From 4fca4916e323224267fe1d6c8ca332824ebf7808 Mon Sep 17 00:00:00 2001 From: Zhaoyang Huang Date: Mon, 17 Feb 2025 10:40:59 +0800 Subject: [RFC PATCH] loop: fix __list_add_valid in loop driver Oops reported by syzbot[1] which introduced by commit 3bee991f2b68 of linux-next(loop: release the lo_work_lock before queue_work). This should be caused by bellow race scenario and could be fixed by subsituting the judge if working_pending to list_empty. mainline before 3bee991f2b68: T0 kworker/u0:1 loop_queue_work() { process_one_work() { if(!work_pending) set_work_pool_and_clear_pending() queue_work() { set_work_data(WORK_PENDING) spin_unlock(worker->lock) } loop_process_work { spin_lock(worker->lock) if(!work_pending) //FAILED HERE list_add(&work->idle_list) linux-next after 3bee991f2b68: T0 & kworker/u0:0 kworker/u0:1 loop_queue_work() { process_one_work() { set_work_pool_and_clear_pending() spin_unlock(worker->lock) loop_process_work() { spin_lock(worker->lock) if(!work_pending) //FIRST ADD list_add(&work->idle_list) if(!work_pending) queue_work() set_work_data(WORK_PENDING) spin_unlock(worker->lock) } process_one_work() { set_work_pool_and_clear_pending() loop_process_work() { spin_lock(worker->lock) if(!work_pending) //SECOND ADD list_add(&work->idle_list) BUG(...) [1] list_add double add: new=ffff88807fe21c70, prev=ffff88807fe21c70, next=ffff888024c29160. ------------[ cut here ]------------ kernel BUG at lib/list_debug.c:37! Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI CPU: 1 UID: 0 PID: 12 Comm: kworker/u8:1 Not tainted 6.14.0-rc2-next-20250212-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 12/27/2024 Workqueue: loop0 loop_workfn RIP: 0010:__list_add_valid_or_report+0xa4/0x130 lib/list_debug.c:35 Code: f7 74 11 b0 01 5b 41 5c 41 5d 41 5e 41 5f 5d c3 cc cc cc cc 48 c7 c7 40 e5 60 8c 4c 89 fe 4c 89 e2 4c 89 f1 e8 fd 88 35 fc 90 <0f> 0b 48 c7 c7 40 e3 60 8c e8 ee 88 35 fc 90 0f 0b 48 c7 c7 e0 e3 RSP: 0018:ffffc90000117628 EFLAGS: 00010046 RAX: 0000000000000058 RBX: 1ffff1100ffc438e RCX: 89e05f8d6ffcb000 RDX: 0000000000000000 RSI: 0000000080000001 RDI: 0000000000000000 RBP: 1ffff1100498522d R08: ffffffff819f562c R09: 1ffff92000022e60 R10: dffffc0000000000 R11: fffff52000022e61 R12: ffff88807fe21c70 R13: dffffc0000000000 R14: ffff888024c29160 R15: ffff88807fe21c70 FS: 0000000000000000(0000) GS:ffff8880b8700000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007ff4b64ffe00 CR3: 000000007cfa4000 CR4: 00000000003526f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: __list_add_valid include/linux/list.h:88 [inline] __list_add include/linux/list.h:150 [inline] list_add_tail include/linux/list.h:183 [inline] loop_process_work+0x1f96/0x21c0 drivers/block/loop.c:1977 process_one_work kernel/workqueue.c:3236 [inline] process_scheduled_works+0xa66/0x1840 kernel/workqueue.c:3317 worker_thread+0x870/0xd30 kernel/workqueue.c:3398 kthread+0x7a9/0x920 kernel/kthread.c:464 ret_from_fork+0x4b/0x80 arch/x86/kernel/process.c:148 ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:244 Modules linked in: ---[ end trace 0000000000000000 ]--- RIP: 0010:__list_add_valid_or_report+0xa4/0x130 lib/list_debug.c:35 Code: f7 74 11 b0 01 5b 41 5c 41 5d 41 5e 41 5f 5d c3 cc cc cc cc 48 c7 c7 40 e5 60 8c 4c 89 fe 4c 89 e2 4c 89 f1 e8 fd 88 35 fc 90 <0f> 0b 48 c7 c7 40 e3 60 8c e8 ee 88 35 fc 90 0f 0b 48 c7 c7 e0 e3 RSP: 0018:ffffc90000117628 EFLAGS: 00010046 RAX: 0000000000000058 RBX: 1ffff1100ffc438e RCX: 89e05f8d6ffcb000 RDX: 0000000000000000 RSI: 0000000080000001 RDI: 0000000000000000 RBP: 1ffff1100498522d R08: ffffffff819f562c R09: 1ffff92000022e60 R10: dffffc0000000000 R11: fffff52000022e61 R12: ffff88807fe21c70 R13: dffffc0000000000 R14: ffff888024c29160 R15: ffff88807fe21c70 FS: 0000000000000000(0000) GS:ffff8880b8700000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007ff4b64ffe00 CR3: 000000007cfa4000 CR4: 00000000003526f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Reported-by: syzbot+c104904eeb2c0edbdb06@syzkaller.appspotmail.com Fixes: 3bee991f2b68 ("loop: release the lo_work_lock before queue_work") Signed-off-by: Zhaoyang Huang --- drivers/block/loop.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/loop.c b/drivers/block/loop.c index 68c943a77e41..2bcdd9646ac3 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -1972,7 +1972,7 @@ static void loop_process_work(struct loop_worker *worker, * *and* the worker will not run again which ensures that it * is safe to free any worker on the idle list */ - if (worker && !work_pending(&worker->work)) { + if (worker && list_empty(&worker->idle_list)) { worker->last_ran_at = jiffies; list_add_tail(&worker->idle_list, &lo->idle_worker_list); loop_set_timer(lo); -- 2.25.1