393
Subj : Re: A question about atomic_ptr
To   : comp.programming.threads
From : Peter Dimov
Date : Mon Apr 18 2005 12:23 pm

Chris Thomasson wrote:
>> Hm. How about the following DWCASless algorithm:
>
> OK, this is similar to logic that I used a while back:
>
> http://groups.google.ca/groups?selm=R6fmb.18865%24Fm2.9642%40attbi_s04
>
> You could run into a similar race-condition described. Your algorithm
> could cause a memory leak:
>
> Think if the following scenario occurred:
>
>
>
>
> static obj_t *sptr = new objA;
>
>
> Threads A and B - ( both writers )

The algorithm was single writer, multiple readers, but it's still wrong. :-(

Reader r1 inc gate, read pointer
Writer w inc gate, replace pointer
Reader r2 inc gate, read new pointer

Reader r1 dec gate, do nothing (!= 0)
Writer dec gate, do nothing (!=0)
Reader r2 dec gate, do nothing (pointer hasn't changed)

.

0