summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
Diffstat
-rw-r--r--fs/ext4/file.c3
-rw-r--r--fs/xfs/xfs_file.c3
-rw-r--r--include/linux/fs.h3
-rw-r--r--include/linux/io_uring_types.h24
-rw-r--r--include/trace/events/io_uring.h15
-rw-r--r--include/uapi/linux/io_uring.h33
-rw-r--r--io_uring/alloc_cache.h39
-rw-r--r--io_uring/filetable.c21
-rw-r--r--io_uring/io-wq.c524
-rw-r--r--io_uring/io_uring.c348
-rw-r--r--io_uring/io_uring.h49
-rw-r--r--io_uring/kbuf.c160
-rw-r--r--io_uring/kbuf.h7
-rw-r--r--io_uring/net.h5
-rw-r--r--io_uring/notif.c8
-rw-r--r--io_uring/notif.h3
-rw-r--r--io_uring/poll.c32
-rw-r--r--io_uring/rsrc.c350
-rw-r--r--io_uring/rsrc.h72
-rw-r--r--io_uring/rw.c8
-rw-r--r--io_uring/timeout.c71
-rw-r--r--io_uring/uring_cmd.c18
22 files changed, 949 insertions, 847 deletions
diff --git a/fs/ext4/file.c b/fs/ext4/file.c
index 0b8b4499e5ca..d101b3b0c7da 100644
--- a/fs/ext4/file.c
+++ b/fs/ext4/file.c
@@ -899,7 +899,8 @@ static int ext4_file_open(struct inode *inode, struct file *filp)
return ret;
}
- filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC;
+ filp->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC |
+ FMODE_DIO_PARALLEL_WRITE;
return dquot_file_open(inode, filp);
}
diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
index 705250f9f90a..863289aaa441 100644
--- a/fs/xfs/xfs_file.c
+++ b/fs/xfs/xfs_file.c
@@ -1171,7 +1171,8 @@ xfs_file_open(
{
if (xfs_is_shutdown(XFS_M(inode->i_sb)))
return -EIO;
- file->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC;
+ file->f_mode |= FMODE_NOWAIT | FMODE_BUF_RASYNC | FMODE_BUF_WASYNC |
+ FMODE_DIO_PARALLEL_WRITE;
return generic_file_open(inode, file);
}
diff --git a/include/linux/fs.h b/include/linux/fs.h
index ef2281a2acce..67495ef79bb2 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -168,6 +168,9 @@ typedef int (dio_iodone_t)(struct kiocb *iocb, loff_t offset,
#define FMODE_NOREUSE ((__force fmode_t)0x800000)
+/* File supports non-exclusive O_DIRECT writes from multiple threads */
+#define FMODE_DIO_PARALLEL_WRITE ((__force fmode_t)0x1000000)
+
/* File was opened by fanotify and shouldn't generate fanotify events */
#define FMODE_NONOTIFY ((__force fmode_t)0x4000000)
diff --git a/include/linux/io_uring_types.h b/include/linux/io_uring_types.h
index 00689c12f6ab..1b2a20a42413 100644
--- a/include/linux/io_uring_types.h
+++ b/include/linux/io_uring_types.h
@@ -188,8 +188,10 @@ struct io_ev_fd {
};
struct io_alloc_cache {
- struct hlist_head list;
+ struct io_wq_work_node list;
unsigned int nr_cached;
+ unsigned int max_cached;
+ size_t elem_size;
};
struct io_ring_ctx {
@@ -239,7 +241,6 @@ struct io_ring_ctx {
* uring_lock, and updated through io_uring_register(2)
*/
struct io_rsrc_node *rsrc_node;
- int rsrc_cached_refs;
atomic_t cancel_seq;
struct io_file_table file_table;
unsigned nr_user_files;
@@ -295,7 +296,7 @@ struct io_ring_ctx {
spinlock_t completion_lock;
bool poll_multi_queue;
- bool cq_waiting;
+ atomic_t cq_wait_nr;
/*
* ->iopoll_list is protected by the ctx->uring_lock for
@@ -325,16 +326,15 @@ struct io_ring_ctx {
struct io_restriction restrictions;
/* slow path rsrc auxilary data, used by update/register */
- struct io_rsrc_node *rsrc_backup_node;
struct io_mapped_ubuf *dummy_ubuf;
struct io_rsrc_data *file_data;
struct io_rsrc_data *buf_data;
- struct delayed_work rsrc_put_work;
- struct callback_head rsrc_put_tw;
- struct llist_head rsrc_put_llist;
+ /* protected by ->uring_lock */
struct list_head rsrc_ref_list;
- spinlock_t rsrc_ref_lock;
+ struct io_alloc_cache rsrc_node_cache;
+ struct wait_queue_head rsrc_quiesce_wq;
+ unsigned rsrc_quiesce;
struct list_head io_buffers_pages;
@@ -366,6 +366,11 @@ struct io_ring_ctx {
unsigned evfd_last_cq_tail;
};
+struct io_tw_state {
+ /* ->uring_lock is taken, callbacks can use io_tw_lock to lock it */
+ bool locked;
+};
+
enum {
REQ_F_FIXED_FILE_BIT = IOSQE_FIXED_FILE_BIT,
REQ_F_IO_DRAIN_BIT = IOSQE_IO_DRAIN_BIT,
@@ -472,7 +477,7 @@ enum {
REQ_F_HASH_LOCKED = BIT(REQ_F_HASH_LOCKED_BIT),
};
-typedef void (*io_req_tw_func_t)(struct io_kiocb *req, bool *locked);
+typedef void (*io_req_tw_func_t)(struct io_kiocb *req, struct io_tw_state *ts);
struct io_task_work {
struct llist_node node;
@@ -562,6 +567,7 @@ struct io_kiocb {
atomic_t refs;
atomic_t poll_refs;
struct io_task_work io_task_work;
+ unsigned nr_tw;
/* for polled requests, i.e. IORING_OP_POLL_ADD and async armed poll */
union {
struct hlist_node hash_node;
diff --git a/include/trace/events/io_uring.h b/include/trace/events/io_uring.h
index 936fd41bf147..69454f1f98b0 100644
--- a/include/trace/events/io_uring.h
+++ b/include/trace/events/io_uring.h
@@ -360,19 +360,18 @@ TRACE_EVENT(io_uring_complete,
);
/**
- * io_uring_submit_sqe - called before submitting one SQE
+ * io_uring_submit_req - called before submitting a request
*
* @req: pointer to a submitted request
- * @force_nonblock: whether a context blocking or not
*
* Allows to track SQE submitting, to understand what was the source of it, SQ
* thread or io_uring_enter call.
*/
-TRACE_EVENT(io_uring_submit_sqe,
+TRACE_EVENT(io_uring_submit_req,
- TP_PROTO(struct io_kiocb *req, bool force_nonblock),
+ TP_PROTO(struct io_kiocb *req),
- TP_ARGS(req, force_nonblock),
+ TP_ARGS(req),
TP_STRUCT__entry (
__field( void *, ctx )
@@ -380,7 +379,6 @@ TRACE_EVENT(io_uring_submit_sqe,
__field( unsigned long long, user_data )
__field( u8, opcode )
__field( u32, flags )
- __field( bool, force_nonblock )
__field( bool, sq_thread )
__string( op_str, io_uring_get_opcode(req->opcode) )
@@ -392,16 +390,15 @@ TRACE_EVENT(io_uring_submit_sqe,
__entry->user_data = req->cqe.user_data;
__entry->opcode = req->opcode;
__entry->flags = req->flags;
- __entry->force_nonblock = force_nonblock;
__entry->sq_thread = req->ctx->flags & IORING_SETUP_SQPOLL;
__assign_str(op_str, io_uring_get_opcode(req->opcode));
),
TP_printk("ring %p, req %p, user_data 0x%llx, opcode %s, flags 0x%x, "
- "non block %d, sq_thread %d", __entry->ctx, __entry->req,
+ "sq_thread %d", __entry->ctx, __entry->req,
__entry->user_data, __get_str(op_str),
- __entry->flags, __entry->force_nonblock, __entry->sq_thread)
+ __entry->flags, __entry->sq_thread)
);
/*
diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h
index 709de6d4feb2..0716cb17e436 100644
--- a/include/uapi/linux/io_uring.h
+++ b/include/uapi/linux/io_uring.h
@@ -250,6 +250,7 @@ enum io_uring_op {
#define IORING_TIMEOUT_REALTIME (1U << 3)
