
}
+#ifdef CONFIG_BLK_CGROUP
+ if (worker->blkcg_css) {
+ kthread_associate_blkcg(NULL);
+ worker->blkcg_css = NULL;
+ }
+#endif
+
return dropped_lock;
}
@@ -200,7 +212,6 @@ static void io_worker_exit(struct io_worker *worker)
{
struct io_wqe *wqe = worker->wqe;
struct io_wqe_acct *acct = io_wqe_get_acct(wqe, worker);
- unsigned nr_workers;
/*
* If we're not at zero, someone else is holding a brief reference
@@ -220,23 +231,19 @@ static void io_worker_exit(struct io_worker *worker)
worker->flags = 0;
preempt_enable();
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
hlist_nulls_del_rcu(&worker->nulls_node);
list_del_rcu(&worker->all_list);
if (__io_worker_unuse(wqe, worker)) {
__release(&wqe->lock);
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
}
acct->nr_workers--;
- nr_workers = wqe->acct[IO_WQ_ACCT_BOUND].nr_workers +
- wqe->acct[IO_WQ_ACCT_UNBOUND].nr_workers;
- spin_unlock_irq(&wqe->lock);
-
- /* all workers gone, wq exit can proceed */
- if (!nr_workers && refcount_dec_and_test(&wqe->wq->refs))
- complete(&wqe->wq->done);
+ raw_spin_unlock_irq(&wqe->lock);
kfree_rcu(worker, rcu);
+ if (refcount_dec_and_test(&wqe->wq->refs))
+ complete(&wqe->wq->done);
}
static inline bool io_wqe_run_queue(struct io_wqe *wqe)
@@ -318,6 +325,7 @@ static void io_worker_start(struct io_wqe *wqe, struct io_worker *worker)
worker->flags |= (IO_WORKER_F_UP | IO_WORKER_F_RUNNING);
worker->restore_files = current->files;
+ worker->restore_nsproxy = current->nsproxy;
worker->restore_fs = current->fs;
io_wqe_inc_running(wqe, worker);
}
@@ -436,6 +444,17 @@ static void io_wq_switch_mm(struct io_worker *worker, struct io_wq_work *work)
work->flags |= IO_WQ_WORK_CANCEL;
}
+static inline void io_wq_switch_blkcg(struct io_worker *worker,
+ struct io_wq_work *work)
+{
+#ifdef CONFIG_BLK_CGROUP
+ if (work->blkcg_css != worker->blkcg_css) {
+ kthread_associate_blkcg(work->blkcg_css);
+ worker->blkcg_css = work->blkcg_css;
+ }
+#endif
+}
+
static void io_wq_switch_creds(struct io_worker *worker,
struct io_wq_work *work)
{
@@ -454,6 +473,7 @@ static void io_impersonate_work(struct io_worker *worker,
if (work->files && current->files != work->files) {
task_lock(current);
current->files = work->files;
+ current->nsproxy = work->nsproxy;
task_unlock(current);
}
if (work->fs && current->fs != work->fs)
@@ -463,6 +483,7 @@ static void io_impersonate_work(struct io_worker *worker,
if (worker->cur_creds != work->creds)
io_wq_switch_creds(worker, work);
current->signal->rlim[RLIMIT_FSIZE].rlim_cur = work->fsize;
+ io_wq_switch_blkcg(worker, work);
}
static void io_assign_current_work(struct io_worker *worker,
@@ -504,7 +525,7 @@ get_next:
else if (!wq_list_empty(&wqe->work_list))
wqe->flags |= IO_WQE_FLAG_STALLED;
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
if (!work)
break;
io_assign_current_work(worker, work);
@@ -538,17 +559,17 @@ get_next:
io_wqe_enqueue(wqe, linked);
if (hash != -1U && !next_hashed) {
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
wqe->hash_map &= ~BIT_ULL(hash);
wqe->flags &= ~IO_WQE_FLAG_STALLED;
/* skip unnecessary unlock-lock wqe->lock */
if (!work)
