December 2012 - Boost - lists.boost.org

Boost SIMD beta release
by joel falcou 25 Dec '12

25 Dec '12

We're rolling a beta release of Boost.SIMD along with a beta release of NT2. This version comes very late because we were busy with other things but hopefully we'll now do rolling releases frequently. Documentation is quite lacking but code is relatively stable and works. * Release Files can be found here: http://nt2.metascale.org/downloads/ * Additional material about the library can be found here : - Boost.SIMD PACT paper : http://www.lri.fr/~falcou/pub/pact-2012.pdf - Boost.SIMD MeetingC++ slides : http://meetingcpp.com/tl_files/mcpp/slides/12/simd.pdf * We welcome comments, questions and random requests/rants/bug reports as to polish the library before actual submission to the boost review. Such comments can be sent there, on our nt2-dev Google groups or in the nt2 issue tracker. Regards

13 57

automatic SVN mirror down?
by Robin Eckert 24 Dec '12

24 Dec '12

Hello, The latest commit on the live boost SVN mirror (git://github.com/ryppl/boost-svn.git) is 11 days old. Is the automatic mirror jammed right now or is the mirror probably shut down due to the modularization effort? Regards, Robin

2 2

[GitHelp] Help desk ready to assist
by Beman Dawes 24 Dec '12

24 Dec '12

See https://svn.boost.org/trac/boost/wiki/GitHelp for details. --Beman

1 0

[Git] ** Attention All Boost Committers **
by Dave Abrahams 24 Dec '12

24 Dec '12

If you have *ever* checked *anything* (e.g. code or documentation) into the Boost Subversion repository, this message applies to you. Git uses email addresses to identify authors and committers. In order to accurately convert Subversion history to Git, we need to map all (or as many as possible) Subversion user IDs to working email addresses. I have posted a provisional UserID mapping table at https://github.com/boost-lib/boost/wiki/AuthorMapping (with a simple mangling of '@'=>'<>' and '.'=>'~'). If your email address is already in the table, you've been Bcc'd on this message, which may explain the appearance of multiple copies in your mailbox. I request that you make corrections to this table. To edit the table directly, you'll need to be logged into GitHub. Direct correction is by far the preferable approach, but if you don't have---and don't want to acquire---a free GitHub account, you can send edits directly to me. Important notes (these also appear on the page with the table) * You may have several userids (e.g. I appear to have "dave," "abrahams," and "david_abrahams" --- probably left over from the transition from SourceForge CVS to our own SVN). Please look carefully to be sure that you have accounted for all your userids, and please use the same name and email address for each of them. Please *make sure that all non-duplicate userids stay in the table*. For example, all three of my userids are still there. * A given userid may appear several times in the table. If that userid belongs to you, please delete all but one line. If you don't, we will just use the first entry with that userid. * If you have a GitHub account, please make sure the email address you use in the table is associated with that account at https://github.com/settings/emails Thanks for your cooperation. -- Dave Abrahams BoostPro Computing Software Development Training http://www.boostpro.com Clang/LLVM/EDG Compilers C++ Boost

3 2

[TypeErasure] Similar simple class.
by Arman Hunanyan 23 Dec '12

23 Dec '12

Hi All, I have implemented xany class https://sourceforge.net/projects/extendableany/?source=directory. Then I noticed that there is a Boost.TypeErasure library which allows to do same as xany does and something more. Implementation of class xany has 1005 lines and it is implemented and tested in a week. Of course xany isn't comparable with Boost.TypeErasure, but it may be interesting to look. Below is a definition of same method for Boost.TypeErasure and for xany. // definition of push_back for Boost.TypeErasure template<class C, class T> struct push_back { static void apply(C& cont, const T& arg) { cont.push_back(arg); } }; namespace boost { namespace type_erasure { template<class C, class T, class Base> struct concept_interface< ::push_back<C, T>, Base, C> : Base { void push_back(typename rebind_any<Base, const T&>::type arg) { call(::push_back<C, T>(), *this, arg); } }; } } any<push_back<_self, int>, _self&> c; // definition of push_back for xany template <typename Xany, typename T> struct push_back_method { typedef void (Xany::* signature) (const T&); template <typename B> struct wrapper : public B { void push_back(const T& e) { this->call(push_back_method(), e); } }; struct implementation { template <typename Container> void operator() (Container& c, const T& e) { c.push_back(e); } }; }; xany<boost::mpl::vector<push_back_method<boost::mpl::_1, int> > > c; xany requires additional signature type to be specified. This allows to use dispatching in implementation. struct implementation { void operator() (OtherCodingStyleContainer& c, const T& e) { c.PushBack(e); } template <typename Container> void operator() (Container& c, const T& e) { c.push_back(e); } }; Documentation contains more examples.