#define IORING_LINK_TIMEOUT_UPDATE (1U << 4)
#define IORING_TIMEOUT_ETIME_SUCCESS (1U << 5)
+#define IORING_TIMEOUT_MULTISHOT (1U << 6)
#define IORING_TIMEOUT_CLOCK_MASK (IORING_TIMEOUT_BOOTTIME | IORING_TIMEOUT_REALTIME)
#define IORING_TIMEOUT_UPDATE_MASK (IORING_TIMEOUT_UPDATE | IORING_LINK_TIMEOUT_UPDATE)
/*
@@ -389,6 +390,9 @@ enum {
#define IORING_OFF_SQ_RING 0ULL
#define IORING_OFF_CQ_RING 0x8000000ULL
#define IORING_OFF_SQES 0x10000000ULL
+#define IORING_OFF_PBUF_RING 0x80000000ULL
+#define IORING_OFF_PBUF_SHIFT 16
+#define IORING_OFF_MMAP_MASK 0xf8000000ULL
/*
* Filled with the offset for mmap(2)
@@ -568,19 +572,6 @@ struct io_uring_rsrc_update2 {
__u32 resv2;
};
-struct io_uring_notification_slot {
- __u64 tag;
- __u64 resv[3];
-};
-
-struct io_uring_notification_register {
- __u32 nr_slots;
- __u32 resv;
- __u64 resv2;
- __u64 data;
- __u64 resv3;
-};
-
/* Skip updating fd indexes set to this value in the fd table */
#define IORING_REGISTER_FILES_SKIP (-2)
@@ -635,12 +626,26 @@ struct io_uring_buf_ring {
};
};
+/*
+ * Flags for IORING_REGISTER_PBUF_RING.
+ *
+ * IOU_PBUF_RING_MMAP: If set, kernel will allocate the memory for the ring.
+ * The application must not set a ring_addr in struct
+ * io_uring_buf_reg, instead it must subsequently call
+ * mmap(2) with the offset set as:
+ * IORING_OFF_PBUF_RING | (bgid << IORING_OFF_PBUF_SHIFT)
+ * to get a virtual mapping for the ring.
+ */
+enum {
+ IOU_PBUF_RING_MMAP = 1,
+};
+
/* argument for IORING_(UN)REGISTER_PBUF_RING */
struct io_uring_buf_reg {
__u64 ring_addr;
__u32 ring_entries;
__u16 bgid;
- __u16 pad;
+ __u16 flags;
__u64 resv[3];
};
diff --git a/io_uring/alloc_cache.h b/io_uring/alloc_cache.h
index c2cde88aeed5..241245cb54a6 100644
--- a/io_uring/alloc_cache.h
+++ b/io_uring/alloc_cache.h
@@ -7,47 +7,60 @@
#define IO_ALLOC_CACHE_MAX 512
struct io_cache_entry {
- struct hlist_node node;
+ struct io_wq_work_node node;
};
static inline bool io_alloc_cache_put(struct io_alloc_cache *cache,
struct io_cache_entry *entry)
{
- if (cache->nr_cached < IO_ALLOC_CACHE_MAX) {
+ if (cache->nr_cached < cache->max_cached) {
cache->nr_cached++;
- hlist_add_head(&entry->node, &cache->list);
+ wq_stack_add_head(&entry->node, &cache->list);
+ /* KASAN poisons object */
+ kasan_slab_free_mempool(entry);
return true;
}
return false;
}
+static inline bool io_alloc_cache_empty(struct io_alloc_cache *cache)
+{
+ return !cache->list.next;
+}
+
static inline struct io_cache_entry *io_alloc_cache_get(struct io_alloc_cache *cache)
{
- if (!hlist_empty(&cache->list)) {
- struct hlist_node *node = cache->list.first;
+ if (cache->list.next) {
+ struct io_cache_entry *entry;
- hlist_del(node);
+ entry = container_of(cache->list.next, struct io_cache_entry, node);
+ kasan_unpoison_range(entry, cache->elem_size);
+ cache->list.next = cache->list.next->next;
cache->nr_cached--;
- return container_of(node, struct io_cache_entry, node);
+ return entry;
}
return NULL;
}
-static inline void io_alloc_cache_init(struct io_alloc_cache *cache)
+static inline void io_alloc_cache_init(struct io_alloc_cache *cache,
+ unsigned max_nr, size_t size)
{
- INIT_HLIST_HEAD(&cache->list);
+ cache->list.next = NULL;
cache->nr_cached = 0;
+ cache->max_cached = max_nr;
