Re: Patch: linux-2.5.58/drivers/base/bus.c ignored pre-existing devices

[Adam J. Richter]

On 2003-01-15, Patrick Mochel wrote:
>The extra code was an attempt at properly handling failure of ->probe(),
>and to make sure that an error from ->probe() (e.g. -ENOMEM) is propogated
>up.
>
>The cause of the problem you're seeing is that if a device wasn't bound to 
>a driver, -ENODEV was returned, causing it to be removed from the bus's 
>list of devices. 
>
>To remedy this, I've changed the semantics of bus_match() to return the 
>following:
>
>* 1     if device was bound to a driver.
>* 0     if it wasn't 
>* <0    if drv->probe() returned an error. 
>
>This allows the caller to know if binding happened, as well as bubble the 
>error up.

	I have't tried your patch, but, from reading it, I infer that
your patch would not work with drivers that do further matching
in their probe function and return -ENODEV, which should be handled
as if match() function failed.  I suggest you use my patch.

Adam J. Richter     __     ______________   575 Oroville Road
[e-mail]     \ /                  Milpitas, California 95035
[phone]         | g g d r a s i l   United States of America
                         "Free Software For The Rest Of Us."