| Message ID | 0f2f1aa5729332612bd01fe0f2f385fd1f06ce7c.1735231717.git.asml.silence@gmail.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [1/1] io_uring/sqpoll: fix sqpoll error handling races | expand |
On 12/26/24 9:49 AM, Pavel Begunkov wrote: > BUG: KASAN: slab-use-after-free in __lock_acquire+0x370b/0x4a10 kernel/locking/lockdep.c:5089 > Call Trace: > <TASK> > ... > _raw_spin_lock_irqsave+0x3d/0x60 kernel/locking/spinlock.c:162 > class_raw_spinlock_irqsave_constructor include/linux/spinlock.h:551 [inline] > try_to_wake_up+0xb5/0x23c0 kernel/sched/core.c:4205 > io_sq_thread_park+0xac/0xe0 io_uring/sqpoll.c:55 > io_sq_thread_finish+0x6b/0x310 io_uring/sqpoll.c:96 > io_sq_offload_create+0x162/0x11d0 io_uring/sqpoll.c:497 > io_uring_create io_uring/io_uring.c:3724 [inline] > io_uring_setup+0x1728/0x3230 io_uring/io_uring.c:3806 > ... > > Kun Hu reports that the SQPOLL creating error path has UAF, which > happens if io_uring_alloc_task_context() fails and then io_sq_thread() > manages to run and complete before the rest of error handling code, > which means io_sq_thread_finish() is looking at already killed task. Might be worth mentioning that this is only really a fault injection thing. But ouside of that, looks fine to me - thanks!
On Thu, 26 Dec 2024 16:49:23 +0000, Pavel Begunkov wrote: > BUG: KASAN: slab-use-after-free in __lock_acquire+0x370b/0x4a10 kernel/locking/lockdep.c:5089 > Call Trace: > <TASK> > ... > _raw_spin_lock_irqsave+0x3d/0x60 kernel/locking/spinlock.c:162 > class_raw_spinlock_irqsave_constructor include/linux/spinlock.h:551 [inline] > try_to_wake_up+0xb5/0x23c0 kernel/sched/core.c:4205 > io_sq_thread_park+0xac/0xe0 io_uring/sqpoll.c:55 > io_sq_thread_finish+0x6b/0x310 io_uring/sqpoll.c:96 > io_sq_offload_create+0x162/0x11d0 io_uring/sqpoll.c:497 > io_uring_create io_uring/io_uring.c:3724 [inline] > io_uring_setup+0x1728/0x3230 io_uring/io_uring.c:3806 > ... > > [...] Applied, thanks! [1/1] io_uring/sqpoll: fix sqpoll error handling races commit: e33ac68e5e21ec1292490dfe061e75c0dbdd3bd4 Best regards,
diff --git a/io_uring/sqpoll.c b/io_uring/sqpoll.c index 6df5e649c413..9e5bd79fd2b5 100644 --- a/io_uring/sqpoll.c +++ b/io_uring/sqpoll.c @@ -405,6 +405,7 @@ void io_sqpoll_wait_sq(struct io_ring_ctx *ctx) __cold int io_sq_offload_create(struct io_ring_ctx *ctx, struct io_uring_params *p) { + struct task_struct *task_to_put = NULL; int ret; /* Retain compatibility with failing for an invalid attach attempt */ @@ -480,6 +481,7 @@ __cold int io_sq_offload_create(struct io_ring_ctx *ctx, } sqd->thread = tsk; + task_to_put = get_task_struct(tsk); ret = io_uring_alloc_task_context(tsk, ctx); wake_up_new_task(tsk); if (ret) @@ -490,11 +492,15 @@ __cold int io_sq_offload_create(struct io_ring_ctx *ctx, goto err; } + if (task_to_put) + put_task_struct(task_to_put); return 0; err_sqpoll: complete(&ctx->sq_data->exited); err: io_sq_thread_finish(ctx); + if (task_to_put) + put_task_struct(task_to_put); return ret; }
BUG: KASAN: slab-use-after-free in __lock_acquire+0x370b/0x4a10 kernel/locking/lockdep.c:5089 Call Trace: <TASK> ... _raw_spin_lock_irqsave+0x3d/0x60 kernel/locking/spinlock.c:162 class_raw_spinlock_irqsave_constructor include/linux/spinlock.h:551 [inline] try_to_wake_up+0xb5/0x23c0 kernel/sched/core.c:4205 io_sq_thread_park+0xac/0xe0 io_uring/sqpoll.c:55 io_sq_thread_finish+0x6b/0x310 io_uring/sqpoll.c:96 io_sq_offload_create+0x162/0x11d0 io_uring/sqpoll.c:497 io_uring_create io_uring/io_uring.c:3724 [inline] io_uring_setup+0x1728/0x3230 io_uring/io_uring.c:3806 ... Kun Hu reports that the SQPOLL creating error path has UAF, which happens if io_uring_alloc_task_context() fails and then io_sq_thread() manages to run and complete before the rest of error handling code, which means io_sq_thread_finish() is looking at already killed task. Cc: stable@vger.kernel.org Reported-by: Kun Hu <huk23@m.fudan.edu.cn> Signed-off-by: Pavel Begunkov <asml.silence@gmail.com> --- io_uring/sqpoll.c | 6 ++++++ 1 file changed, 6 insertions(+)