tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: The missing membar_X() directive

On 14 Jul 2010, at 12:25 , Jean-Yves Migeon wrote:

> On 14.07.2010 18:26, Dennis Ferguson wrote:
>> I didn't know there was a name for that but, yes, that's how it
>> works.  When something is removed from the structure the memory
>> allocator needs to hang on to the now-detached memory and keep
>> it unchanged until it can prove that all lookups which might have
>> started before the change have finished.
> IIRC, you cannot implement RCU in non GPL software (unless IBM gives
> approval for it).

While, to be clear, I think I can implement anything I want since I'm
neither a US nor an EU resident, I think the issue the patents address
(now that I've become aware of them) has nothing to do with the data structure
itself (the data structure does derive from one which is independently
patented, but with a patent owner who I know won't be suing).  They
instead deal with what I called the memory allocator's problem above,
that is what method do you use to prove that concurrent accesses
that started before the change have finished.  There are a lot of ways
to skin that cat besides the method Linux uses, and the patents
related to that: if something else can't be independently invented the
method described by US patent 4,809,168 looks workable to me, and that one
is expired.

In any case, let's not call it RCU so no one can confuse it with anything
that is patented.  If and when the time comes to think about using this
for something, an appropriate method for doing this can be chosen to ensure
that no one gets sued.

Dennis Ferguson

Home | Main Index | Thread Index | Old Index