goto get_next;
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
}
} while (work);
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
} while (1);
}
@@ -563,7 +584,7 @@ static int io_wqe_worker(void *data)
while (!test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
set_current_state(TASK_INTERRUPTIBLE);
loop:
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
if (io_wqe_run_queue(wqe)) {
__set_current_state(TASK_RUNNING);
io_worker_handle_work(worker);
@@ -574,7 +595,7 @@ loop:
__release(&wqe->lock);
goto loop;
}
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
if (signal_pending(current))
flush_signals(current);
if (schedule_timeout(WORKER_IDLE_TIMEOUT))
@@ -586,11 +607,11 @@ loop:
}
if (test_bit(IO_WQ_BIT_EXIT, &wq->state)) {
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
if (!wq_list_empty(&wqe->work_list))
io_worker_handle_work(worker);
else
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
}
io_worker_exit(worker);
@@ -630,14 +651,14 @@ void io_wq_worker_sleeping(struct task_struct *tsk)
worker->flags &= ~IO_WORKER_F_RUNNING;
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
io_wqe_dec_running(wqe, worker);
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
}
static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
{
- struct io_wqe_acct *acct =&wqe->acct[index];
+ struct io_wqe_acct *acct = &wqe->acct[index];
struct io_worker *worker;
worker = kzalloc_node(sizeof(*worker), GFP_KERNEL, wqe->node);
@@ -656,7 +677,7 @@ static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
return false;
}
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
list_add_tail_rcu(&worker->all_list, &wqe->all_list);
worker->flags |= IO_WORKER_F_FREE;
@@ -665,11 +686,12 @@ static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index)
if (!acct->nr_workers && (worker->flags & IO_WORKER_F_BOUND))
worker->flags |= IO_WORKER_F_FIXED;
acct->nr_workers++;
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
if (index == IO_WQ_ACCT_UNBOUND)
atomic_inc(&wq->user->processes);
+ refcount_inc(&wq->refs);
wake_up_process(worker->task);
return true;
}
@@ -685,28 +707,63 @@ static inline bool io_wqe_need_worker(struct io_wqe *wqe, int index)
return acct->nr_workers < acct->max_workers;
}
+static bool io_wqe_worker_send_sig(struct io_worker *worker, void *data)
+{
+ send_sig(SIGINT, worker->task, 1);
+ return false;
+}
+
+/*
+ * 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_wqe *wqe,
+ bool (*func)(struct io_worker *, void *),
+ void *data)
+{
+ struct io_worker *worker;
+ bool ret = false;
+
+ list_for_each_entry_rcu(worker, &wqe->all_list, all_list) {
+ 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)
+{
+ wake_up_process(worker->task);
+ return false;
+}
+
/*
* Manager thread. Tasked with creating new workers, if we need them.
*/
static int io_wq_manager(void *data)
{
struct io_wq *wq = data;
- int workers_to_create = num_possible_nodes();
int node;
/* create fixed workers */
- refcount_set(&wq->refs, workers_to_create);
+ refcount_set(&wq->refs, 1);
for_each_node(node) {
if (!node_online(node))
continue;
- if (!create_io_worker(wq, wq->wqes[node], IO_WQ_ACCT_BOUND))