+ cache->elem_size = size;
}
static inline void io_alloc_cache_free(struct io_alloc_cache *cache,
void (*free)(struct io_cache_entry *))
{
- while (!hlist_empty(&cache->list)) {
- struct hlist_node *node = cache->list.first;
+ while (1) {
+ struct io_cache_entry *entry = io_alloc_cache_get(cache);
- hlist_del(node);
- free(container_of(node, struct io_cache_entry, node));
+ if (!entry)
+ break;
+ free(entry);
}
cache->nr_cached = 0;
}
diff --git a/io_uring/filetable.c b/io_uring/filetable.c
index b80614e7d605..0f6fa791a47d 100644
--- a/io_uring/filetable.c
+++ b/io_uring/filetable.c
@@ -64,7 +64,6 @@ static int io_install_fixed_file(struct io_ring_ctx *ctx, struct file *file,
u32 slot_index)
__must_hold(&req->ctx->uring_lock)
{
- bool needs_switch = false;
struct io_fixed_file *file_slot;
int ret;
@@ -81,18 +80,13 @@ static int io_install_fixed_file(struct io_ring_ctx *ctx, struct file *file,
if (file_slot->file_ptr) {
struct file *old_file;
- ret = io_rsrc_node_switch_start(ctx);
- if (ret)
- goto err;
-
old_file = (struct file *)(file_slot->file_ptr & FFS_MASK);
- ret = io_queue_rsrc_removal(ctx->file_data, slot_index,
- ctx->rsrc_node, old_file);
+ ret = io_queue_rsrc_removal(ctx->file_data, slot_index, old_file);
if (ret)
- goto err;
+ return ret;
+
file_slot->file_ptr = 0;
io_file_bitmap_clear(&ctx->file_table, slot_index);
- needs_switch = true;
}
ret = io_scm_file_account(ctx, file);
@@ -101,9 +95,6 @@ static int io_install_fixed_file(struct io_ring_ctx *ctx, struct file *file,
io_fixed_file_set(file_slot, file);
io_file_bitmap_set(&ctx->file_table, slot_index);
}
-err:
- if (needs_switch)
- io_rsrc_node_switch(ctx, ctx->file_data);
return ret;
}
@@ -156,9 +147,6 @@ int io_fixed_fd_remove(struct io_ring_ctx *ctx, unsigned int offset)
return -ENXIO;
if (offset >= ctx->nr_user_files)
return -EINVAL;
- ret = io_rsrc_node_switch_start(ctx);
- if (ret)
- return ret;
offset = array_index_nospec(offset, ctx->nr_user_files);
file_slot = io_fixed_file_slot(&ctx->file_table, offset);
@@ -166,13 +154,12 @@ int io_fixed_fd_remove(struct io_ring_ctx *ctx, unsigned int offset)
return -EBADF;
file = (struct file *)(file_slot->file_ptr & FFS_MASK);
- ret = io_queue_rsrc_removal(ctx->file_data, offset, ctx->rsrc_node, file);
+ ret = io_queue_rsrc_removal(ctx->file_data, offset, file);
if (ret)
return ret;
file_slot->file_ptr = 0;
io_file_bitmap_clear(&ctx->file_table, offset);
- io_rsrc_node_switch(ctx, ctx->file_data);
return 0;
}
diff --git a/io_uring/io-wq.c b/io_uring/io-wq.c
index f81c0a7136a5..b2715988791e 100644
--- a/io_uring/io-wq.c
+++ b/io_uring/io-wq.c
@@ -15,6 +15,7 @@
#include <linux/cpu.h>
#include <linux/task_work.h>
#include <linux/audit.h>
+#include <linux/mmu_context.h>
#include <uapi/linux/io_uring.h>
#include "io-wq.h"
@@ -39,7 +40,7 @@ enum {
};
/*
- * One for each thread in a wqe pool
+ * One for each thread in a wq pool
*/
struct io_worker {
refcount_t ref;
@@ -47,7 +48,7 @@ struct io_worker {
struct hlist_nulls_node nulls_node;
struct list_head all_list;
struct task_struct *task;
- struct io_wqe *wqe;
+ struct io_wq *wq;
struct io_wq_work *cur_work;
struct io_wq_work *next_work;
@@ -73,7 +74,7 @@ struct io_worker {
