- Boost-users - lists.boost.org

Re: [Boost-Users] Re: Newbie needs pointer to currently executing thread.
by dick.bridges＠tais.com 10 Apr '02

10 Apr '02

I'm implementing the Leader/Followers thread pool pattern - reference "Pattern-Oriented Software Architecture Volume 2" (Schmidt, et al). When a thread is released via notify_one(), it is [assumed to be] in the "leader" role and stores it's id. A sanity check should be performed to verify that there is no other thread in the leader role (i.e., current_leader is "none") before the thread stores it's id there. When one of the event sources is signaled and the leader is unblocked, it must, among other tasks, demote itself (i.e., set the current_leader value to "none") and promote one of the followers by notify_one() before continuing to service the request. Because of the potential for race conditions, (especially in the context of future program updates, modifications, etc.) I should enforce the rule that only the leader can promote a follower. Something along the lines of: current_leader == Thread::this(). "dmoore99atwork" <dmoore@ashtontec To: Boost-Users(a)yahoogroups.com hgroup.com> cc: Subject: [Boost-Users] Re: Newbie needs pointer to currently executing thread. 04/09/2002 06:53 PM Please respond to Boost-Users --- In Boost-Users@y..., dick.bridges@t... wrote: > Threads in a thread_group - derived pool are waiting on a condition. > Condition gets notify_one(). How do I get id/pointer for the thread that > was released? I've got a feeling I'm gonna be *REAL* embarrassed when I > find out, but I can't figure it out. > > Thanks for any help. Dick, In general, condition variables (both the Boost.Threads version and the POSIX version) don't tell the "notifier" which thread woke up. That's not necessarily a problem, though. A thread pool usually implies a set of threads that are equivalent in some way, even running the same thread function. Referring to the example at: http://www.boost.org/libs/thread/doc/condition.html#Example, Take a look the the receive() function. while (buffered == 0) buffer_not_empty.wait(lk); You might have three threads which reach this point when buffered==0 (meaning the buffer is empty). When a new sender in a fourth thread adds something to the buffer and calls notify_one(), exactly one of the three threads waiting on the condition will wake up and do some work. You won't care which thread wakes up, because any one of them will be prepared to do the necessary work. If this doesn't get you unstuck, please post more details, in particular why you feel it is important to know which thread "woke up" Regards, Dave Info: <http://www.boost.org> Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl> Unsubscribe: <mailto:boost-users-unsubscribe@yahoogroups.com> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

2 1

Problem with File Dependency Example
by Martin Sander 10 Apr '02

10 Apr '02

Hi! (I'm a newbie to boost and English is not my native language) I changed the examples/file_dependencies.cpp from boost 1.0.27 a little bit and I think the output is wrong. The changed File is at http://www.ebx.de/qqq.cpp (just files_e, name and used_by are changed). The program printed "FF" and "GG" with the same time_slot, but they have an "Edge". With "#if 0" instead of "#if 1" on line 106 the result get even worse, "DD" has a lower time_slot than "GG", but "GG" depends on "DD". What is wrong? My code? Did I misuse the example? M.

1 0

Cuthill Mckee Ordering Algorithm
by Sukandar, Erman H. 10 Apr '02

10 Apr '02

On the Graph documentation it is shown an example showing how to get reverse Cuthill-McKee ordering. It uses rbegin() to get it. My question is: how to get the non-reverse Cuthill-McKee ordering. It still give the reverse ordering even after I use begin(). Did I do something wrong? Can someone advise me the correct one (with example, if possible). Erman

1 0

Detect derivation from e.g. addable
by whastingsnj 10 Apr '02

10 Apr '02

Is there a way to detect that a class has been derived from boost::addable<T>, for example? I would like to be able to write something like this: template <typename T> void foo() { //if T derived from boost::addable<T,...> // . . . } The problem, of course, is the base class chaining feature. I do not see how to use boost::is_base_and_derived in this case.

2 1

Newbie needs pointer to currently executing thread.
by dick.bridges＠tais.com 10 Apr '02

10 Apr '02

Threads in a thread_group - derived pool are waiting on a condition. Condition gets notify_one(). How do I get id/pointer for the thread that was released? I've got a feeling I'm gonna be *REAL* embarrassed when I find out, but I can't figure it out. Thanks for any help.

