From d7e10d47691d1702db1cd1edcc689d3031eefc67 Mon Sep 17 00:00:00 2001 From: Xiaoguang Wang Date: Thu, 4 Feb 2021 17:20:56 +0800 Subject: [PATCH 1/2] io_uring: don't modify identity's files uncess identity is cowed Abaci Robot reported following panic: BUG: kernel NULL pointer dereference, address: 0000000000000000 PGD 800000010ef3f067 P4D 800000010ef3f067 PUD 10d9df067 PMD 0 Oops: 0002 [#1] SMP PTI CPU: 0 PID: 1869 Comm: io_wqe_worker-0 Not tainted 5.11.0-rc3+ #1 Hardware name: Red Hat KVM, BIOS 0.5.1 01/01/2011 RIP: 0010:put_files_struct+0x1b/0x120 Code: 24 18 c7 00 f4 ff ff ff e9 4d fd ff ff 66 90 0f 1f 44 00 00 41 57 41 56 49 89 fe 41 55 41 54 55 53 48 83 ec 08 e8 b5 6b db ff 41 ff 0e 74 13 48 83 c4 08 5b 5d 41 5c 41 5d 41 5e 41 5f e9 9c RSP: 0000:ffffc90002147d48 EFLAGS: 00010293 RAX: 0000000000000000 RBX: ffff88810d9a5300 RCX: 0000000000000000 RDX: ffff88810d87c280 RSI: ffffffff8144ba6b RDI: 0000000000000000 RBP: 0000000000000080 R08: 0000000000000001 R09: ffffffff81431500 R10: ffff8881001be000 R11: 0000000000000000 R12: ffff88810ac2f800 R13: ffff88810af38a00 R14: 0000000000000000 R15: ffff8881057130c0 FS: 0000000000000000(0000) GS:ffff88813bc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000000 CR3: 000000010dbaa002 CR4: 00000000003706f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: __io_clean_op+0x10c/0x2a0 io_dismantle_req+0x3c7/0x600 __io_free_req+0x34/0x280 io_put_req+0x63/0xb0 io_worker_handle_work+0x60e/0x830 ? io_wqe_worker+0x135/0x520 io_wqe_worker+0x158/0x520 ? __kthread_parkme+0x96/0xc0 ? io_worker_handle_work+0x830/0x830 kthread+0x134/0x180 ? kthread_create_worker_on_cpu+0x90/0x90 ret_from_fork+0x1f/0x30 Modules linked in: CR2: 0000000000000000 ---[ end trace c358ca86af95b1e7 ]--- I guess case below can trigger above panic: there're two threads which operates different io_uring ctxs and share same sqthread identity, and later one thread exits, io_uring_cancel_task_requests() will clear task->io_uring->identity->files to be NULL in sqpoll mode, then another ctx that uses same identity will panic. Indeed we don't need to clear task->io_uring->identity->files here, io_grab_identity() should handle identity->files changes well, if task->io_uring->identity->files is not equal to current->files, io_cow_identity() should handle this changes well. Cc: stable@vger.kernel.org # 5.5+ Reported-by: Abaci Robot Signed-off-by: Xiaoguang Wang Reviewed-by: Pavel Begunkov Signed-off-by: Jens Axboe --- fs/io_uring.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/fs/io_uring.c b/fs/io_uring.c index 38c6cbe1ab38..5d3348d66f06 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -8982,12 +8982,6 @@ static void io_uring_cancel_task_requests(struct io_ring_ctx *ctx, if ((ctx->flags & IORING_SETUP_SQPOLL) && ctx->sq_data) { atomic_dec(&task->io_uring->in_idle); - /* - * If the files that are going away are the ones in the thread - * identity, clear them out. - */ - if (task->io_uring->identity->files == files) - task->io_uring->identity->files = NULL; io_sq_thread_unpark(ctx->sq_data); } } From aec18a57edad562d620f7d19016de1fc0cc2208c Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Thu, 4 Feb 2021 19:22:46 +0000 Subject: [PATCH 2/2] io_uring: drop mm/files between task_work_submit Since SQPOLL task can be shared and so task_work entries can be a mix of them, we need to drop mm and files before trying to issue next request. Cc: stable@vger.kernel.org # 5.10+ Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe --- fs/io_uring.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/io_uring.c b/fs/io_uring.c index 5d3348d66f06..1f68105a41ed 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -2205,6 +2205,9 @@ static void __io_req_task_submit(struct io_kiocb *req) else __io_req_task_cancel(req, -EFAULT); mutex_unlock(&ctx->uring_lock); + + if (ctx->flags & IORING_SETUP_SQPOLL) + io_sq_thread_drop_mm_files(); } static void io_req_task_submit(struct callback_head *cb)