Re: Re: RFC: patch to allow lock-free traversal of lists

[Paul E. McKenney]

>On Sat, 13 Oct 2001, Linus Torvalds wrote:
>>
>> In short, RCU seems to be a case of "hey, that's cool", but it's a
>> solution in search of a problem so severe that it is worth it.
>
>Oh, and before people start telling me that RCU was successfully used in
>AIX/projectX/xxxx/etc, you have to realize that I don't give a rats *ss
>about the fact that there are OS's out there that are "more scalable".

I understand and agree: RCU must prove itself -in- -Linux-.

>The last time I looked, Solaris and AIX and all the rest of the "scalable"
>systems were absolute pigs on smaller hardware, and the "scalability" in
>them often translates into "we scale linearly to many CPU's by being
>really bad even on one".

Again understood and agreed.  Simplicity and single-CPU performance are
at least as important as scalability.  Although I believe that RCU can
be an important part of all three in some situations, this belief is
clearly in need of more proof.

						Thanx, Paul