1 0

[Inspect] hadn't run since... August
by Artyom Beilis 23 Dec '12

23 Dec '12

Boost.Inspect hadn't run on trunk or release sinse august: From: http://beta.boost.org/development/testing.html trunk - http://boost.cowic.de/rc/inspect-trunk.html release - http://boost.cowic.de/rc/inspect-snapshot.html Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.com/ CppDB - C++ SQL Connectivity: http://cppcms.com/sql/cppdb/

2 1

[atomic, x86] need another pair of eyes
by Tim Blechmann 23 Dec '12

23 Dec '12

hi all, i need another pair of eyes regarding boost.atomic on x86: the implementation of memory barrier is merely a compiler barrier, but not a CPU barrier, as it is using code like: __asm__ __volatile__ ("" ::: "memory"); afaict, one should use a `real' CPU barrier like "mfence" or "lock; addl $0,0(%%esp)". is this correct? apart from that, i've seen that compare_exchange is using explicit memory barriers before/after "cmpxchg" instructions. i somehow though that cmpxchg and the 8b/16b variants implicitly issue a memory barrier, so the resulting code would generate multiple memory barriers. can someone with some insights in the x86 architecture confirm this? thanks, tim

6 35

Re: [boost] Boost SIMD beta release
by Brian Wood 23 Dec '12

23 Dec '12

From: Mathias Gaunard: On 20/12/2012 20:34, Dave Abrahams wrote: >> Yes; there's a runtime test. > > Let's put a runtime test at every step of our critical loop, sounds like > a good idea. > I actually somewhat looked into this, and in this case segmented > iterators don't allow any optimization compared to normal iterators. I think less sarcasm would be helpful. >> I'm afraid you have not been specific enough about the >> problem you claim to see for me to be able to evaluate >> whether it exists or not. > > I'm afraid you're the one trying to make everything fit into this > model you've somehow taken a liking to. I'd prefer to focus on results rather than motives. Shalom Brian Ebenezer Enterprises http://webEbenezer.net (651) 251-9384

1 0

[BTREE] Get BTREE library into boost - Market Research
by Joel Young 23 Dec '12

23 Dec '12

My organization is investigating contracting for enhancements to the boost::btree library developed by Beman Dawes with all enhancements to remain under the boost license and author(s) retaining all copyright. No travel is expected to be required as part of this project. We are considering the following enhancements: 1. Develop detailed developer and tutorial documentation in customary boost formats 2. Enhancing the interface to provide more explicit preload and caching control to include the amount of RAM to reserve, perhaps with the ability to specify posix alignment to maximize benefit of linux anonymous huge page support. 3. Make the library C++ 11 ready 4. Finalize data layout and support for variable length payloads 5. Provide convenient interface for storing std::string (or the contents thereof) in payloads 6. Prepare the library for submission for formal review Any comments and suggestions on this are greatly appreciated. Of particular interest are discussions of technical approaches and estimates of hours required. Please do not send formal proposals. In addition to replying on this list (preferred) I can be reached at the telephone numbers and email below. Comments sent by 21 December will be most valuable. Discussion after that is still very welcome! Please don’t include any proprietary data. Also note that I am not authorized to commit the government in any way--this is simply market research. Thanks! Joel Dr. JOEL D. YOUNG Assistant Professor Computer Science Department Naval Postgraduate School Glasgow East Rm 338 jdyoung(a)nps.edu (831) 656-3518

6 11

Security bugs policy
by Artyom Beilis 23 Dec '12

23 Dec '12

Hello, I would like to know if there is any policy about security bug fixes, updates or releases. What I'm talking about. Consider a security related bug discovered in boost library. For example: For example recently reported bug in Boost.Locale that will be fixed in 1.53 https://svn.boost.org/trac/boost/ticket/7743 I consider it security bug as an application that validates the input (checks if it is UTF-8) may accept an invalid sequence. Or very critical bug in UUID (fixed in 1.43): https://svn.boost.org/trac/boost/ticket/3971 That any application that used them may be target to UUID guessing. These kind of bugs in general should be considered critical and somehow published, it may be also important to backport the fixes to stable releases, because not everybody can move forward and they may need to stick with older releases (as Boost is neither ABI nor API backward compatible even between minor releases). Annoucements of such bugs may be critical for example for stable Linux distributions that ship some older Boost versions and likely need to backport such bug fixes to their own boost tree because Boost doesn't do it. Questions: - Do we have any policy about it? - If not we probably need one including backporting of security bug fixes to stable releases. What brings me to other issues... Should Boost support older releases with critical bug fixes/security bug fixes? For how long time? Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.com/ CppDB - C++ SQL Connectivity: http://cppcms.com/sql/cppdb/

2 1