- goto err;
- workers_to_create--;
+ if (create_io_worker(wq, wq->wqes[node], IO_WQ_ACCT_BOUND))
+ continue;
+ set_bit(IO_WQ_BIT_ERROR, &wq->state);
+ set_bit(IO_WQ_BIT_EXIT, &wq->state);
+ goto out;
}
- while (workers_to_create--)
- refcount_dec(&wq->refs);
-
complete(&wq->done);
while (!kthread_should_stop()) {
@@ -720,12 +777,12 @@ static int io_wq_manager(void *data)
if (!node_online(node))
continue;
- spin_lock_irq(&wqe->lock);
+ raw_spin_lock_irq(&wqe->lock);
if (io_wqe_need_worker(wqe, IO_WQ_ACCT_BOUND))
fork_worker[IO_WQ_ACCT_BOUND] = true;
if (io_wqe_need_worker(wqe, IO_WQ_ACCT_UNBOUND))
fork_worker[IO_WQ_ACCT_UNBOUND] = true;
- spin_unlock_irq(&wqe->lock);
+ raw_spin_unlock_irq(&wqe->lock);
if (fork_worker[IO_WQ_ACCT_BOUND])
create_io_worker(wq, wqe, IO_WQ_ACCT_BOUND);
if (fork_worker[IO_WQ_ACCT_UNBOUND])
@@ -738,12 +795,18 @@ static int io_wq_manager(void *data)
if (current->task_works)
task_work_run();
- return 0;
-err:
- set_bit(IO_WQ_BIT_ERROR, &wq->state);
- set_bit(IO_WQ_BIT_EXIT, &wq->state);
- if (refcount_sub_and_test(workers_to_create, &wq->refs))
+out:
+ if (refcount_dec_and_test(&wq->refs)) {
complete(&wq->done);
+ return 0;
+ }
+ /* if ERROR is set and we get here, we have workers to wake */
+ if (test_bit(IO_WQ_BIT_ERROR, &wq->state)) {
+ rcu_read_lock();
+ for_each_node(node)
+ io_wq_for_each_worker(wq->wqes[node], io_wq_worker_wake, NULL);
+ rcu_read_unlock();
+ }
return 0;
}
@@ -821,10 +884,10 @@ static void io_wqe_enqueue(struct io_wqe *wqe, struct io_wq_work *work)
}
work_flags = work->flags;
- spin_lock_irqsave(&wqe->lock, flags);
+ raw_spin_lock_irqsave(&wqe->lock, flags);
io_wqe_insert_work(wqe, work);
wqe->flags &= ~IO_WQE_FLAG_STALLED;
- spin_unlock_irqrestore(&wqe->lock, flags);
+ raw_spin_unlock_irqrestore(&wqe->lock, flags);
if ((work_flags & IO_WQ_WORK_CONCURRENT) ||
!atomic_read(&acct->nr_running))
@@ -850,37 +913,6 @@ void io_wq_hash_work(struct io_wq_work *work, void *val)
work->flags |= (IO_WQ_WORK_HASHED | (bit << IO_WQ_HASH_SHIFT));
}
-static bool io_wqe_worker_send_sig(struct io_worker *worker, void *data)
-{
- send_sig(SIGINT, worker->task, 1);
- return false;
-}
-
-/*
- * 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_wqe *wqe,
- bool (*func)(struct io_worker *, void *),
- void *data)
-{
- struct io_worker *worker;
- bool ret = false;
-
- list_for_each_entry_rcu(worker, &wqe->all_list, all_list) {
- 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;
-}
-
void io_wq_cancel_all(struct io_wq *wq)
{
int node;
@@ -951,13 +983,13 @@ static void io_wqe_cancel_pending_work(struct io_wqe *wqe,
unsigned long flags;
retry:
- spin_lock_irqsave(&wqe->lock, flags);
+ raw_spin_lock_irqsave(&wqe->lock, flags);
wq_list_for_each(node, prev, &wqe->work_list) {
work = container_of(node, struct io_wq_work, list);
if (!match->fn(work, match->data))
continue;
io_wqe_remove_pending(wqe, work, prev);
- spin_unlock_irqrestore(&wqe->lock, flags);
+ raw_spin_unlock_irqrestore(&wqe->lock, flags);
io_run_cancel(work, wqe);
match->nr_pending++;
if (!match->cancel_all)
@@ -966,7 +998,7 @@ retry:
