Re: next-20160517 - lockdep splat in pcie code

[Valdis.Kletnieks@vt.edu]

On Wed, 18 May 2016 11:54:59 +0300, Mika Westerberg said:

> Can you check if the patch fixes the issue?

Tested, no complaints at boot, and everything seems to be working.

Feel free to add this to the patch and send it upstream:

Tested-By: Valdis Kletnieks <valdis.kletnieks@vt.edu>

> -----8<-----8<-----8<-----8<-----8<-----8<-----8<------8<-----8<------
>
> From: Mika Westerberg <mika.westerberg@linux.intel.com>
> Date: Wed, 18 May 2016 11:35:09 +0300
> Subject: [PATCH] PCI: pcie: Call pm_runtime_no_callbacks() after device is registered
>
> Commit 0195d2813547 ("PCI: Add runtime PM support for PCIe ports") added
> call to pm_runtime_no_callbacks() for each port service device to prevent
> them exposing unnecessary runtime PM sysfs files. However, that function
> tries to acquire dev->power.lock which is not yet initialized.
>
> This triggers following splat:
>
>  BUG: spinlock bad magic on CPU#0, swapper/0/1
>   lock: 0xffff8801be2aa8e8, .magic: 00000000, .owner: <none>/-1, .owner_cpu: 0
>  CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.6.0+ #820
>   0000000000000000 ffff8801beb97be0 ffffffff812cf42d 0000000000000000
>   ffff8801be2aa8e8 ffff8801beb97c00 ffffffff8109ee58 ffff8801be2aa8e8
>   ffff8801be2aa8e8 ffff8801beb97c30 ffffffff8109efd9 ffff8801be2aa8e8
>  Call Trace:
>   [<ffffffff812cf42d>] dump_stack+0x4f/0x72
>   [<ffffffff8109ee58>] spin_dump+0x78/0xc0
>   [<ffffffff8109efd9>] do_raw_spin_lock+0xf9/0x150
>   [<ffffffff81674e10>] _raw_spin_lock_irq+0x20/0x30
>   [<ffffffff813d738e>] pm_runtime_no_callbacks+0x1e/0x40
>   [<ffffffff81322cad>] pcie_port_device_register+0x1fd/0x4e0
>   [<ffffffff813232f8>] pcie_portdrv_probe+0x38/0xa0
>   [<ffffffff81314fb5>] local_pci_probe+0x45/0xa0
>   [<ffffffff81315fe0>] ? pci_match_device+0xe0/0x110
>   [<ffffffff813163ab>] pci_device_probe+0xdb/0x130
>   [<ffffffff813ccefc>] driver_probe_device+0x22c/0x440
>   [<ffffffff813cd1e1>] __driver_attach+0xd1/0xf0
>   [<ffffffff813cd110>] ? driver_probe_device+0x440/0x440
>   [<ffffffff813caab4>] bus_for_each_dev+0x64/0xa0
>   [<ffffffff813cc60e>] driver_attach+0x1e/0x20
>   [<ffffffff813cc0cb>] bus_add_driver+0x1eb/0x280
>   [<ffffffff81d36cdc>] ? pcie_port_setup+0x7c/0x7c
>   [<ffffffff813cdb90>] driver_register+0x60/0xe0
>   [<ffffffff813148e0>] __pci_register_driver+0x60/0x70
>   [<ffffffff81d36d3f>] pcie_portdrv_init+0x63/0x75
>   [<ffffffff810003db>] do_one_initcall+0xab/0x1c0
>   [<ffffffff81cff083>] kernel_init_freeable+0x153/0x1d9
>   [<ffffffff8166dfce>] kernel_init+0xe/0x100
>   [<ffffffff81675612>] ret_from_fork+0x22/0x40
>   [<ffffffff8166dfc0>] ? rest_init+0x90/0x90
>
> Fix this by calling pm_runtime_no_callbacks() after device_register() just
> like other buses, like I2C is doing already.
>
> Reported-by: Valdis Kletnieks <Valdis.Kletnieks@vt.edu>
> Suggested-by: Lukas Wunner <lukas@wunner.de>
> Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
> ---
>  drivers/pci/pcie/portdrv_core.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/pci/pcie/portdrv_core.c b/drivers/pci/pcie/portdrv_core.c
> index 65b1a624826b..97927dfbbf5f 100644
> --- a/drivers/pci/pcie/portdrv_core.c
> +++ b/drivers/pci/pcie/portdrv_core.c
> @@ -344,7 +344,6 @@ static int pcie_device_init(struct pci_dev *pdev, int service, int irq)
>  		     get_descriptor_id(pci_pcie_type(pdev), service));
>  	device->parent = &pdev->dev;
>  	device_enable_async_suspend(device);
> -	pm_runtime_no_callbacks(device);
>  
>  	retval = device_register(device);
>  	if (retval) {
> @@ -352,6 +351,8 @@ static int pcie_device_init(struct pci_dev *pdev, int service, int irq)
>  		return retval;
>  	}
>  
> +	pm_runtime_no_callbacks(device);
> +
>  	return 0;
>  }
>  
> -- 
> 2.8.1
>
>