Re: gcc atomic word

[David Sugar]

This sounds very interesting, and I would like to see if we can add it 
for the next point release.  One area I have been thinking of is adding 
support for lockfree data structures.  Whether we work with an existing 
lockfree library or not, I would like to get our detection and use of 
atomic operations in Common C++ clean enough so that we can look at 
making at least some of the core Common C++ classes which currently rely 
on mutex locks optionally able to use atomic lockfree algorithms. 
Functional atomic counter operations seem a good place to start.

Valient Gough wrote:
> I didn't see a link for archives of this list, so I don't know if this 
> has been discussed.
>
> I have uploaded a patch to commoncpp2-1.3.1 which adds support for gcc's 
> _Atomic_word counter (in bits/atomicity.h).
> Changes:
>
> - m4/ost_pthread.m4 is modified to check for support
> - configure.ac is modified to specify C++ as the language to use for 
> tests (I don't know why it wasn't already, since this is a C++ library)
> - include/cc++/thread.h and src/mutex.cpp are modified with the 
> implementation.
>
> I've only verified that it correctly detects support for my compiler and 
> builds, nothing more.
>
> regards,
> Valient
>
>
>
> _______________________________________________
> Bug-commoncpp mailing list
> address@hidden
> http://lists.gnu.org/mailman/listinfo/bug-commoncpp

dyfet.vcf
Description: Vcard