Le 12/01/13 14:26, Andrey Semashev a écrit :
On Saturday 12 January 2013 12:08:10 Vicente J. Botet Escriba wrote:
Anyway, can Boost.Thread be modified in such a way so that Boost.Atomic use is not exposed to the user? E.g. so that call_once invokes a compiled function implemented within Boost.Thread library that uses Boost.Atomic to modify the once flag. yes, this will be great. I don't know Boost.Atomic details to try to do
Le 12/01/13 11:51, Andrey Semashev a écrit : this. Andrey do you mind to provide a patch that doesn't needs to link with boost_atomic? I attached the patch (for posix only). It appeared a bit hacky and I'm not sure if you're ok with it.
Or, can Boost.Atomic be header-only? It can't be completely header-only because it requires a spinkock pool singleton. You can find the discussion about it in the nearby thread "[boost] [atomic] [release] possible linking problem with atomic" and other older threads.
Thanks for the quick patch. Shouldn't the patch concern only boost/thread/pthread/once.hpp and libs/thread/src/pthread/once.hpp? Could you send the resulting files also? Thanks again, Vicente