December 2008 - Boost-users - lists.preview.boost.org

boost::interprocess on FreeBSD 7
by Andy Wiese 17 Dec '08

17 Dec '08

My apologies if this is a resend. I think I might have sent the first time before I was approved in the list. Using boost::interprocess from boost_1_37_0 on FreeBSD 7.0, specifically wishing to allocate boost::interprocess::map structures in shared memory pools. I get a segmentation fault in pthread_mutex_lock() from /lib/libthr.so. 3, in a variety of places including anything related to a ShmemAllocator. For example, when I run the example from http://www.boost.org/doc/libs/1_37_0/doc/html/interprocess/allocators_conta… I get the segfault at this line: //Initialize the STL-like allocator const ShmemAllocator alloc_inst (segment.get_segment_manager()); I encountered this segfault (always in pthread_mutex_lock) when trying to use mapped file boost::interprocess::map structure that works very well on darwin. Can anyone suggest what I might do to get this working on FreeBSD? Is anyone familiar with FreeBSD 7.0 and the scary inner workings of boost::interprocess interested in a small consulting project? ----------------------------------------------------------------------------------------------- FWIW, here is the complete text of the example, taken verbatim from the tutorial page: void test_ipc2() { using namespace boost::interprocess; shared_memory_object::remove("MySharedMemory"); try{ //A managed shared memory where we can construct objects //associated with a c-string managed_shared_memory segment(create_only, "MySharedMemory", //segment name 65536); //Alias an STL-like allocator of ints that allocates ints from the segment typedef allocator<int, managed_shared_memory::segment_manager> ShmemAllocator; //Alias a vector that uses the previous STL-like allocator typedef vector<int, ShmemAllocator> MyVector; int initVal[] = {0, 1, 2, 3, 4, 5, 6 }; const int *begVal = initVal; const int *endVal = initVal + sizeof(initVal)/sizeof(initVal[0]); //Initialize the STL-like allocator const ShmemAllocator alloc_inst (segment.get_segment_manager()); //Construct the vector in the shared memory segment with the STL- like allocator //from a range of iterators MyVector *myvector = segment.construct<MyVector> ("MyVector")/*object name*/ (begVal /*first ctor parameter*/, endVal /*second ctor parameter*/, alloc_inst /*third ctor parameter*/); BOOST_CHECK(myvector); //Use vector as your want std::sort(myvector->rbegin(), myvector->rend()); // . . . //When done, destroy and delete vector from the segment segment.destroy<MyVector>("MyVector"); } catch(...){ shared_memory_object::remove("MySharedMemory"); throw; } shared_memory_object::remove("MySharedMemory"); }

1 0

Re: [Boost-users] [boost] [non Boost] Template troubles.
by David Abrahams 17 Dec '08

17 Dec '08

on Tue Dec 16 2008, Noah Roberts <roberts.noah-AT-gmail.com> wrote: > Robert Jones wrote: >> On Fri, Dec 12, 2008 at 5:52 PM, Noah Roberts <roberts.noah(a)gmail.com>wrote: >> >>> Noah Roberts wrote: >>> >> >> Neil, Noah, Kim - Thanks all for your input. Although no perfect >> solution has emerged it is always good to see a few good minds >> looking at it. > > Don't recall if it was said already but you can also explicitly say > which type you're using: Seriously, this is OT (non-Boost after all), isn't it? As a moderator I overlooked the first few postings, but you should probably take this off the Boost lists now. There are appropriate newsgroups for this sort of thing. Thanks, -- Dave Abrahams BoostPro Computing http://www.boostpro.com

1 0

Re: [Boost-users] [boost] tweaking the review process
by David Abrahams 16 Dec '08