/* not safe to continue after unlock */
goto retry;
}
- spin_unlock_irqrestore(&wqe->lock, flags);
+ raw_spin_unlock_irqrestore(&wqe->lock, flags);
}
static void io_wqe_cancel_running_work(struct io_wqe *wqe,
@@ -1074,7 +1106,7 @@ struct io_wq *io_wq_create(unsigned bounded, struct io_wq_data *data)
}
atomic_set(&wqe->acct[IO_WQ_ACCT_UNBOUND].nr_running, 0);
wqe->wq = wq;
- spin_lock_init(&wqe->lock);
+ raw_spin_lock_init(&wqe->lock);
INIT_WQ_LIST(&wqe->work_list);
INIT_HLIST_NULLS_HEAD(&wqe->free_list, 0);
INIT_LIST_HEAD(&wqe->all_list);
@@ -1113,12 +1145,6 @@ bool io_wq_get(struct io_wq *wq, struct io_wq_data *data)
return refcount_inc_not_zero(&wq->use_refs);
}
-static bool io_wq_worker_wake(struct io_worker *worker, void *data)
-{
- wake_up_process(worker->task);
- return false;
-}
-
static void __io_wq_destroy(struct io_wq *wq)
{
int node;
diff --git a/fs/io-wq.h b/fs/io-wq.h
index ddaf9614cf9b..84bcf6a85523 100644
--- a/fs/io-wq.h
+++ b/fs/io-wq.h
@@ -87,7 +87,11 @@ struct io_wq_work {
struct io_wq_work_node list;
struct files_struct *files;
struct mm_struct *mm;
+#ifdef CONFIG_BLK_CGROUP
+ struct cgroup_subsys_state *blkcg_css;
+#endif
const struct cred *creds;
+ struct nsproxy *nsproxy;
struct fs_struct *fs;
unsigned long fsize;
unsigned flags;
diff --git a/fs/io_uring.c b/fs/io_uring.c
index f58b3d6bba8a..fc6de6b4784e 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -79,6 +79,8 @@
#include <linux/splice.h>
#include <linux/task_work.h>
#include <linux/pagemap.h>
+#include <linux/io_uring.h>
+#include <linux/blk-cgroup.h>
#define CREATE_TRACE_POINTS
#include <trace/events/io_uring.h>
@@ -98,6 +100,8 @@
#define IORING_MAX_FILES_TABLE (1U << IORING_FILE_TABLE_SHIFT)
#define IORING_FILE_TABLE_MASK (IORING_MAX_FILES_TABLE - 1)
#define IORING_MAX_FIXED_FILES (64 * IORING_MAX_FILES_TABLE)
+#define IORING_MAX_RESTRICTIONS (IORING_RESTRICTION_LAST + \
+ IORING_REGISTER_LAST + IORING_OP_LAST)
struct io_uring {
u32 head ____cacheline_aligned_in_smp;
@@ -187,6 +191,7 @@ struct io_mapped_ubuf {
size_t len;
struct bio_vec *bvec;
unsigned int nr_bvecs;
+ unsigned long acct_pages;
};
struct fixed_file_table {
@@ -205,7 +210,7 @@ struct fixed_file_data {
struct fixed_file_table *table;
struct io_ring_ctx *ctx;
- struct percpu_ref *cur_refs;
+ struct fixed_file_ref_node *node;
struct percpu_ref refs;
struct completion done;
struct list_head ref_list;
@@ -219,6 +224,27 @@ struct io_buffer {
__u16 bid;
};
+struct io_restriction {
+ DECLARE_BITMAP(register_op, IORING_REGISTER_LAST);
+ DECLARE_BITMAP(sqe_op, IORING_OP_LAST);
+ u8 sqe_flags_allowed;
+ u8 sqe_flags_required;
+ bool registered;
+};
+
+struct io_sq_data {
+ refcount_t refs;
+ struct mutex lock;
+
+ /* ctx's that are using this sqd */
+ struct list_head ctx_list;
+ struct list_head ctx_new_list;
+ struct mutex ctx_lock;
+
+ struct task_struct *thread;
+ struct wait_queue_head wait;
+};
+
struct io_ring_ctx {
struct {
struct percpu_ref refs;
@@ -231,6 +257,7 @@ struct io_ring_ctx {
unsigned int cq_overflow_flushed: 1;
unsigned int drain_next: 1;
unsigned int eventfd_async: 1;
+ unsigned int restricted: 1;
/*
