Forums before death by AOL, social media and spammers... "We can't have nice things"
|    comp.lang.c++.moderated    |    Moderated discussion of C++ superhackery    |    33,346 messages    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
|    Message 31,575 of 33,346    |
|    Alexander Terekhov to All    |
|    Re: atomic counter    |
|    17 Oct 11 14:50:10    |
      0b751a3e       From: terekhov@web.de              SG wrote:       [...]       > on. As long as programmers have a vague idea of what a data race is,       > what they can do to avoid them (amotics, . . .              Yeah, right, 'amotics' ;-)              Lock-free atomics are inherently racy and the default memory model for       lock-free atomics shall be the most relaxed one, not sequentially       consistency (SC).              On PPC/POWER/PowerPC, for SC, you would need most heavy fence 'hwsync'       between each pair of SC lock-free atomic operations, there are examples       showing this for each pair: RR, RW, WR, and WW.              Lock-based atomics do not need 'hwsync' instruction but nobody needs       lock-based atomics because locking can be done more efficiently on       higher level.              regards,       alexander.                     --        [ See http://www.gotw.ca/resources/clcm.htm for info about ]        [ comp.lang.c++.moderated. First time posters: Do this! ]              --- SoupGate-Win32 v1.05        * Origin: you cannot sedate... all the things you hate (1:229/2)    |
[   << oldest   |   < older   |   list   |   newer >   |   newest >>   ]
(c) 1994, bbs@darkrealms.ca