16 Dec '08

on Mon Dec 15 2008, John Phillips <phillips-AT-mps.ohio-state.edu> wrote: > David Abrahams wrote: >> on Fri Nov 21 2008, "Stjepan Rajko" <stjepan.rajko-AT-gmail.com> wrote: >> >>> On Thu, Nov 20, 2008 at 11:40 PM, vicente.botet >>> <vicente.botet(a)wanadoo.fr> wrote: >>>> I would like to make some suggestion to improve the review management: >>>> >>> Thank you for starting this discussion. >>> >>>> * I had notice only at the end of the the review that the review tokes place on > two >>> mailing lists (devel ans user). Maybe this is usaual on Boost reviews but I was not >>> aware. It will be more transaparent if all the reviews were posted to the same > mailing >>> list, maybe a specific one should be created. >>> >>> Yes, this is a bit of an issue. I mentioned that both lists are used >>> in my post that opened the review (in my notes to first-time >>> reviewers) to try to give people a heads-up (but... I often get >>> long-winded and I'm sure it's easy to miss parts of my posts :-)). >>> >>> Having a dedicated, or at least recommended mailing list (either dev >>> or user) might be a good thing. >> >> It has always been the -devel list, at least up until recently. How >> reviews began being posted to -users, I do not know. >> > > I don't know if I was the first to do this, but I have done it. Looking at the > current contents of the instructions to review managers, either they have changed > since I started managing reviews or I somehow hallucinated extra words to add the user > list to the lists that should receive an announcement of the review. An announcement is one thing; instructing people to post their reviews in the same list is a different thing entirely. -- Dave Abrahams BoostPro Computing http://www.boostpro.com

1 0

can't extend boost::any on windows
by Michael Linck 16 Dec '08

16 Dec '08

Hi, I've implemented a simple, low level json formatting utility to complement tinyjson's lightweight parser. To do this, I've extended boost::any only to add a "format()" function to it that is aware of what types the content can take in a json structure and produces the appropriate string. The default constructor and the constructor with a value are the only things I implemented, aside from the format function. Both just call the base constructor. Compiled in gcc, everything works fabulously. Apparently, a functioning assignment operator and copy constructor are appropriately inherited and everything runs great. On windows, however, some subtle difference causes assignment and copies to seg-fault. The real problem is that, for the life of me, I can't find a way to implement assignment or copies explicitly, that doesn't also bomb out. I can probably come up with a work around that stores my type in shared pointers so they're never actually copied, but I'm really curious if anyone could explain what causes this. Here are some code snippets to illustrate the behavior: void test(){ std::cout<<"messing around with boost any"<<std::endl; boost::any tAny; boost::any tStringAny(std::string("some string")); tAny = tStringAny; std::cout<<"string any assignable to any"<<std::endl; std::cout<<"messing around with value type"<<std::endl; sysshep::json::Value tValue; std::cout<<"formatted blank value "<<tValue.format()<<std::endl; std::cout<<"making string value"<<std::endl; sysshep::json::Value tStringValue(std::string("some_string")); std::cout<<"formatted string value "<<tStringValue.format()<<std::endl; std::cout<<"assigning string value to blank value"<<std::endl; tValue = tStringValue; /// FAILS std::cout<<"formatted blank value after assignment"<<tValue.format()<<std::endl; } class Value : public boost::any { public: Value():boost::any(){}; template<typename ValueTypeTA> Value(const ValueTypeTA & aValue): : boost::any(aValue) {} virtual ~Value(); std::string format() const; }; I'm not going to list the different ways I've attempted to implement copy and assignment, some more plausible than others, out of respect for brevity. If you try a build on windows, VC8, you should be able to mess around with this to your hearts content. It's really annoying because it was a factor in the fundamental elegance of the parsing implementation, because I see no reason why it shouldn't work, because it worked fine in linux, and because it put my windows port at least a day behind customer expectations. thanks for any insight.

3 3

issues when compiling graph library with expat
by Sean Farrow 16 Dec '08

16 Dec '08

Hi: When compiling the graph library with expat to add graph ml, the libraries fails to compile, I've declared environment variables EXPAT_INCLUDE and EXPAT_LIBRARY but it's failing expat.h even though the directory is in the pat as "c:\program files\expat\source\lib\" Any help apreciated. Sean.