* Ring buffer of indices into array of io_uring_sqe, which is
@@ -264,9 +291,25 @@ struct io_ring_ctx {
/* IO offload */
struct io_wq *io_wq;
- struct task_struct *sqo_thread; /* if using sq thread polling */
- struct mm_struct *sqo_mm;
- wait_queue_head_t sqo_wait;
+
+ /*
+ * For SQPOLL usage - we hold a reference to the parent task, so we
+ * have access to the ->files
+ */
+ struct task_struct *sqo_task;
+
+ /* Only used for accounting purposes */
+ struct mm_struct *mm_account;
+
+#ifdef CONFIG_BLK_CGROUP
+ struct cgroup_subsys_state *sqo_blkcg_css;
+#endif
+
+ struct io_sq_data *sq_data; /* if using sq thread polling */
+
+ struct wait_queue_head sqo_sq_wait;
+ struct wait_queue_entry sqo_wait_entry;
+ struct list_head sqd_list;
/*
* If used, fixed file set. Writers must ensure that ->refs is dead,
@@ -275,8 +318,6 @@ struct io_ring_ctx {
*/
struct fixed_file_data *file_data;
unsigned nr_user_files;
- int ring_fd;
- struct file *ring_file;
/* if used, fixed mapped user buffers */
unsigned nr_user_bufs;
@@ -338,6 +379,7 @@ struct io_ring_ctx {
struct llist_head file_put_llist;
struct work_struct exit_work;
+ struct io_restriction restrictions;
};
/*
@@ -392,13 +434,16 @@ struct io_cancel {
struct io_timeout {
struct file *file;
- u64 addr;
- int flags;
u32 off;
u32 target_seq;
struct list_head list;
};
+struct io_timeout_rem {
+ struct file *file;
+ u64 addr;
+};
+
struct io_rw {
/* NOTE: kiocb has the file as the first member, so don't do it here */
struct kiocb kiocb;
@@ -514,15 +559,6 @@ struct io_async_rw {
struct wait_page_queue wpq;
};
-struct io_async_ctx {
- union {
- struct io_async_rw rw;
- struct io_async_msghdr msg;
- struct io_async_connect connect;
- struct io_timeout_data timeout;
- };
-};
-
enum {
REQ_F_FIXED_FILE_BIT = IOSQE_FIXED_FILE_BIT,
REQ_F_IO_DRAIN_BIT = IOSQE_IO_DRAIN_BIT,
@@ -544,7 +580,6 @@ enum {
REQ_F_BUFFER_SELECTED_BIT,
REQ_F_NO_FILE_TABLE_BIT,
REQ_F_WORK_INITIALIZED_BIT,
- REQ_F_TASK_PINNED_BIT,
/* not a real bit, just to check we're not overflowing the space */
__REQ_F_LAST_BIT,
@@ -590,8 +625,6 @@ enum {
REQ_F_NO_FILE_TABLE = BIT(REQ_F_NO_FILE_TABLE_BIT),
/* io_wq_work is initialized */
REQ_F_WORK_INITIALIZED = BIT(REQ_F_WORK_INITIALIZED_BIT),
- /* req->task is refcounted */
- REQ_F_TASK_PINNED = BIT(REQ_F_TASK_PINNED_BIT),
};
struct async_poll {
@@ -614,6 +647,7 @@ struct io_kiocb {
struct io_sync sync;
struct io_cancel cancel;
struct io_timeout timeout;
+ struct io_timeout_rem timeout_rem;
struct io_connect connect;
struct io_sr_msg sr_msg;
struct io_open open;
@@ -629,7 +663,8 @@ struct io_kiocb {
struct io_completion compl;
};
- struct io_async_ctx *io;
+ /* opcode allocated if it needs to store data for async defer */
+ void *async_data;
u8 opcode;
/* polled IO has completed */
u8 iopoll_completed;
@@ -697,8 +732,6 @@ struct io_submit_state {
};
struct io_op_def {
- /* needs req->io allocated for deferral/async */
- unsigned async_ctx : 1;
/* needs current->mm setup, does mm access */
unsigned needs_mm : 1;
/* needs req->file assigned */
@@ -720,35 +753,49 @@ struct io_op_def {
unsigned pollout : 1;
/* op supports buffer selection */
unsigned buffer_select : 1;
