- Boost-users - lists.boost.org

Boost Threading Multiple Reader / Single Writer Usage Questions
by Brian Nantz 19 Feb '10

19 Feb '10

I am working in legacy code and I have a list collection of pointers that is accessed by multiple threads. I am having some thread locking issues with it. Mostly the list is read and very few changes (adds or deletes) are made to the list. I thought this would be a great opportunity to use a Reader Writer Lock (RWL). It seems that the shared_mutex is the Boost RWL. There are a few requirements that I am not sure shared_mutex will do for me: 1. There could be nested locks and the RWL should handle unwinding them correctly. I see that recursive_mutex seems to handle that. Does shared_mutex handle that as well? 2. Is there a way to give a higher priority to writes? I am worried about the readers starving the writers, so I would like any pending writers to be given preference over pending readers. How can I get this functionality from the Boost thread library? If I need to build my own, what class do you recommend using as a base to build from?

2 1

Re: [Boost-users] Object pools?
by michele.caini＠gmail.com 19 Feb '10

19 Feb '10

Hi, of course, in my opinion pool has to manage objects life time. However sharing objects with pool owner by shared ptr seems to be a good idea, because pool still manages objects life time. Using shared pointer also inside pool solves the problem, maybe. It is correct? ----- Reply message ----- Da: "Ryan McConnehey" <mccorywork(a)gmail.com> Data: gio, feb 18, 2010 02:50 Oggetto: [Boost-users] Object pools? A: <boost-users(a)lists.boost.org> Michele Caini wrote: > Pool will create objects, pool will release and reclaim objects, and pool will delete > objects. > If the pool manages the life time of the object then the pool can't release the object, correct. The pool will at most loan the object to the user. > You are looking for something that create objects and forbid about them, > until you return that objects to the home. Is correct? > Did you mean forget instead of forbid. The kind of pool I'm talking about would only forget about those objects that it released. It still needs to manage the life time of the objects it didn't release. > Really, it's possible but it is not as a pool works, simply. > Are you stating that Boost.Pool doesn't work this way or all pools? Is there a definition for a pool that I'm missing. I thought a pool was a collection of objects and an object could be provided to the user. The implementation and life time of the object were left up to the pool. > Your idea breaks also raii pattern and you risk to have memory leaks, > because not reclaimed objects maybe be lost during their life time. > In my requirements, to avoid memory leaks, the objects were passed to the user with a boost::shared_ptr. > What about a monostate pattern that globally manages objects as a pool > and will be destroyed only when program terminates? That's can be a > valid alternative, or not? > I believe the monostate pattern shares its (static) data. Wouldn't this mean the amount of data can't be increased? If the data can't be increased it not as useful to the user when all the objects have been used but more are still needed. Ryan _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

4 6

[1.41.0] Intel 11.1 Linux vs. cmath header
by Richard Hadsell 19 Feb '10

19 Feb '10

I am trying to build the Boost 1.41.0 libraries with an evaluation copy of Intel C++ 11.1 for 64-bit Linux. My system is Fedora 12 with G++ 4.4.3 installed as the default compiler. This fatal error is preventing most of the libraries from building: /usr/include/c++/4.4.3/cmath(500): error: identifier "__builtin_fpclassify" is undefined return __builtin_fpclassify(FP_NAN, FP_INFINITE, FP_NORMAL, ^ detected during: instantiation of "__gnu_cxx::__enable_if<std::__is_arithmetic<_Tp>::__value, int>::__type std::fpclassify(_Tp) [with _Tp=float]" at line 65 of "./boost/functional/hash/detail/hash_float.hpp" instantiation of "size_t={unsigned long} boost::hash_detail::float_hash_value(T) [with T=float]" at line 282 of "./boost/functional/hash/hash.hpp" I am building the libraries with this command: bjam toolset=intel-linux threading=multi variant=release link=static,shared --stagedir=stageIF12 stage Release notes seem to indicate that someone has successfully built Boost libraries with Intel 11.1 and G++ 4.4.1 (whose cmath header looks the same as for 4.4.3). I have not used the Intel compiler for several years, so I may be missing something that is familiar to other users. Help? -- Dick Hadsell 203-992-6320 Fax: 203-992-6001 Reply-to: hadsell(a)blueskystudios.com Blue Sky Studios http://www.blueskystudios.com 1 American Lane, Greenwich, CT 06831-2560

2 2

[spirit] first steps
by Igor R 18 Feb '10

18 Feb '10

Hello, How to use the string parser? I tried this way: std::string s1; bool r = qi::phrase_parse(first, last, ( string[ref(s1) = _1] ) , space); But it doesn't compile. Thanks.

