[boost] Re: [shared_ptr] possible bug in sp_counted_base.hpp

From: Rene Rivera (grafik.list_at_[hidden])
Date: 2005-04-14 16:09:59

Peter Dimov wrote:
> Rene Rivera wrote:
> 
>>> No.. just one that tests all three of the functions. The timing test
>>> you have only tests one of them.
> 
> No, it tests two of them. 

Really? I'm not so sure because when I look at the dissaembly of the 
test and I only saw "xadd" from one of the functions. Didn't see an 
"inc" from the other one.

>>> Well certainly an _mt timing test would be best. As it's not that
>>> useful to only time the non-contention cases.
> 
> There are two _mt tests that are also timing tests, you even reported 
> crashes with one of them. ;-)

Didn't realize they spit out timing information :-) Here's the timing 
info the weak_ptr test..

CodeWarrior 8.3...

sp_counted_base_w32.hpp, 17.555
sp_counted_base_nt.hpp, 17.355
sp_counted_base_cw_x86.hpp, 17.215

VC 7.1...

sp_counted_base_w32.hpp, 13.259
sp_counted_base_nt.hpp, 13.319

> The intent of shared_ptr_timing_test is precisely to time the 
> uncontended case, because that's what I want to measure, the cost of MT 
> safety when none is needed.

Oh, I see. So you really need that weak_ptr_timing_test ;-)

-- 
-- Grafik - Don't Assume Anything
-- Redshift Software, Inc. - http://redshift-software.com
-- rrivera/acm.org - grafik/redshift-software.com
-- 102708583/icq - grafikrobot/aim - Grafik/jabber.org