Skip to contentSkip to conversation

SQPOLL cancel files fix

PATCH In review Pavel Begunkov 2 participants 4 messages Feb 10, 2021

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover SQPOLL cancel files fix

2/2 is for the recent syzbot report. Quick and dirty, but easy to backport. I plan to restructure (and clean) task vs files cancel later.

Pavel Begunkov (2):
io_uring: cancel files inflight counting
io_uring; fix files cancel hangs

 fs/io_uring.c | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)
Signature · 1 line
2.24.0
#2 lore raw

PATCH 2/2 io_uring; fix files cancel hangs

We park SQPOLL task before going into io_uring_cancel_files(), so the task won't run task_works including those that might be important for the cancellation passes. In this case it's io_poll_remove_one(), which frees requests via io_put_req_deferred().

Unpark it for while waiting, it's ok as we disable submissions beforehand, so no new will be generated.

INFO: task syz-executor893:8493 blocked for more than 143 seconds.
Call Trace:
 context_switch kernel/sched/core.c:4327 [inline]
 __schedule+0x90c/0x21a0 kernel/sched/core.c:5078
 schedule+0xcf/0x270 kernel/sched/core.c:5157
 io_uring_cancel_files fs/io_uring.c:8912 [inline]
 io_uring_cancel_task_requests+0xe70/0x11a0 fs/io_uring.c:8979
 __io_uring_files_cancel+0x110/0x1b0 fs/io_uring.c:9067
 io_uring_files_cancel include/linux/io_uring.h:51 [inline]
 do_exit+0x2fe/0x2ae0 kernel/exit.c:780
 do_group_exit+0x125/0x310 kernel/exit.c:922
 __do_sys_exit_group kernel/exit.c:933 [inline]
 __se_sys_exit_group kernel/exit.c:931 [inline]
 __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:931
 do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
 entry_SYSCALL_64_after_hwframe+0x44/0xa9
---
 fs/io_uring.c | 5 +++++
 1 file changed, 5 insertions(+)
fs/io_uring.c +5 -0
index 6b73e38aa1a9..1e803a9afc8e 100644
@@ -9056,11 +9056,16 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 			break;
 
 		io_uring_try_cancel_requests(ctx, task, files);
+
+		if (ctx->sq_data)
+			io_sq_thread_unpark(ctx->sq_data);
 		prepare_to_wait(&task->io_uring->wait, &wait,
 				TASK_UNINTERRUPTIBLE);
 		if (inflight == io_uring_count_inflight(ctx, task, files))
 			schedule();
 		finish_wait(&task->io_uring->wait, &wait);
+		if (ctx->sq_data)
+			io_sq_thread_park(ctx->sq_data);
 	}
 }
 
Signature · 1 line
2.24.0
#1 lore raw

PATCH 1/2 io_uring: cancel files inflight counting

In reply to Pavel Begunkov

To not skip events, io_uring_cancel_files() is pretty much counts on io_uring_count_inflight() to be monotonous for the time it's used. That's not the case when it includes requests of other thats that are PF_EXITING.

Cancel as before, but don't account extra in io_uring_count_inflight(), we can hang otherwise.

---
 fs/io_uring.c | 16 +++++++++++-----
 1 file changed, 11 insertions(+), 5 deletions(-)
fs/io_uring.c +11 -5
index 17194e0d62ff..6b73e38aa1a9 100644
@@ -1090,15 +1090,14 @@ static inline void io_set_resource_node(struct io_kiocb *req)
 	}
 }
 
-static bool io_match_task(struct io_kiocb *head,
-			  struct task_struct *task,
-			  struct files_struct *files)
+static bool __io_match_task(struct io_kiocb *head, struct task_struct *task,
+			    struct files_struct *files, bool match_exiting)
 {
 	struct io_kiocb *req;
 
 	if (task && head->task != task) {
 		/* in terms of cancelation, always match if req task is dead */
-		if (head->task->flags & PF_EXITING)
+		if (match_exiting && (head->task->flags & PF_EXITING))
 			return true;
 		return false;
 	}
@@ -1117,6 +1116,13 @@ static bool io_match_task(struct io_kiocb *head,
 	return false;
 }
 
+static bool io_match_task(struct io_kiocb *head,
+			  struct task_struct *task,
+			  struct files_struct *files)
+{
+	return __io_match_task(head, task, files, true);
+}
+
 static void io_sq_thread_drop_mm_files(void)
 {
 	struct files_struct *files = current->files;
@@ -9032,7 +9038,7 @@ static int io_uring_count_inflight(struct io_ring_ctx *ctx,
 
 	spin_lock_irq(&ctx->inflight_lock);
 	list_for_each_entry(req, &ctx->inflight_list, inflight_entry)
-		cnt += io_match_task(req, task, files);
+		cnt += __io_match_task(req, task, files, false);
 	spin_unlock_irq(&ctx->inflight_lock);
 	return cnt;
 }
Signature · 1 line
2.24.0
maintainer
#3 lore raw
In reply to Pavel Begunkov
Quoted 9 lines · Pavel Begunkov

2/2 is for the recent syzbot report. Quick and dirty, but easy to backport. I plan to restructure (and clean) task vs files cancel later.

Pavel Begunkov (2):
io_uring: cancel files inflight counting
io_uring; fix files cancel hangs

 fs/io_uring.c | 21 ++++++++++++++++-----
 1 file changed, 16 insertions(+), 5 deletions(-)

Applied 2/2 for now, thanks.

Signature · 1 line
Jens Axboe