4 15

mpl for_each: stop applying functor when runtime predicate is true
by Hicham Mouline 18 Feb '10

18 Feb '10

With a mpl sequence, applying a functor f with for_each on that sequence applies f at runtime on every element of the sequence. If I wish to stop applying f once I've found the element in the sequence which satisfies some runtime predicate, how can I do that? I'm thinking of putting a bool state in f, and turn it to false once I decide not to apply anymore, and on invocations of, just do nothing when the state is false. Any better way of doing this? Regards,

1 0

[Iostreams] unresolved problem
by Michał Nowotka 18 Feb '10

18 Feb '10

Hello, I have the same problem as described here: http://thread.gmane.org/gmane.comp.lib.boost.user/34531 . There wasn't any answer to this post and it was asked over a year ago. I think this is quite important. Can anyone can help? -- Regards Michał Nowotka

1 0

tokenize string delimiter
by Lloyd 18 Feb '10

18 Feb '10

Hi, Is it possibleto tokenize a string based on a string delimiter? Thanks, Lloyd ______________________________________ Scanned and protected by Email scanner

11 19

fusion: work of 4 sequences at the same time
by Hicham Mouline 18 Feb '10

18 Feb '10

Hi, I am trying to apply fusion::for_each on a sequence, but inside the unary functor applied to each element, I need to know the index of the element in question, because I need to extract elements with the same index from other sequences. fusion::vector<...> fseq1; fusion::vector<...> fseq2; fusion::vector<...> fseq3; fusion::vector<...> fseq4; fusion::for_each( fseq1, functor<...>(fseq2, fseq3, fseq4) ); template <typename Seq2, typename Seq3, typename Seq4) struct functor { /// ctor template <typename T> void operator()( const T& t) const { /// find out index of t in fseq1, /// use elements from fseq2,3,4 at the same index } }; I could use boost::fusion::find to get an iterator and then get its index, but this doesn not look clean, Any ideas, Rds,

2 2

Re: [Boost-users] fusion: work of 4 sequences at the same time
by Hicham Mouline 18 Feb '10

18 Feb '10

> -----Original Message----- > From: Hicham Mouline [mailto:hicham@mouline.org] > Sent: 18 February 2010 17:38 > To: 'boost-users(a)lists.boost.org' > Subject: fusion: work of 4 sequences at the same time > > Hi, > > I am trying to apply fusion::for_each on a sequence, but inside the unary > functor applied to each element, I need to know the index of the element in > question, because I need to extract elements with the same index from other > sequences. > > fusion::vector<...> fseq1; > fusion::vector<...> fseq2; > fusion::vector<...> fseq3; > fusion::vector<...> fseq4; > > fusion::for_each( fseq1, functor<...>(fseq2, fseq3, fseq4) ); > > template <typename Seq2, typename Seq3, typename Seq4) > struct functor { > /// ctor > > template <typename T> > void operator()( const T& t) const > { > /// find out index of t in fseq1, > /// use elements from fseq2,3,4 at the same index > } > }; > > I could use boost::fusion::find to get an iterator and then get its index, but > this doesn not look clean, > > Any ideas, > > Rds, > Ah! fusion::zip is my friend. fusion::for_each( fusion::zip(fseq1, fseq2, fseq3, fseq4), functor() ); Could u confirm though that I will be able to modify fseq1, even through the zipping? Rds,

1 0

Transfering the process state to another process
by Lloyd 18 Feb '10

18 Feb '10

Hi, We want to have a program which cannot be terminated by the user (even admin). I think the only way to do this is to write a watchdog process. So there will be a master and a slave process, the master will be the active process, and the slave keep on watching the master. I case if the master is terminated, the slave become master and initiates a new slave process. But we need to transfer (or share) the process state of "old master" with the "new master" (So that the new master can continue form where the old master has stopped). Is this possible with shared memory? Or is there is a better way? Thanks, Lloyd ______________________________________ Scanned and protected by Email scanner

8 11