- Boost-users - lists.boost.org

Re: [Boost-users] [boost.thread] destruction of condition variables
by Vladimir Pozdyaev 29 Jul '08

29 Jul '08

Howard Hinnant: > On Jul 28, 2008, at 6:49 AM, Vladimir Pozdyaev wrote: > > > > Looking at, e.g., > http://www.opengroup.org/onlinepubs/009695399/functions/pthread_cond_init.h… > > reveals that "It shall be safe to destroy an initialized condition > > variable upon which no threads are currently blocked." Am I right in > > understanding that the wait-destroy sequence should be quite safe > > this way (provided no one else waits)? > > I believe your code as shown is subject to "spurious wakeup". If the > wait ends (spuriously) before the notify_all() completes, or is even > initiated, then the notify happens on a destructed cv. POSIX says > that in this case the pthread_cond_broadcast will return EINVAL. I believe I should apologise for this distraction. I am aware of possibility of spurious wakeups; they are one reason for the claim that the original code is not an example of good practice. The point is, spurious wakeups have nothing to do with the problem at hand, so I just opted for a more concise code sample. > The sentence you refer to means that the thread doing the notifying > can safely destroy the cv even if the waiting thread has not yet > returned from the wait (as long as that thread or any other do not try > to initiate a new wait after the final signal/broadcast). Note though This is a somewhat narrower interpretation, since the original sentence does not consider roles of threads (notifying or waiting). Is it a matter of bad wording on the part of the latter? Because I don't see in what way does my code contradict the condition "no threads are currently blocked" (on a cv). AFAICT, after the only cv.wait has finished... > that it is not safe to destruct the mutex which is being waited on, > until the waiting thread wakes from the wait (locking the mutex) and > then unlocks that mutex. ... and the only (scoped) lock on a mutex released, no threads are blocked on them, so it shouldn't be problematic to destroy both mutex and cv, should it? (The problem doesn't seem to reproduce with a trunk version of the lib, so the point may well be moot anyway, but still I'm wondering...) ---- Vladimir

2 1

[CRC] Compilation fails with #include <boost/crc.hpp>
by Norbert Riedlin 29 Jul '08

29 Jul '08

Hi all, I want to use boost CRC, but when I #include <boost/crc.hpp>, I get the following error: "error: ISO C++ forbids variable-size array 'table_type'" in line 468 of crc.hpp. gcc --version tells me, I'm using the following gcc on my Mac (OSX 10.5): "i686-apple-darwin9-gcc-4.0.1 (GCC) 4.0.1 (Apple Inc. build 5484)" I use boost 1.34.1, but as crc.hpp of version 1.35 does not differ, I suppose updating the library would not help. The error is in the following context: <code> struct crc_table_t { BOOST_STATIC_CONSTANT( std::size_t, byte_combos = (1ul << CHAR_BIT) ); typedef mask_uint_t<Bits> masking_type; typedef typename masking_type::fast value_type; #if defined(__BORLANDC__) && defined(_M_IX86) && (__BORLANDC__ == 0x560) // for some reason Borland's command line compiler (version // 0x560) // chokes over this unless we do the calculation for it: typedef value_type table_type[ 0x100 ]; #else typedef value_type table_type[ byte_combos ];//<-- error here #endif static void init_table(); static table_type table_; }; // boost::detail::crc_table_t </code> I suppose it has to do with a bug of gcc, that cannot calculate "byte_combos" at compile time (or at least it acts as if it had to calculate "byte_combos" at runtime). A fix would be to enhance the "#if defined..." to the gcc versions with the said bug. I just realized there was an entry in the bug tracking system at sourceforge concerning this problem: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1478435&gro up_id=7586 but I have no permission to view the issue. The entry was made in early 2006. Is there any work on this issue? TIA Norbert P.S.: I also realize I can get around the problem by not using "-pedantic", but for me this is a valuable switch.

2 1

[BOOST::ASIO] How can I catch disconnect event from remote side ?
by Ken zhang 29 Jul '08

29 Jul '08

Hi Dear all. How can I catch the disconnect event from remote side in boost::asio? Thanks

2 1

Specifying which version of gcc to use when compiling boost
by Simon Lacroix 29 Jul '08

29 Jul '08

Hello all, I have a system with both gcc version 3.4 and gcc version 3.2 installed (distro is RHE4). I am trying to compile the filesystem package using 3.2 but for some reason it always seem to pick gcc 3.4. I tried the following (as specified in section 5.2.2 of the unix getting started guide) : ./configure .. -with-toolset=gcc-3.2.3 Performing a make after that tells me that it doesn't know anything about the tool and then give me a list of tool (basically all tools name with no version) Doing this however works but it uses version 3.4 ./configure .. -with-toolset=gcc Both cc and gcc executable points to gcc 3.2.3 the version that I want to use. Anybody can point me out what I am missing ? Thanks in advance ! Simon

3 2

Re: [Boost-users] compiling boost on AIX gives pthread errors
by Andrej van der Zee 29 Jul '08