#define IO_WQ_NR_HASH_BUCKETS (1u << IO_WQ_HASH_ORDER)
-struct io_wqe_acct {
+struct io_wq_acct {
unsigned nr_workers;
unsigned max_workers;
int index;
@@ -90,26 +91,6 @@ enum {
};
/*
- * Per-node worker thread pool
- */
-struct io_wqe {
- raw_spinlock_t lock;
- struct io_wqe_acct acct[IO_WQ_ACCT_NR];
-
- int node;
-
- struct hlist_nulls_head free_list;
- struct list_head all_list;
-
- struct wait_queue_entry wait;
-
- struct io_wq *wq;
- struct io_wq_work *hash_tail[IO_WQ_NR_HASH_BUCKETS];
-
- cpumask_var_t cpu_mask;
-};
-
-/*
* Per io_wq state
*/
struct io_wq {
@@ -127,7 +108,19 @@ struct io_wq {
struct task_struct *task;
- struct io_wqe *wqes[];
+ 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;
};
static enum cpuhp_state io_wq_online;
@@ -140,10 +133,10 @@ struct io_cb_cancel_data {
bool cancel_all;
};
-static bool create_io_worker(struct io_wq *wq, struct io_wqe *wqe, int index);
-static void io_wqe_dec_running(struct io_worker *worker);
-static bool io_acct_cancel_pending_work(struct io_wqe *wqe,
- struct io_wqe_acct *acct,
+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);
static void create_worker_cb(struct callback_head *cb);
static void io_wq_cancel_tw_create(struct io_wq *wq);
@@ -159,20 +152,20 @@ static void io_worker_release(struct io_worker *worker)
complete(&worker->ref_done);
}
-static inline struct io_wqe_acct *io_get_acct(struct io_wqe *wqe, bool bound)
+static inline struct io_wq_acct *io_get_acct(struct io_wq *wq, bool bound)
{
- return &wqe->acct[bound ? IO_WQ_ACCT_BOUND : IO_WQ_ACCT_UNBOUND];
+ return &wq->acct[bound ? IO_WQ_ACCT_BOUND : IO_WQ_ACCT_UNBOUND];
}
-static inline struct io_wqe_acct *io_work_get_acct(struct io_wqe *wqe,
- struct io_wq_work *work)
+static inline struct io_wq_acct *io_work_get_acct(struct io_wq *wq,
+ struct io_wq_work *work)
{
- return io_get_acct(wqe, !(work->flags & IO_WQ_WORK_UNBOUND));
+ return io_get_acct(wq, !(work->flags & IO_WQ_WORK_UNBOUND));
}
-static inline struct io_wqe_acct *io_wqe_get_acct(struct io_worker *worker)
+static inline struct io_wq_acct *io_wq_get_acct(struct io_worker *worker)
{
- return io_get_acct(worker->wqe, worker->flags & IO_WORKER_F_BOUND);
+ return io_get_acct(worker->wq, worker->flags & IO_WORKER_F_BOUND);
}
static void io_worker_ref_put(struct io_wq *wq)
@@ -183,14 +176,13 @@ static void io_worker_ref_put(struct io_wq *wq)
static void io_worker_cancel_cb(struct io_worker *worker)
{
- struct io_wqe_acct *acct = io_wqe_get_acct(worker);
- struct io_wqe *wqe = worker->wqe;
- struct io_wq *wq = wqe->wq;
+ struct io_wq_acct *acct = io_wq_get_acct(worker);
+ struct io_wq *wq = worker->wq;
atomic_dec(&acct->nr_running);
- raw_spin_lock(&worker->wqe->lock);
+ raw_spin_lock(&wq->lock);
acct->nr_workers--;
- raw_spin_unlock(&worker->wqe->lock);
+ raw_spin_unlock(&wq->lock);
io_worker_ref_put(wq);
clear_bit_unlock(0, &worker->create_state);
io_worker_release(worker);
@@ -208,8 +200,7 @@ static bool io_task_worker_match(struct callback_head *cb, void *data)
static void io_worker_exit(struct io_worker *worker)
{
- struct io_wqe *wqe = worker->wqe;
- struct io_wq *wq = wqe->wq;
+ struct io_wq *wq = worker->wq;
while (1) {
struct callback_head *cb = task_work_cancel_match(wq->task,
