On Sun, 15 Dec 2002 01:50:35 +0100
"Oliver Neukum" <[EMAIL PROTECTED]> wrote:

> Am Samstag, 14. Dezember 2002 22:23 schrieb Kari Hameenaho:
> > On Sat, 14 Dec 2002 16:44:30 +0100
> >
> > "Oliver Neukum" <[EMAIL PROTECTED]> wrote:
> > > Yes. Either you have a private array, which will lead people to do
> > > proper locking, or you have a function and reference counting.
> > > But a function which returns a pointer to a memory area that
> > > could go away is a recipe for desaster.
> >
> > I thought that the usb_interface never goes away ?
> 
> They perish with their device.
> If that weren't the case we'd have a memory leak.

Ok, then there is no reson to put minor there. I thought that
there were fixed number of interfaces and they just are 
assigned to devices.

> > > If you do that, the list accessing functions can have internal locking.
> > > And you've caused a temptation. You have now a way to get at a driver's
> > > interfaces without the driver's knowledge.
> >
> > You can do it now too, nothing is changed here ?
> 
> Not with knowledge of only the minor number.
> 

Oh, you mean access to usb_interface, that can easily be fixed and
find function can return the local data pointer. The pointer must be void
then, I dont know if it is good idea. But local data pointer can even now
be by searching a list, "minor table" is a list too.

---
Kari H�meenaho


-------------------------------------------------------
This sf.net email is sponsored by:
With Great Power, Comes Great Responsibility
Learn to use your power at OSDN's High Performance Computing Channel
http://hpc.devchannel.org/
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel

Reply via email to