Merge v6.12.34linux-rolling-lts

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

5 files changed, 52 insertions, 22 deletions

diff --git a/io_uring/fdinfo.c b/io_uring/fdinfo.c
index ecdbe473a49f..c6c624eb9866 100644
--- a/io_uring/fdinfo.c
+++ b/io_uring/fdinfo.c
@@ -146,18 +146,26 @@ __cold void io_uring_show_fdinfo(struct seq_file *m, struct file *file)
 
 	if (has_lock && (ctx->flags & IORING_SETUP_SQPOLL)) {
 		struct io_sq_data *sq = ctx->sq_data;
+		struct task_struct *tsk;
 
+		rcu_read_lock();
+		tsk = rcu_dereference(sq->thread);
 		/*
 		 * sq->thread might be NULL if we raced with the sqpoll
 		 * thread termination.
 		 */
-		if (sq->thread) {
+		if (tsk) {
+			get_task_struct(tsk);
+			rcu_read_unlock();
+			getrusage(tsk, RUSAGE_SELF, &sq_usage);
+			put_task_struct(tsk);
 			sq_pid = sq->task_pid;
 			sq_cpu = sq->sq_cpu;
-			getrusage(sq->thread, RUSAGE_SELF, &sq_usage);
 			sq_total_time = (sq_usage.ru_stime.tv_sec * 1000000
 					 + sq_usage.ru_stime.tv_usec);
 			sq_work_time = sq->work_time;
+		} else {
+			rcu_read_unlock();
 		}
 	}
 
diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c
index bd3b3f7a6f6c..64870f51b678 100644
--- a/io_uring/io_uring.c
+++ b/io_uring/io_uring.c
@@ -2916,7 +2916,7 @@ static __cold void io_ring_exit_work(struct work_struct *work)
 			struct task_struct *tsk;
 
 			io_sq_thread_park(sqd);
-			tsk = sqd->thread;
+			tsk = sqpoll_task_locked(sqd);
 			if (tsk && tsk->io_uring && tsk->io_uring->io_wq)
 				io_wq_cancel_cb(tsk->io_uring->io_wq,
 						io_cancel_ctx_cb, ctx, true);
@@ -3153,7 +3153,7 @@ __cold void io_uring_cancel_generic(bool cancel_all, struct io_sq_data *sqd)
 	s64 inflight;
 	DEFINE_WAIT(wait);
 
-	WARN_ON_ONCE(sqd && sqd->thread != current);
+	WARN_ON_ONCE(sqd && sqpoll_task_locked(sqd) != current);
 
 	if (!current->io_uring)
 		return;
diff --git a/io_uring/register.c b/io_uring/register.c
index eca26d4884d9..a325b493ae12 100644
--- a/io_uring/register.c
+++ b/io_uring/register.c
@@ -268,6 +268,8 @@ static __cold int io_register_iowq_max_workers(struct io_ring_ctx *ctx,
 	if (ctx->flags & IORING_SETUP_SQPOLL) {
 		sqd = ctx->sq_data;
 		if (sqd) {
+			struct task_struct *tsk;
+
 			/*
 			 * Observe the correct sqd->lock -> ctx->uring_lock
 			 * ordering. Fine to drop uring_lock here, we hold
@@ -277,8 +279,9 @@ static __cold int io_register_iowq_max_workers(struct io_ring_ctx *ctx,
 			mutex_unlock(&ctx->uring_lock);
 			mutex_lock(&sqd->lock);
 			mutex_lock(&ctx->uring_lock);
-			if (sqd->thread)
-				tctx = sqd->thread->io_uring;
+			tsk = sqpoll_task_locked(sqd);
+			if (tsk)
+				tctx = tsk->io_uring;
 		}
 	} else {
 		tctx = current->io_uring;
diff --git a/io_uring/sqpoll.c b/io_uring/sqpoll.c
index 430922c54168..9a6306894895 100644
--- a/io_uring/sqpoll.c
+++ b/io_uring/sqpoll.c
@@ -30,7 +30,7 @@ enum {
 void io_sq_thread_unpark(struct io_sq_data *sqd)
 	__releases(&sqd->lock)
 {
-	WARN_ON_ONCE(sqd->thread == current);
+	WARN_ON_ONCE(sqpoll_task_locked(sqd) == current);
 
