Re: Future features (was Re: Discussing issues)

Havoc Pennington <[email protected]>
Newsgroups gmane.comp.xfree86.forum
Message-ID <[email protected]>
On Wed, Apr 16, 2003 at 07:27:35PM -0400, Alan Cox wrote:
> > >    I would say that loading the driver is the domain of the operating
> > >    system, and apps/desktop only care about devices the OS has already
> > >    "found."
> 
> Not all drivers are kernel side. Its like X input devices. My pen driver
> is in XFree86 not in drivers/input, and I'd like to keep it that way 8)
> 

Right, that's an issue I have in my rough notes on this. The case I
was thinking of was printers.

When a driver is in the kernel though, from talking to you the other
day I gathered the device "exists" to most userspace apps when the
driver has been loaded. Until then we don't know whether there will be
0, 1, or N virtual devices (major/minor pairs) for a piece of
hardware. Hopefully I understood that right. ;-)

For drivers that are in userspace, we could use a similar model (the
device exists when CUPS or X has loaded it), or do something
different.
 
> USB core driver - new device 0x1111 , 0x4115, class USB storage
> USB to user space "Hey do something with this info"

So user space here is probably just a single special-purpose app that
loads usb-storage.

> User loads usb-storage
> USB storage to user space "Hey this disk has appeared"

User space here is probably all kinds of random apps that care.

Havoc
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.