29 Jul '08

Hi, > We are using AIX 5.3 and the 4.2.0 compiler and obviously require IBM's mpCC > MPI compiter to be supported, out-of-the-box. You should not be using the gcc compiler that is shipped with AIX 5.3 because it is hopelessly broken, just like their alternative binary version from the official IBM website. I wonder why? Note that IBMs compiler xlc/c++ costs almost 4000 Euro's for one user's lisence. And still I wonder why they ship a broken gcc compiler with their AIX distribution ;) Luckily you also can get binary distrubutions from here: http://gcc.gnu.org/install/binaries.html. Besides that there is a patch you should apply to the Boost source tree to make it all work (or you can use boost-trunk from SVN). Then you will be able to compile boost on AIX 5.3, but you will get some occasional complains from the linker and/or compiler. Anyway, I would be greatful if Boost would officially start to support gcc on AIX! Hopefuly it will be picked up soon... Cheers, Andrej __________________________________________________________ Not happy with your email address?. Get the one you really want - millions of new email addresses available now at Yahoo! http://uk.docs.yahoo.com/ymail/new.html

2 1

XML library?
by Michiel Helvensteijn 29 Jul '08

29 Jul '08

I was surprised not to find an XML parser/generator library in boost. Or is it hidden somewhere? Is it planned for a future release, perhaps? In the meantime, can anyone recommend another open source C++ XML library I can use (with DOM parser)? Thanks! -- Michiel Helvensteijn

8 10

Re: [Boost-users] compiling boost on AIX gives pthread errors
by Henrik Nagel 29 Jul '08

29 Jul '08

K. Noel Belcourt wrote: > > We don't test with gcc on Aix, but we should. > I am having trouble installing boost 1.35.0 on AIX too. How about porting boost to AIX? We have an IBM p575+ computer with AIX that you can get an account on for testing purposes. We are using AIX 5.3 and the 4.2.0 compiler and obviously require IBM's mpCC MPI compiter to be supported, out-of-the-box. Best regards, Henrik -- View this message in context: http://www.nabble.com/compiling-boost-on-AIX-gives-pthread-errors-tp1733930… Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

Suggestions for Simple Expression Evaluation
by Richard L. Aurbach 28 Jul '08

28 Jul '08

I am looking for some code to handle simple expression evaluation -- cases like "a | b | c", or "(a + b)/2", where 'a', 'b', and 'c' are named constants that I can evaluate by lookup. I.e., code to evaluate simple expressions to produce integer results. I'm looking for something light-weight and easy to use -- something that justifies a decision to look elsewhere rather than writing the 10- million-and-first simple evaluator. (I know that lots of folks have written this type of code as a homework assignment in a data structures course. As it happens, I don't have an old assignment to look up and am hoping to spare the time doing it now :-) ) I didn't see anything in the boost library that didn't appear to be incredible overkill, but I'm happy to be proven wrong about that. Anyway, I would be very interested in some suggestions (using boost or otherwise) that can help me implement run-time evaluation of simple expressions without a major investment in the size or complexity of my application. Cheers, Rick Aurbach Aurbach & Associates, Inc.

2 1

new version
by Gang Ma 28 Jul '08

28 Jul '08

Thanks, guys. I change my code a bit. The results are right. But I suspect it never used thread, but excute the thread one by one. new code void mtread(char* filename,const int stidx, const int edidx,int id) { fstream infile; infile.open(filename); string line; int linenum = 0; if (infile.good()){ while (! infile.eof()) { boost::mutex::scoped_lock lock(io_mutex); // infile.seekg(linenum,ios::beg); getline(infile,line); if (linenum>=stidx && linenum<=edidx){ cout<<id << " : " <<line<<endl; //cout<<linenum<<endl; linenum++; } else { linenum++; if (linenum>edidx) break; } } } else { cerr<< "file " << filename << " cannot be opened for reading" <<endl; } infile.close(); } I got the same results running the program twice. 1 : 1 1 : 2 1 : 3 1 : 4 1 : 5 2 : 6 2 : 7 2 : 8 2 : 9 2 : 10 Thank for your helps. regards, gma

1 0

Re: [Boost-users] Initialization of boost::asio::ip::tcp::socket geneates unhandled exception
by Igor R 28 Jul '08

28 Jul '08

Hi, The problem is somewhere in the design of your program, so it's hardly possible to find out it from this short description. Sounds like dangling reference/pointer - probably, io_service was destroyed/not initialized yet? (remember that the members are initialized in the order of their definition) 2008/7/28, Jean-Sebastien Stoezel <js.stoezel(a)gmail.com>: > Hi, > > I'm getting confused by the unhandled exception I get when > initializing a TCP socket. This happens when the socket is > instantiated and initialized with an IO service object, from the > constructor of a class called CProcessor. > > Here are the variables declaration, inside the class: > > boost::asio::ip::tcp::socket m_Socket; > boost::asio::io_service m_IoService; > > Here is what the default constructor looks like for this class: > CProcessor::CProcessor(void): > m_Socket(m_IoService) > {} > > The exception is generated by this constructor, as soon as the > m_Socket object is initialized. > > The type of unhandled exception I got is as follow: > "Unhandled exception at whatever in hatever.exe: 0xC0000005: Access > violation reading location 0xcccccce0." > The program breaks in win_mutex.hpp, at a line calling: > ::EnterCriticalSection(&crit_section_); > > I run VS9. > > Any idea?! > > Thanks, > Jean > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

2 1