 	/*
 	 * Do the dance but not conditional clear_bit() because it'd race with
@@ -45,24 +45,32 @@ void io_sq_thread_unpark(struct io_sq_data *sqd)
 void io_sq_thread_park(struct io_sq_data *sqd)
 	__acquires(&sqd->lock)
 {
-	WARN_ON_ONCE(data_race(sqd->thread) == current);
+	struct task_struct *tsk;
 
 	atomic_inc(&sqd->park_pending);
 	set_bit(IO_SQ_THREAD_SHOULD_PARK, &sqd->state);
 	mutex_lock(&sqd->lock);
-	if (sqd->thread)
-		wake_up_process(sqd->thread);
+
+	tsk = sqpoll_task_locked(sqd);
+	if (tsk) {
+		WARN_ON_ONCE(tsk == current);
+		wake_up_process(tsk);
+	}
 }
 
 void io_sq_thread_stop(struct io_sq_data *sqd)
 {
-	WARN_ON_ONCE(sqd->thread == current);
+	struct task_struct *tsk;
+
 	WARN_ON_ONCE(test_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state));
 
 	set_bit(IO_SQ_THREAD_SHOULD_STOP, &sqd->state);
 	mutex_lock(&sqd->lock);
-	if (sqd->thread)
-		wake_up_process(sqd->thread);
+	tsk = sqpoll_task_locked(sqd);
+	if (tsk) {
+		WARN_ON_ONCE(tsk == current);
+		wake_up_process(tsk);
+	}
 	mutex_unlock(&sqd->lock);
 	wait_for_completion(&sqd->exited);
 }
@@ -277,7 +285,8 @@ static int io_sq_thread(void *data)
 	/* offload context creation failed, just exit */
 	if (!current->io_uring) {
 		mutex_lock(&sqd->lock);
-		sqd->thread = NULL;
+		rcu_assign_pointer(sqd->thread, NULL);
+		put_task_struct(current);
 		mutex_unlock(&sqd->lock);
 		goto err_out;
 	}
@@ -386,7 +395,8 @@ static int io_sq_thread(void *data)
 		io_sq_tw(&retry_list, UINT_MAX);
 
 	io_uring_cancel_generic(true, sqd);
-	sqd->thread = NULL;
+	rcu_assign_pointer(sqd->thread, NULL);
+	put_task_struct(current);
 	list_for_each_entry(ctx, &sqd->ctx_list, sqd_list)
 		atomic_or(IORING_SQ_NEED_WAKEUP, &ctx->rings->sq_flags);
 	io_run_task_work();
@@ -496,7 +506,10 @@ __cold int io_sq_offload_create(struct io_ring_ctx *ctx,
 			goto err_sqpoll;
 		}
 
-		sqd->thread = tsk;
+		mutex_lock(&sqd->lock);
+		rcu_assign_pointer(sqd->thread, tsk);
+		mutex_unlock(&sqd->lock);
+
 		task_to_put = get_task_struct(tsk);
 		ret = io_uring_alloc_task_context(tsk, ctx);
 		wake_up_new_task(tsk);
@@ -507,9 +520,6 @@ __cold int io_sq_offload_create(struct io_ring_ctx *ctx,
 		ret = -EINVAL;
 		goto err;
 	}
-
-	if (task_to_put)
-		put_task_struct(task_to_put);
 	return 0;
 err_sqpoll:
 	complete(&ctx->sq_data->exited);
@@ -527,10 +537,13 @@ __cold int io_sqpoll_wq_cpu_affinity(struct io_ring_ctx *ctx,
 	int ret = -EINVAL;
 
 	if (sqd) {
+		struct task_struct *tsk;
+
 		io_sq_thread_park(sqd);
 		/* Don't set affinity for a dying thread */
-		if (sqd->thread)
-			ret = io_wq_cpu_affinity(sqd->thread->io_uring, mask);
+		tsk = sqpoll_task_locked(sqd);
+		if (tsk)
+			ret = io_wq_cpu_affinity(tsk->io_uring, mask);
 		io_sq_thread_unpark(sqd);
 	}
 
diff --git a/io_uring/sqpoll.h b/io_uring/sqpoll.h
index 4171666b1cf4..b83dcdec9765 100644
--- a/io_uring/sqpoll.h
+++ b/io_uring/sqpoll.h
@@ -8,7 +8,7 @@ struct io_sq_data {
 	/* ctx's that are using this sqd */
 	struct list_head	ctx_list;
 
-	struct task_struct	*thread;
+	struct task_struct __rcu *thread;
 	struct wait_queue_head	wait;
 
 	unsigned		sq_thread_idle;
@@ -29,3 +29,9 @@ void io_sq_thread_unpark(struct io_sq_data *sqd);
 void io_put_sq_data(struct io_sq_data *sqd);
 void io_sqpoll_wait_sq(struct io_ring_ctx *ctx);
 int io_sqpoll_wq_cpu_affinity(struct io_ring_ctx *ctx, cpumask_var_t mask);
+
+static inline struct task_struct *sqpoll_task_locked(struct io_sq_data *sqd)
+{
+	return rcu_dereference_protected(sqd->thread,
+					 lockdep_is_held(&sqd->lock));
+}