Re: USB_DEBUG mess

[Iain Hibbert <plunky%ogmig.net@localhost>]

On Sat, 5 Jan 2013, Christos Zoulas wrote:

> USB_DEBUG is defopt'ed and it is used as a global knob in usb.h:
[...]
> This is done for drivers which don't include "opt_usb.h" so they will not
> recognize when USB_DEBUG changes.

Since when has this defopt been the case?

Ok, answering my own question, it was 18 months ago by Matt.. but why was
this added?

  
http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/dev/usb/files.usb?rev=1.106&content-type=text/x-cvsweb-markup

Normally, the XXX_DEBUG options are not specified in any files.* files,
meaning that as they are unknown options, they will translate into a
CPPFLAG of -DXXX_DEBUG in the kernel Makefile

> in the long term.
>
> - Do we want to move the .c defines in usb.h, and have usb.h include 
> "opt_usb.h"
> - Do we want to not enable all the driver debugging code when USB_DEBUG is 
> defined
>   and leave it to the individual drivers what to do (and leave USB_DEBUG in 
> the
>   generic driver code where it is used directly)?

historically, if you define USB_DEBUG then you get all known debug options
relating to USB enabled, otherwise you can turn them on individually if
you choose.

why would we want to change that?

iain