@@ -223,23 +214,23 @@ static void io_worker_exit(struct io_worker *worker)
io_worker_release(worker);
wait_for_completion(&worker->ref_done);
- raw_spin_lock(&wqe->lock);
+ 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(&wqe->lock);
- io_wqe_dec_running(worker);
+ raw_spin_unlock(&wq->lock);
+ io_wq_dec_running(worker);
worker->flags = 0;
preempt_disable();
current->flags &= ~PF_IO_WORKER;
preempt_enable();
kfree_rcu(worker, rcu);
- io_worker_ref_put(wqe->wq);
+ io_worker_ref_put(wq);
do_exit(0);
}
-static inline bool io_acct_run_queue(struct io_wqe_acct *acct)
+static inline bool io_acct_run_queue(struct io_wq_acct *acct)
{
bool ret = false;
@@ -256,8 +247,8 @@ static inline bool io_acct_run_queue(struct io_wqe_acct *acct)
* Check head of free list for an available worker. If one isn't available,
* caller must create one.
*/
-static bool io_wqe_activate_free_worker(struct io_wqe *wqe,
- struct io_wqe_acct *acct)
+static bool io_wq_activate_free_worker(struct io_wq *wq,
+ struct io_wq_acct *acct)
__must_hold(RCU)
{
struct hlist_nulls_node *n;
@@ -268,10 +259,10 @@ static bool io_wqe_activate_free_worker(struct io_wqe *wqe,
* 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, &wqe->free_list, nulls_node) {
+ hlist_nulls_for_each_entry_rcu(worker, n, &wq->free_list, nulls_node) {
if (!io_worker_get(worker))
continue;
- if (io_wqe_get_acct(worker) != acct) {
+ if (io_wq_get_acct(worker) != acct) {
io_worker_release(worker);
continue;
}
@@ -289,7 +280,7 @@ static bool io_wqe_activate_free_worker(struct io_wqe *wqe,
* 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_wqe_create_worker(struct io_wqe *wqe, struct io_wqe_acct *acct)
+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
@@ -298,21 +289,21 @@ static bool io_wqe_create_worker(struct io_wqe *wqe, struct io_wqe_acct *acct)
if (unlikely(!acct->max_workers))
pr_warn_once("io-wq is not configured for unbound workers");
- raw_spin_lock(&wqe->lock);
+ raw_spin_lock(&wq->lock);
if (acct->nr_workers >= acct->max_workers) {
- raw_spin_unlock(&wqe->lock);
+ raw_spin_unlock(&wq->lock);
return true;
}
acct->nr_workers++;
- raw_spin_unlock(&wqe->lock);
+ raw_spin_unlock(&wq->lock);
atomic_inc(&acct->nr_running);
- atomic_inc(&wqe->wq->worker_refs);
- return create_io_worker(wqe->wq, wqe, acct->index);
+ atomic_inc(&wq->worker_refs);
+ return create_io_worker(wq, acct->index);
}
-static void io_wqe_inc_running(struct io_worker *worker)
+static void io_wq_inc_running(struct io_worker *worker)
{
- struct io_wqe_acct *acct = io_wqe_get_acct(worker);
+ struct io_wq_acct *acct = io_wq_get_acct(worker);
atomic_inc(&acct->nr_running);
}
@@ -321,22 +312,22 @@ static void create_worker_cb(struct callback_head *cb)
{
struct io_worker *worker;
struct io_wq *wq;
- struct io_wqe *wqe;
- struct io_wqe_acct *acct;
+
+ struct io_wq_acct *acct;
bool do_create = false;
worker = container_of(cb, struct io_worker, create_work);
- wqe = worker->wqe;
- wq = wqe->wq;
- acct = &wqe->acct[worker->create_index];
- raw_spin_lock(&wqe->lock);
+ 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(&wqe->lock);
+ raw_spin_unlock(&wq->lock);
if (do_create) {
- create_io_worker(wq, wqe, worker->create_index);
+ create_io_worker(wq, worker->create_index);
