[PATCH] USB: move usb debugfs directory creation to the usb common core

Felipe Balbi felipe.balbi at linux.intel.com
Wed Jun 5 00:28:12 PDT 2019


Hi,

Greg Kroah-Hartman <gregkh at linuxfoundation.org> writes:
>> Greg Kroah-Hartman <gregkh at linuxfoundation.org> writes:
>> > diff --git a/drivers/usb/core/usb.c b/drivers/usb/core/usb.c
>> > index 7fcb9f782931..f3d6b1ab80cb 100644
>> > --- a/drivers/usb/core/usb.c
>> > +++ b/drivers/usb/core/usb.c
>> > @@ -1185,19 +1185,17 @@ static struct notifier_block usb_bus_nb = {
>> >  	.notifier_call = usb_bus_notify,
>> >  };
>> >  
>> > -struct dentry *usb_debug_root;
>> > -EXPORT_SYMBOL_GPL(usb_debug_root);
>> > +static struct dentry *usb_devices_root;
>> >  
>> >  static void usb_debugfs_init(void)
>> >  {
>> > -	usb_debug_root = debugfs_create_dir("usb", NULL);
>> > -	debugfs_create_file("devices", 0444, usb_debug_root, NULL,
>> > -			    &usbfs_devices_fops);
>> > +	usb_devices_root = debugfs_create_file("devices", 0444, usb_debug_root,
>> 
>> don't we have a race now? Can usbcore ever probe before usb common?
>
> How can that happen if usb_debug_root is in usb common?  The module
> loader will not let that happen.  Or it shouldn't :)

argh, indeed. The very fact that usbcore tries to resolve usb_debug_root
already forces a dependency :-p

-- 
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 832 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-mediatek/attachments/20190605/cea41719/attachment.sig>


More information about the Linux-mediatek mailing list