Thanks for the RTFP pointer, the Triplett ref, hash tab via rt is
really interesting!
I think I'll have to read it again and again to get all out of it...
This brings me to a new question: how much is pypy python memory model
allowed to diverge from that of cpython?
For example, dict lookups, et
On Mon, Dec 27, 2010 at 19:54, Dima Tisnek wrote:
> what do you think of liburcu? lttng.org/urcu
I've studied RCU in the Linux kernel (from which URCU derives) and
thought for a long time of using it for this problem. In short, once
you have GC, RCU (as in the Linux kernel) becomes (almost) trivi
what do you think of liburcu? lttng.org/urcu
p.s. if/when I get around to profiling pypy, which is something I want
to do, I'll be sure to share the results here.
d.
On 27 December 2010 10:57, Paolo Giarrusso wrote:
> On Mon, Dec 27, 2010 at 07:00, Dima Tisnek wrote:
>> (dan) random binary tre
On Mon, Dec 27, 2010 at 07:00, Dima Tisnek wrote:
> (dan) random binary trees: O(log2(n)) is 7.2 for builtins, though the
> constant factor for locks, etc, might make it worthwhile
"7.2"? What is that? Do you mean 7.2=log2(no. of builtins)?
> (paolo) non blocking hash maps: memory barriers can b
On Mon, Dec 27, 2010 at 09:31, Carl Friedrich Bolz wrote:
> On 12/27/2010 07:00 AM, Dima Tisnek wrote:
>> (dan) random binary trees: O(log2(n)) is 7.2 for builtins, though the
>> constant factor for locks, etc, might make it worthwhile
>>
>> (paolo) non blocking hash maps: memory barriers can be q
Oh, ok. Fixed now by adding "-b default " to the hg log command.
(still learning mercurial :)
2010/12/26 Amaury Forgeot d'Arc :
> Hi,
>
> 2010/12/25 Miquel Torres
>>
>> I finally added Mercurial support to Codespeed, following Antonio's
>> suggestion of parsing the (templated) command line outp
On 27/12/10 09:31, Carl Friedrich Bolz wrote:
> The only case where improving dicts would help is for user-defined
> dictionaries, i.e. when you write dicts with curly braces.
then it's easy to optimize, you just write "dict()" instead of "{}".
(sorry, I could not resist :-))
__
On 12/27/2010 07:00 AM, Dima Tisnek wrote:
> (dan) random binary trees: O(log2(n)) is 7.2 for builtins, though the
> constant factor for locks, etc, might make it worthwhile
>
> (paolo) non blocking hash maps: memory barriers can be quite costly too
>
> different options will have to be implemented