} else {
atomic_dec(&acct->nr_running);
io_worker_ref_put(wq);
@@ -346,11 +337,10 @@ static void create_worker_cb(struct callback_head *cb)
}
static bool io_queue_worker_create(struct io_worker *worker,
- struct io_wqe_acct *acct,
+ struct io_wq_acct *acct,
task_work_func_t func)
{
- struct io_wqe *wqe = worker->wqe;
- struct io_wq *wq = wqe->wq;
+ struct io_wq *wq = worker->wq;
/* raced with exit, just ignore create call */
if (test_bit(IO_WQ_BIT_EXIT, &wq->state))
@@ -392,10 +382,10 @@ fail:
return false;
}
-static void io_wqe_dec_running(struct io_worker *worker)
+static void io_wq_dec_running(struct io_worker *worker)
{
- struct io_wqe_acct *acct = io_wqe_get_acct(worker);
- struct io_wqe *wqe = worker->wqe;
+ struct io_wq_acct *acct = io_wq_get_acct(worker);
+ struct io_wq *wq = worker->wq;
if (!(worker->flags & IO_WORKER_F_UP))
return;
@@ -406,7 +396,7 @@ static void io_wqe_dec_running(struct io_worker *worker)
return;
atomic_inc(&acct->nr_running);
- atomic_inc(&wqe->wq->worker_refs);
+ atomic_inc(&wq->worker_refs);
io_queue_worker_create(worker, acct, create_worker_cb);
}
@@ -414,29 +404,25 @@ static void io_wqe_dec_running(struct io_worker *worker)
* 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_wqe *wqe, struct io_worker *worker)
+static void __io_worker_busy(struct io_wq *wq, struct io_worker *worker)
{
if (worker->flags & IO_WORKER_F_FREE) {
worker->flags &= ~IO_WORKER_F_FREE;
- raw_spin_lock(&wqe->lock);
+ raw_spin_lock(&wq->lock);
hlist_nulls_del_init_rcu(&worker->nulls_node);
- raw_spin_unlock(&wqe->lock);
+ raw_spin_unlock(&wq->lock);
}
}
/*
- * No work, worker going to sleep. Move to freelist, and unuse mm if we
- * have one attached. Dropping the mm may potentially sleep, so we drop
- * the lock in that case and return success. Since the caller has to
- * retry the loop in that case (we changed task state), we don't regrab
- * the lock if we return success.
+ * No work, worker going to sleep. Move to freelist.
*/
-static void __io_worker_idle(struct io_wqe *wqe, struct io_worker *worker)
- __must_hold(wqe->lock)
+static void __io_worker_idle(struct io_wq *wq, struct io_worker *worker)
+ __must_hold(wq->lock)
{
if (!(worker->flags & IO_WORKER_F_FREE)) {
worker->flags |= IO_WORKER_F_FREE;
- hlist_nulls_add_head_rcu(&worker->nulls_node, &wqe->free_list);
+ hlist_nulls_add_head_rcu(&worker->nulls_node, &wq->free_list);
}
}
@@ -445,17 +431,16 @@ 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_wqe *wqe, unsigned int hash)
+static bool io_wait_on_hash(struct io_wq *wq, unsigned int hash)
{
- struct io_wq *wq = wqe->wq;
bool ret = false;
spin_lock_irq(&wq->hash->wait.lock);
- if (list_empty(&wqe->wait.entry)) {
- __add_wait_queue(&wq->hash->wait, &wqe->wait);
+ 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(&wqe->wait.entry);
+ list_del_init(&wq->wait.entry);
ret = true;
}
}
@@ -463,14 +448,14 @@ static bool io_wait_on_hash(struct io_wqe *wqe, unsigned int hash)
return ret;
}
-static struct io_wq_work *io_get_next_work(struct io_wqe_acct *acct,
+static struct io_wq_work *io_get_next_work(struct io_wq_acct *acct,
struct io_worker *worker)
__must_hold(acct->lock)
{
struct io_wq_work_node *node, *prev;
struct io_wq_work *work, *tail;
