I got this traceback from lockdep while running last night's Fedora Core (rawhide) kernel, which is version 2.6.17-1.2356.fc6 . (I believe it's based on a fairly current git pull, davej will know for sure.) I'm not 100% sure if I'm doing something wrong in userland, or is this a kernel bug, but I'm pretty sure I shouldn't see lockdep complain either way. <4>============================================= <4>[ INFO: possible recursive locking detected ] <4>--------------------------------------------- <4>anaconda/451 is trying to acquire lock: <4> (&bdev->bd_mutex){--..}, at: [] __blkdev_put+0x22/0x123 <4> <4>but task is already holding lock: <4> (&bdev->bd_mutex){--..}, at: [] do_open+0x68/0x385 <4> <4>other info that might help us debug this: <4>1 lock held by anaconda/451: <4> #0: (&bdev->bd_mutex){--..}, at: [] do_open+0x68/0x385 <4> <4>stack backtrace: <4> [] show_trace_log_lvl+0x54/0xfd <4> [] show_trace+0xd/0x10 <4> [] dump_stack+0x19/0x1b <4> [] __lock_acquire+0x760/0x987 <4> [] lock_acquire+0x4b/0x6d <4> [] mutex_lock_nested+0xcc/0x243 <4> [] __blkdev_put+0x22/0x123 <4> [] blkdev_put+0xa/0xc <4> [] do_open+0x313/0x385 <4> [] blkdev_open+0x1f/0x48 <4> [] __dentry_open+0xb8/0x186 <4> [] nameidata_to_filp+0x1c/0x2e <4> [] do_filp_open+0x2f/0x36 <4> [] do_sys_open+0x40/0xbb <4> [] sys_open+0x16/0x18 <4> [] syscall_call+0x7/0xb The full syslog is attached. -- Peter