+ /* needs rlimit(RLIMIT_FSIZE) assigned */
unsigned needs_fsize : 1;
+ /* must always have async data allocated */
+ unsigned needs_async_data : 1;
+ /* needs blkcg context, issues async io potentially */
+ unsigned needs_blkcg : 1;
+ /* size of async data needed, if any */
+ unsigned short async_size;
};
-static const struct io_op_def io_op_defs[] = {
+static const struct io_op_def io_op_defs[] __read_mostly = {
[IORING_OP_NOP] = {},
[IORING_OP_READV] = {
- .async_ctx = 1,
.needs_mm = 1,
.needs_file = 1,
.unbound_nonreg_file = 1,
.pollin = 1,
.buffer_select = 1,
+ .needs_async_data = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_WRITEV] = {
- .async_ctx = 1,
.needs_mm = 1,
.needs_file = 1,
.hash_reg_file = 1,
.unbound_nonreg_file = 1,
.pollout = 1,
.needs_fsize = 1,
+ .needs_async_data = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_FSYNC] = {
.needs_file = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_READ_FIXED] = {
.needs_file = 1,
.unbound_nonreg_file = 1,
.pollin = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_WRITE_FIXED] = {
.needs_file = 1,
@@ -756,6 +803,8 @@ static const struct io_op_def io_op_defs[] = {
.unbound_nonreg_file = 1,
.pollout = 1,
.needs_fsize = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_POLL_ADD] = {
.needs_file = 1,
@@ -764,27 +813,33 @@ static const struct io_op_def io_op_defs[] = {
[IORING_OP_POLL_REMOVE] = {},
[IORING_OP_SYNC_FILE_RANGE] = {
.needs_file = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_SENDMSG] = {
- .async_ctx = 1,
.needs_mm = 1,
.needs_file = 1,
.unbound_nonreg_file = 1,
.needs_fs = 1,
.pollout = 1,
+ .needs_async_data = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_msghdr),
},
[IORING_OP_RECVMSG] = {
- .async_ctx = 1,
.needs_mm = 1,
.needs_file = 1,
.unbound_nonreg_file = 1,
.needs_fs = 1,
.pollin = 1,
.buffer_select = 1,
+ .needs_async_data = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_msghdr),
},
[IORING_OP_TIMEOUT] = {
- .async_ctx = 1,
.needs_mm = 1,
+ .needs_async_data = 1,
+ .async_size = sizeof(struct io_timeout_data),
},
[IORING_OP_TIMEOUT_REMOVE] = {},
[IORING_OP_ACCEPT] = {
@@ -796,28 +851,33 @@ static const struct io_op_def io_op_defs[] = {
},
[IORING_OP_ASYNC_CANCEL] = {},
[IORING_OP_LINK_TIMEOUT] = {
- .async_ctx = 1,
.needs_mm = 1,
+ .needs_async_data = 1,
+ .async_size = sizeof(struct io_timeout_data),
},
[IORING_OP_CONNECT] = {
- .async_ctx = 1,
.needs_mm = 1,
.needs_file = 1,
.unbound_nonreg_file = 1,
.pollout = 1,
+ .needs_async_data = 1,
+ .async_size = sizeof(struct io_async_connect),
},
[IORING_OP_FALLOCATE] = {
.needs_file = 1,
.needs_fsize = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_OPENAT] = {
.file_table = 1,
.needs_fs = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_CLOSE] = {
.needs_file = 1,
.needs_file_no_error = 1,
.file_table = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_FILES_UPDATE] = {
.needs_mm = 1,
@@ -827,6 +887,7 @@ static const struct io_op_def io_op_defs[] = {
.needs_mm = 1,
.needs_fs = 1,
.file_table = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_READ] = {
.needs_mm = 1,
