Hi Maarten
On Mon, 2013-02-04 at 22:59 +0100, Maarten Lankhorst wrote:
Op 04-02-13 22:30, Marcin Slusarz schreef:
- Lockdep thinks all nouveau subdevs belong to the same class and can be
locked in arbitrary order, which is not true (at least in general case). Tell it to distinguish subdevs by (o)class type.
Apart from this specific case, is there any other reason why we require being able to nest 2 subdev locks?
Add a NVOBJ_FLAG_CREATE_EXCLUSIVE flag to nouveau_engctx_create and return -EBUSY if there is any existing object. Problem solved, and lockdep is still generic.
- DRM client can be locked under user client lock - tell lockdep to put DRM
client lock in a separate class.
Can you copy paste a typical lockdep splat for this? Also this should be a separate patch. :-)
PS - Deep call chain. Has anyone looked into max stack depth yet?
[ 5.995881] ============================================= [ 5.995886] [ INFO: possible recursive locking detected ] [ 5.995892] 3.8.0-next-20130125+ttypatch-xeon+lockdep #20130125+ttypatch Not tainted [ 5.995898] --------------------------------------------- [ 5.995904] modprobe/275 is trying to acquire lock: [ 5.995909] (&subdev->mutex){+.+.+.}, at: [<ffffffffa00d10b8>] nouveau_instobj_create_+0x48/0x90 [nouveau] [ 5.995955] [ 5.995955] but task is already holding lock: [ 5.995961] (&subdev->mutex){+.+.+.}, at: [<ffffffffa00da3b5>] nv50_disp_data_ctor+0x65/0xd0 [nouveau] [ 5.995989] [ 5.995989] other info that might help us debug this: [ 5.995995] Possible unsafe locking scenario: [ 5.995995] [ 5.996001] CPU0 [ 5.996004] ---- [ 5.996005] lock(&subdev->mutex); [ 5.996005] lock(&subdev->mutex); [ 5.996005] [ 5.996005] *** DEADLOCK *** [ 5.996005] [ 5.996005] May be due to missing lock nesting notation [ 5.996005] [ 5.996005] 4 locks held by modprobe/275: [ 5.996005] #0: (&__lockdep_no_validate__){......}, at: [<ffffffff8146fa5b>] __driver_attach+0x5b/0xb0 [ 5.996005] #1: (&__lockdep_no_validate__){......}, at: [<ffffffff8146fa69>] __driver_attach+0x69/0xb0 [ 5.996005] #2: (drm_global_mutex){+.+.+.}, at: [<ffffffffa0028d86>] drm_get_pci_dev+0xc6/0x2d0 [drm] [ 5.996005] #3: (&subdev->mutex){+.+.+.}, at: [<ffffffffa00da3b5>] nv50_disp_data_ctor+0x65/0xd0 [nouveau] [ 5.996005] [ 5.996005] stack backtrace: [ 5.996005] Pid: 275, comm: modprobe Not tainted 3.8.0-next-20130125+ttypatch-xeon+lockdep #20130125+ttypatch [ 5.996005] Call Trace: [ 5.996005] [<ffffffff810bd437>] __lock_acquire+0x687/0x1a70 [ 5.996005] [<ffffffff810bf70b>] ? mark_held_locks+0x9b/0x100 [ 5.996005] [<ffffffff810bf87d>] ? trace_hardirqs_on_caller+0x10d/0x1a0 [ 5.996005] [<ffffffff810bf91d>] ? trace_hardirqs_on+0xd/0x10 [ 5.996005] [<ffffffff8170a87a>] ? _raw_write_unlock_irqrestore+0x4a/0x90 [ 5.996005] [<ffffffff810bedc8>] lock_acquire+0x98/0x150 [ 5.996005] [<ffffffffa00d10b8>] ? nouveau_instobj_create_+0x48/0x90 [nouveau] [ 5.996005] [<ffffffffa00d10b8>] ? nouveau_instobj_create_+0x48/0x90 [nouveau] [ 5.996005] [<ffffffff81707f60>] mutex_lock_nested+0x50/0x390 [ 