unsigned int stall_hash = -1U;
- struct io_wqe *wqe = worker->wqe;
+ struct io_wq *wq = worker->wq;
wq_list_for_each(node, prev, &acct->work_list) {
unsigned int hash;
@@ -485,11 +470,11 @@ static struct io_wq_work *io_get_next_work(struct io_wqe_acct *acct,
hash = io_get_work_hash(work);
/* all items with this hash lie in [work, tail] */
- tail = wqe->hash_tail[hash];
+ tail = wq->hash_tail[hash];
/* hashed, can run if not already running */
- if (!test_and_set_bit(hash, &wqe->wq->hash->map)) {
- wqe->hash_tail[hash] = NULL;
+ if (!test_and_set_bit(hash, &wq->hash->map)) {
+ wq->hash_tail[hash] = NULL;
wq_list_cut(&acct->work_list, &tail->list, prev);
return work;
}
@@ -508,12 +493,12 @@ static struct io_wq_work *io_get_next_work(struct io_wqe_acct *acct,
*/
set_bit(IO_ACCT_STALLED_BIT, &acct->flags);
raw_spin_unlock(&acct->lock);
- unstalled = io_wait_on_hash(wqe, stall_hash);
+ unstalled = io_wait_on_hash(wq, stall_hash);
raw_spin_lock(&acct->lock);
if (unstalled) {
clear_bit(IO_ACCT_STALLED_BIT, &acct->flags);
- if (wq_has_sleeper(&wqe->wq->hash->wait))
- wake_up(&wqe->wq->hash->wait);
+ if (wq_has_sleeper(&wq->hash->wait))
+ wake_up(&wq->hash->wait);
}
}
@@ -534,13 +519,10 @@ static void io_assign_current_work(struct io_worker *worker,
raw_spin_unlock(&worker->lock);
}
-static void io_wqe_enqueue(struct io_wqe *wqe, struct io_wq_work *work);
-
static void io_worker_handle_work(struct io_worker *worker)
{
- struct io_wqe_acct *acct = io_wqe_get_acct(worker);
- struct io_wqe *wqe = worker->wqe;
- struct io_wq *wq = wqe->wq;
+ struct io_wq_acct *acct = io_wq_get_acct(worker);
+ struct io_wq *wq = worker->wq;
bool do_kill = test_bit(IO_WQ_BIT_EXIT, &wq->state);
do {
@@ -557,7 +539,7 @@ static void io_worker_handle_work(struct io_worker *worker)
work = io_get_next_work(acct, worker);
raw_spin_unlock(&acct->lock);
if (work) {
- __io_worker_busy(wqe, worker);
+ __io_worker_busy(wq, worker);
/*
* Make sure cancelation can find this, even before
@@ -595,7 +577,7 @@ static void io_worker_handle_work(struct io_worker *worker)
}
io_assign_current_work(worker, work);
if (linked)
- io_wqe_enqueue(wqe, linked);
+ io_wq_enqueue(wq, linked);
if (hash != -1U && !next_hashed) {
/* serialize hash clear with wake_up() */
@@ -610,12 +592,11 @@ static void io_worker_handle_work(struct io_worker *worker)
} while (1);
}
-static int io_wqe_worker(void *data)
+static int io_wq_worker(void *data)
{
struct io_worker *worker = data;
- struct io_wqe_acct *acct = io_wqe_get_acct(worker);
- struct io_wqe *wqe = worker->wqe;
- struct io_wq *wq = wqe->wq;
+ 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];
@@ -631,20 +612,20 @@ static int io_wqe_worker(void *data)
while (io_acct_run_queue(acct))
io_worker_handle_work(worker);
- raw_spin_lock(&wqe->lock);
+ 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(&wqe->lock);
+ raw_spin_unlock(&wq->lock);
__set_current_state(TASK_RUNNING);
break;
}
last_timeout = false;
- __io_worker_idle(wqe, worker);
- raw_spin_unlock(&wqe->lock);
+ __io_worker_idle(wq, worker);
+ raw_spin_unlock(&wq->lock);
if (io_run_task_work())
continue;
ret = schedule_timeout(WORKER_IDLE_TIMEOUT);
@@ -658,7 +639,7 @@ static int io_wqe_worker(void *data)
if (!ret) {
la