Am Samstag, 10. Juli 2004 22:16 schrieb Alan Stern:
> > But I have no problem with that; it's what locktree does
> > already, after all. What it _needs_ to do in fact.
>
> My biggest stumbling block is what the API should be (and how to document
> it!). Should the rwsem be exposed? Should routines directly do
> down(&udev->serialize) for locking all devices after the first? Should
> there be a separate pair of routines to encapsulate the notion of nested
> locks? How should such a pair cope with the way locktree() doesn't nest
> its multiple locks? How can this be documented in a way that won't make
> people wonder why such a simple idea needs to be made so complex?
Is it really needed? How much contention do we have, so that we cannot
live with a single lock per bus? It seems to me that most operations
happen in khubd's context anyway.
Regards
Oliver
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
