lkml.org 
[lkml]   [1999]   [Nov]   [29]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: AVL trees vs. Red-Black trees
Thank you for an informative summary.

Manfred Spraul wrote:
> 5) if the CPU sees a "lock;", "xchg", "in", "out" or any serializing
> instruction, then it reevaluates all speculative reads that the cpu
> has made.

I noticed Andrea Arcangeli repeatedly pointing out the difference
between atomicity and serialisation, and that "lock;" guarantees
atomicity not serialisation. Yet here you suggest that "any serializing
instruction" includes the atomic ones. What am I misunderstanding?

And the that matter, whch instructions are serialising in this way? Is
"sfence" serialising? What about a far jump? (Far jump serialises the
execution unit in as much as you need one to switch between real &
protected modes, for example, but I don't think it serialises memory
accesses).

And of course, are the rules for non-Intel "x86 compatible" processors
just as strong?

thanks,
-- Jamie

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:55    [from the cache]
©2003-2011 Jasper Spaans