5.996005] [<ffffffffa00d10b8>] ? nouveau_instobj_create_+0x48/0x90 [nouveau] [ 5.996005] [<ffffffffa00b6376>] ? nouveau_object_ref+0x46/0xd0 [nouveau] [ 5.996005] [<ffffffffa00b64b5>] ? nouveau_object_create_+0x65/0xa0 [nouveau] [ 5.996005] [<ffffffffa00d10b8>] nouveau_instobj_create_+0x48/0x90 [nouveau] [ 5.996005] [<ffffffffa00d1be1>] nv50_instobj_ctor+0x51/0xf0 [nouveau] [ 5.996005] [<ffffffffa00b62a8>] nouveau_object_ctor+0x38/0xc0 [nouveau] [ 5.996005] [<ffffffffa00d1b36>] nv50_instmem_alloc+0x26/0x30 [nouveau] [ 5.996005] [<ffffffffa00b49a7>] nouveau_gpuobj_create_+0x247/0x2f0 [nouveau] [ 5.996005] [<ffffffff8170afed>] ? _raw_spin_unlock_irqrestore+0x6d/0x90 [ 5.996005] [<ffffffff810bf87d>] ? trace_hardirqs_on_caller+0x10d/0x1a0 [ 5.996005] [<ffffffffa00b34dc>] nouveau_engctx_create_+0x26c/0x2b0 [nouveau] [ 5.996005] [<ffffffffa00da411>] nv50_disp_data_ctor+0xc1/0xd0 [nouveau] [ 5.996005] [<ffffffffa00b62a8>] nouveau_object_ctor+0x38/0xc0 [nouveau] [ 5.996005] [<ffffffffa00b6b82>] nouveau_object_new+0x112/0x240 [nouveau] [ 5.996005] [<ffffffffa0155fe5>] nv50_display_create+0x1a5/0x890 [nouveau] [ 5.996005] [<ffffffff8107837b>] ? __cancel_work_timer+0x8b/0xe0 [ 5.996005] [<ffffffffa013c80b>] nouveau_display_create+0x3cb/0x670 [nouveau] [ 5.996005] [<ffffffffa012ba0f>] nouveau_drm_load+0x26f/0x590 [nouveau] [ 5.996005] [<ffffffff8146caee>] ? device_register+0x1e/0x30 [ 5.996005] [<ffffffffa002a626>] ? drm_sysfs_device_add+0x86/0xb0 [drm] [ 5.996005] [<ffffffffa0028e46>] drm_get_pci_dev+0x186/0x2d0 [drm] [ 5.996005] [<ffffffffa012bf9a>] nouveau_drm_probe+0x26a/0x2c0 [nouveau] [ 5.996005] [<ffffffff8170afca>] ? _raw_spin_unlock_irqrestore+0x4a/0x90 [ 5.996005] [<ffffffff81393a8b>] local_pci_probe+0x4b/0x80 [ 5.996005] [<ffffffff81393da1>] pci_device_probe+0x111/0x120 [ 5.996005] [<ffffffff8146f6eb>] driver_probe_device+0x8b/0x3a0 [ 5.996005] [<ffffffff8146faab>] __driver_attach+0xab/0xb0 [ 5.996005] [<ffffffff8146fa00>] ? driver_probe_device+0x3a0/0x3a0 [ 5.996005] [<ffffffff8146d635>] bus_for_each_dev+0x55/0x90 [ 5.996005] [<ffffffff8146f12e>] driver_attach+0x1e/0x20 [ 5.996005] [<ffffffff8146ebe1>] bus_add_driver+0x121/0x2b0 [ 5.996005] [<ffffffffa01ad000>] ? 0xffffffffa01acfff [ 5.996005] [<ffffffff814701b7>] driver_register+0x77/0x170 [ 5.996005] [<ffffffffa01ad000>] ? 0xffffffffa01acfff [ 5.996005] [<ffffffff81392b14>] __pci_register_driver+0x64/0x70 [ 5.996005] [<ffffffffa00290aa>] drm_pci_init+0x11a/0x130 [drm] [ 5.996005] [<ffffffffa01ad000>] ? 0xffffffffa01acfff [ 5.996005] [<ffffffffa01ad000>] ? 0xffffffffa01acfff [ 5.996005] [<ffffffffa01ad04d>] nouveau_drm_init+0x4d/0x1000 [nouveau] [ 5.996005] [<ffffffff8100215a>] do_one_initcall+0x12a/0x180 [ 5.996005] [<ffffffff810cc6a7>] load_module+0x12f7/0x1be0 [ 5.996005] [<ffffffff8137ce20>] ? ddebug_proc_open+0xd0/0xd0 [ 5.996005] [<ffffffff8136993e>] ? trace_hardirqs_on_thunk+0x3a/0x3f [ 5.996005] [<ffffffff810cd067>] sys_init_module+0xd7/0x120 [ 5.996005] [<ffffffff81712f59>] system_call_fastpath+0x16/0x1b