3 5

CRC - error: shift count is too large
by Mike Dalgity 16 Dec '08

16 Dec '08

Hi, I'm attempting to get the Boost CRC to compile for a TI C6713 DSP. I'm using C6000 Code Gen Tools 6.1.6 which is based on EDG. STL implementation is from Dinkumware. When I include <boost/crc.hpp> I get the following errors: "C:/Program Files/boost_1_37_0/boost/crc.hpp", line 286: error: shift count is too large detected during instantiation of class "boost::detail::high_uint_t<Bits> [with Bits=64U]" at line 433 "C:/Program Files/boost_1_37_0/boost/crc.hpp", line 288: error: shift count is too large detected during instantiation of class "boost::detail::high_uint_t<Bits> [with Bits=64U]" at line 433 I'm new to Boost so I could use some help figuring this out. I looked at crc.hpp and high_uint_t appears to be a structure containing two uint_t types corresponding with the errors. In integer.hpp the uint_t struct does not evaluate long long. On this architecture a long is 48 bits and long long is 64 bits. I've only included the header so why is it trying to instantiate a 64 bit type? Thanks, Mike

1 1

[xpressive] Overlapping match results?
by Dave Jenkins 16 Dec '08

16 Dec '08

Is there a way to get overlapping match results from Xpressive? For example, when searching for "aa" in the string "aaaa", I want three results [as]as, a[aa]a, and aa[aa]. Currently, I'm only getting the two non-overlapping results [aa]aa and aa[aa]. I tried the regex 'a' >> before('a'). That almost gave me what I wanted (three results with the correct positions). The problem is that the lengths are wrong. What I need are lengths that include the before() characters. Is that possible? I know in this simple example I could just add 1 to the lengths, but I'm looking for overlapping results on more complex regexes. Thanks, Dave Jenkins

2 2

boost compile to arm
by leslie lige 16 Dec '08

16 Dec '08

Hi, I'm tring to cross compile boost 1.37 to arm, I found that I need to add "__arm__" definition to get compile correctly, but I don't know where do I define __arm__ macro will be global to all. Defautly, the arch is defined as __i386__ when I do cross compile with arm toolchain. regards, Leslie _________________________________________________________________ Explore the seven wonders of the world http://search.msn.com/results.aspx?q=7+wonders+world&mkt=en-US&form=QBRE

2 3

building regex with icu
by Sean Farrow 16 Dec '08

16 Dec '08

Hi: what aditional command line switches/parameters do I need to add to bjam to build regex with icu? Chers Sean.

4 3

this_thread::sleep + undefined reference?
by John D 16 Dec '08

16 Dec '08

Hello This will be an exceptionally stupid problem for whomever is kind enough to clarify a cause and solution. I will keep it brief in hopes that it is not an issue in need of complexity. The code: boost::this_thread::sleep(boost::posix_time::milliseconds(20)); The attempt: g++ -lcurl -lboost_thread -lpthread -pthread c.cpp -o app /tmp/cc2YRa76.o: In function `void boost::this_thread::sleep<boost::date_time::subsecond_duration<boost::posix_time::time_duration, 1000ll> >(boost::date_time::subsecond_duration<boost::posix_time::time_duration, 1000ll> const&)': c.cpp:(.text._ZN5boost11this_thread5sleepINS_9date_time18subsecond_durationINS_10posix_time13time_durationELx1000EEEEEvRKT_[void boost::this_thread::sleep<boost::date_time::subsecond_duration<boost::posix_time::time_duration, 1000ll> >(boost::date_time::subsecond_duration<boost::posix_time::time_duration, 1000ll> const&)]+0x50): undefined reference to `boost::this_thread::sleep(boost::posix_time::ptime const&)' collect2: ld returned 1 exit status I will post more code if need be, but even general guidance will likely lead me to a solution. Thanks John

4 6