@@ -834,6 +895,8 @@ static const struct io_op_def io_op_defs[] = {
.unbound_nonreg_file = 1,
.pollin = 1,
.buffer_select = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_WRITE] = {
.needs_mm = 1,
@@ -841,18 +904,23 @@ static const struct io_op_def io_op_defs[] = {
.unbound_nonreg_file = 1,
.pollout = 1,
.needs_fsize = 1,
+ .needs_blkcg = 1,
+ .async_size = sizeof(struct io_async_rw),
},
[IORING_OP_FADVISE] = {
.needs_file = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_MADVISE] = {
.needs_mm = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_SEND] = {
.needs_mm = 1,
.needs_file = 1,
.unbound_nonreg_file = 1,
.pollout = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_RECV] = {
.needs_mm = 1,
@@ -860,10 +928,12 @@ static const struct io_op_def io_op_defs[] = {
.unbound_nonreg_file = 1,
.pollin = 1,
.buffer_select = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_OPENAT2] = {
.file_table = 1,
.needs_fs = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_EPOLL_CTL] = {
.unbound_nonreg_file = 1,
@@ -873,6 +943,7 @@ static const struct io_op_def io_op_defs[] = {
.needs_file = 1,
.hash_reg_file = 1,
.unbound_nonreg_file = 1,
+ .needs_blkcg = 1,
},
[IORING_OP_PROVIDE_BUFFERS] = {},
[IORING_OP_REMOVE_BUFFERS] = {},
@@ -900,13 +971,10 @@ static void io_queue_linked_timeout(struct io_kiocb *req);
static int __io_sqe_files_update(struct io_ring_ctx *ctx,
struct io_uring_files_update *ip,
unsigned nr_args);
-static int io_prep_work_files(struct io_kiocb *req);
static void __io_clean_op(struct io_kiocb *req);
-static int io_file_get(struct io_submit_state *state, struct io_kiocb *req,
- int fd, struct file **out_file, bool fixed);
-static void __io_queue_sqe(struct io_kiocb *req,
- const struct io_uring_sqe *sqe,
- struct io_comp_state *cs);
+static struct file *io_file_get(struct io_submit_state *state,
+ struct io_kiocb *req, int fd, bool fixed);
+static void __io_queue_sqe(struct io_kiocb *req, struct io_comp_state *cs);
static void io_file_put_work(struct work_struct *work);
static ssize_t io_import_iovec(int rw, struct io_kiocb *req,
@@ -918,7 +986,7 @@ static int io_setup_async_rw(struct io_kiocb *req, const struct iovec *iovec,
static struct kmem_cache *req_cachep;
-static const struct file_operations io_uring_fops;
+static const struct file_operations io_uring_fops __read_mostly;
struct sock *io_uring_get_socket(struct file *file)
{
@@ -933,14 +1001,6 @@ struct sock *io_uring_get_socket(struct file *file)
}
EXPORT_SYMBOL(io_uring_get_socket);
-static void io_get_req_task(struct io_kiocb *req)
-{
- if (req->flags & REQ_F_TASK_PINNED)
- return;
- get_task_struct(req->task);
- req->flags |= REQ_F_TASK_PINNED;
-}
-
static inline void io_clean_op(struct io_kiocb *req)
{
if (req->flags & (REQ_F_NEED_CLEANUP | REQ_F_BUFFER_SELECTED |
@@ -948,13 +1008,6 @@ static inline void io_clean_op(struct io_kiocb *req)
__io_clean_op(req);
}
-/* not idempotent -- it doesn't clear REQ_F_TASK_PINNED */
-static void __io_put_req_task(struct io_kiocb *req)
-{
- if (req->flags & REQ_F_TASK_PINNED)
- put_task_struct(req->task);
-}
-
static void io_sq_thread_drop_mm(void)
{