2 1

Where it is Boost.Signal?
by Raul Huertas Diaz 08 Apr '02

08 Apr '02

I kept reading about a library named Boost.Signal. I think that this library is about signal and slots. But I can't find it in Boost. Can someone tell me where it is, please? I look for it in the Files section and in the CVS repository but i can't find it. Thanks in advance. -- Raúl Huertas Díaz rhuertas at tcpsi dot es http://www.tcpsi.es

2 1

Re: [Boost-Users] Building graphviz under VC6
by Wolf, Frank 08 Apr '02

08 Apr '02

Sorry for being inaccurate and thus wasting bandwidth. I meant the part under ...\libs\graph\src\ for the use of read_graphviz I really would be grateful for any help Thanks, Frank -----Ursprüngliche Nachricht----- Von: Jeff Garland [mailto:jeff@crystalclearsoftware.com] Gesendet: Mittwoch, 3. April 2002 17:49 An: Boost-Users(a)yahoogroups.com Betreff: RE: [Boost-Users] Building graphviz under VC6 Graphviz is not a boost library. This question would be more appropriate for graphviz-interest(a)research.att.com. Jeff > -----Original Message----- > From: Wolf, Frank [mailto:Wolf@bms.de] > Sent: Wednesday, April 03, 2002 1:56 AM > To: Boost-Users(a)yahoogroups.com > Subject: [Boost-Users] Building graphviz under VC6 > > > Has anyone succesfully build the graphviz libs under VC6? > I tried to do so, but ran into fatal error LNK1179 > > I really would be grateful for any help > Thanks. > Frank > Info: <http://www.boost.org> Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl> Unsubscribe: <mailto:boost-users-unsubscribe@yahoogroups.com> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

1 0

Tokenizer
by macrodisk 05 Apr '02

05 Apr '02

I have a question about boost::tokenizer. I have a Semicolon separated string, something like: "All;;;;;;Bus;;;;;;BRT;;;;;;All" When I use boost tokenizer, I get "All", "Bus", "BRT",and "All". I want to get the the empty string between the semicolons, seems like the tokenizer takes consecutive semicolons as a single token; is there any way I can change this to represent different strings? boost::char_separator<char> sep(";"); typedef boost::tokenizer<boost::char_separator<char> > tokenizer; tokenizer Id_tokens(IdString, sep); for(tokenizer::iterator Id_tok_iter= Id_tokens.begin(), End_Id_tok_iter = Id_tokens.end(); Id_tok_Iter != End_Id_tok_iter) { //Do Something } Another question, is there some way to assign a different string once a tokenizer is constructed? tokenizer Id_tokens(IdString,sep); Id_tokens.assign(NewString,sep); //Is this possible with the boost tokenizer? I couldn't find any examples to this end in the places I searched. Thank you, Vijay

3 4

Sun WorkShop assertion failed when compiling with -g
by the_d.geo 05 Apr '02

05 Apr '02

FYI, the following piece of code is an excerpt from boost/type_traits/composite_traits.hpp (boost version 1_27_0). Using sun's CC (Sun WorkShop 6 update 2 C++ 5.3. Patch 111685-05 2002/02/03) and debug on (-g), it fails compilation with the following message >> Assertion: (../links/dbg_cstabs.cc, line 611) while processing test.cpp at line 0. I'd assume that it is a bug in the compiler since it compiles fine without the -g flag. --- start code #include <boost/type_traits/ice.hpp> template <typename T> struct is_pointer { BOOST_STATIC_CONSTANT(bool, value = (::boost::type_traits::ice_and< ::boost::detail::is_pointer_helper<T>::value, ::boost::type_traits::ice_not< ::boost::is_member_pointer<T>::value >::value >::value)); }; --- end code

1 0

preprocessor bug
by jkharris01 04 Apr '02

04 Apr '02

Using MSVC6, SP5: BOOST_PP_LIST_AT() doesn't work in a transform function, but BOOST_PP_LIST_FIRST() and BOOST_PP_LIST_REST() work fine I expected the following code to expand to: a , b but instead it expands to BOOST_PP_TUPLE3_ELEM0 BOOST_PP_TUPLE2_ELEM0 BOOST_PP_WHILE0(BOOST_PP_LIST_REST_N_C,BOOST_PP_LIST_REST_N_F,((a,(7,(_,_,0),1),1),0)) , BOOST_PP_TUPLE3_ELEM0 BOOST_PP_TUPLE2_ELEM0 BOOST_PP_WHILE0(BOOST_PP_LIST_REST_N_C,BOOST_PP_LIST_REST_N_F,((b,(8,(_,_,0),1),1),0)) If I change the definition of F to use BOOST_PP_FIRST(X), it works ************* Example code*************** #define F(a,b,X) BOOST_PP_LIST_AT(X,0) #define LST BOOST_PP_TUPLE_TO_LIST(2,( \ BOOST_PP_TUPLE_TO_LIST(2,(a,7)), \ BOOST_PP_TUPLE_TO_LIST(2,(b,8)))) BOOST_PP_LIST_ENUM(BOOST_PP_LIST_TRANSFORM(F,dummy,LST))

1 0