On 2013-05-03 21:39, H Hartley Sweeten wrote:
Since the merge window is now open, this series is posted as an RFC.
Any comments would be appreciated. I'm hoping to get any issues
addressed so the series will be ready when the staging tree is
opened after the merge window closes.
This comedi driver is for a USB device and should be using the
comedi (*auto_attach) mechanism to attach to the device. Currently
the driver is a bit of a mess has a huge disconnect between the
loading of the usb_driver and the attaching of the comedi_device.
Right now this driver will even allow attaching to the comedi_device
even if the usb device is not attached to the system.
This series cleans up the driver and moves all the USB (*probe)
functionality into the comedi (*auto_attach) and the USB (*disconnect)
into the comedi (*detach). This ensures that a usb device is attached
before the comedi core does the (*auto_attach) to attach to the device.
This also allows consolidating all the private data into one struct
and makes the driver a lot easier to understand.
Some of these patches could probably be combined. Any comments are
welcome.
Jolly good. I started doing this myself on Friday but only got half way.
Some other things I noticed that haven't been fixed yet....
The timeout values such as '1 * HZ' in the calls to usb_bulk_msg()
should be in milliseconds, not jiffies. This was a change to the API
that never got applied to this driver for some reason. So instead of '1
* HZ', use '1000'.
dt9812_configure_gain() should probably set rmw->value to something in
the default case, although it shouldn't get there. I just made made it
the same as the DT9812_GAIN_1 case, setting rmw->value to 0x00.
rt9812_reset_device() (or dt9812_reset_device() if renamed) should bail
out with an error if the device ID is not supported, i.e. not one of
DT9812_DEVID_DT9812_10 or DT9812_DEVID_DT9812_2PT5.
--
-=( Ian Abbott @ MEV Ltd. E-mail: <abbotti@xxxxxxxxx> )=-
-=( Tel: +44 (0)161 477 1898 FAX: +44 (0)161 718 3587 )=-
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/devel