struct mm_struct *mm = current->mm;
@@ -969,9 +1022,10 @@ static int __io_sq_thread_acquire_mm(struct io_ring_ctx *ctx)
{
if (!current->mm) {
if (unlikely(!(ctx->flags & IORING_SETUP_SQPOLL) ||
- !mmget_not_zero(ctx->sqo_mm)))
+ !ctx->sqo_task->mm ||
+ !mmget_not_zero(ctx->sqo_task->mm)))
return -EFAULT;
- kthread_use_mm(ctx->sqo_mm);
+ kthread_use_mm(ctx->sqo_task->mm);
}
return 0;
@@ -985,6 +1039,26 @@ static int io_sq_thread_acquire_mm(struct io_ring_ctx *ctx,
return __io_sq_thread_acquire_mm(ctx);
}
+static void io_sq_thread_associate_blkcg(struct io_ring_ctx *ctx,
+ struct cgroup_subsys_state **cur_css)
+
+{
+#ifdef CONFIG_BLK_CGROUP
+ /* puts the old one when swapping */
+ if (*cur_css != ctx->sqo_blkcg_css) {
+ kthread_associate_blkcg(ctx->sqo_blkcg_css);
+ *cur_css = ctx->sqo_blkcg_css;
+ }
+#endif
+}
+
+static void io_sq_thread_unassociate_blkcg(void)
+{
+#ifdef CONFIG_BLK_CGROUP
+ kthread_associate_blkcg(NULL);
+#endif
+}
+
static inline void req_set_fail_links(struct io_kiocb *req)
{
if ((req->flags & (REQ_F_LINK | REQ_F_HARDLINK)) == REQ_F_LINK)
@@ -1054,7 +1128,8 @@ static struct io_ring_ctx *io_ring_ctx_alloc(struct io_uring_params *p)
goto err;
ctx->flags = p->flags;
- init_waitqueue_head(&ctx->sqo_wait);
+ init_waitqueue_head(&ctx->sqo_sq_wait);
+ INIT_LIST_HEAD(&ctx->sqd_list);
init_waitqueue_head(&ctx->cq_wait);
INIT_LIST_HEAD(&ctx->cq_overflow_list);
init_completion(&ctx->ref_comp);
@@ -1121,6 +1196,10 @@ static bool io_req_clean_work(struct io_kiocb *req)
mmdrop(req->work.mm);
req->work.mm = NULL;
}
+#ifdef CONFIG_BLK_CGROUP
+ if (req->work.blkcg_css)
+ css_put(req->work.blkcg_css);
+#endif
if (req->work.creds) {
put_cred(req->work.creds);
req->work.creds = NULL;
@@ -1146,20 +1225,45 @@ static bool io_req_clean_work(struct io_kiocb *req)
static void io_prep_async_work(struct io_kiocb *req)
{
const struct io_op_def *def = &io_op_defs[req->opcode];
+ struct io_ring_ctx *ctx = req->ctx;
io_req_init_async(req);
if (req->flags & REQ_F_ISREG) {
- if (def->hash_reg_file || (req->ctx->flags & IORING_SETUP_IOPOLL))
+ if (def->hash_reg_file || (ctx->flags & IORING_SETUP_IOPOLL))
io_wq_hash_work(&req->work, file_inode(req->file));
} else {
if (def->unbound_nonreg_file)
req->work.flags |= IO_WQ_WORK_UNBOUND;
}
+ if (!req->work.files && io_op_defs[req->opcode].file_table &&
+ !(req->flags & REQ_F_NO_FILE_TABLE)) {
+ req->work.files = get_files_struct(current);
+ get_nsproxy(current->nsproxy);
+ req->work.nsproxy = current->nsproxy;
+ req->flags |= REQ_F_INFLIGHT;
+
+ spin_lock_irq(&ctx->inflight_lock);
+ list_add(&req->inflight_entry, &ctx->inflight_list);
+ spin_unlock_irq(&ctx->inflight_lock);
+ }
if (!req->work.mm && def->needs_mm) {
mmgrab(current->mm);
req->work.mm = current->mm;
}
+#ifdef CONFIG_BLK_CGROUP
+ if (!req->work.blkcg_css && def->needs_blkcg) {
+ rcu_read_lock();
+ req->work.blkcg_css = blkcg_css();
+ /*
+ * This should be rare, either the cgroup is dying or the task
+ * is moving cgroups. Just punt to root for the handful of ios.
+ */
+ if (!css_tryget_online(req->work.blkcg_css))
+ req->work.blkcg_css = NULL;
+ rcu_read_unlock();
+ }
+#endif
if (!req->w