[PATCH 1/1] io_uring: don't leak creds on SQO attach error

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Attaching to already dead/dying SQPOLL task is disallowed in
io_sq_offload_create(), but cleanup is hand coded by calling
io_put_sq_data()/etc., that miss to put ctx->sq_creds.

Defer everything to error-path io_sq_thread_finish(), adding
ctx->sqd_list in the error case as well as finish will handle it.

Signed-off-by: Pavel Begunkov <asml.silence@xxxxxxxxx>
---
 fs/io_uring.c | 17 ++++++-----------
 1 file changed, 6 insertions(+), 11 deletions(-)

diff --git a/fs/io_uring.c b/fs/io_uring.c
index dcdb6e83f1a2..f258264a2e89 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -7904,22 +7904,17 @@ static int io_sq_offload_create(struct io_ring_ctx *ctx,
 
 		ret = 0;
 		io_sq_thread_park(sqd);
+		list_add(&ctx->sqd_list, &sqd->ctx_list);
+		io_sqd_update_thread_idle(sqd);
 		/* don't attach to a dying SQPOLL thread, would be racy */
-		if (attached && !sqd->thread) {
+		if (attached && !sqd->thread)
 			ret = -ENXIO;
-		} else {
-			list_add(&ctx->sqd_list, &sqd->ctx_list);
-			io_sqd_update_thread_idle(sqd);
-		}
 		io_sq_thread_unpark(sqd);
 
-		if (ret < 0) {
-			io_put_sq_data(sqd);
-			ctx->sq_data = NULL;
-			return ret;
-		} else if (attached) {
+		if (ret < 0)
+			goto err;
+		if (attached)
 			return 0;
-		}
 
 		if (p->flags & IORING_SETUP_SQ_AFF) {
 			int cpu = p->sq_thread_cpu;
-- 
2.24.0




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux