April 2003 - Boost-users - lists.boost.org

boost::shared_ptr and stl algorithms
by Bill Eidson 15 Apr '05

15 Apr '05

std::vector< boost::shared_ptr<T> > TVec; I want to write a function object to run std::for_each. to do somework against members of T and accumulate results in functor. My attempts have failed in VC7 with memory errors in my operator() method. struct fo { bool operator()( boost::shared_ptr<T>& t ) { return true; } }; Thanks in advance bille

5 9

BGL and reversing / moving edges
by joshrose 17 Dec '03

17 Dec '03

G'day, I was hoping someone might be able to tell me if they've run into any situations using BGL where they needed to move (or reverse (i.e. flip the source and target)) an edge within a graph implemented as an adjacency_list. I'm working on an implementation of an algorithm for max. weight bipartite matching ("the assignment problem"), and one of the assumptions in the algorithm is that non-matched edges are directed from one subgraph, A, to the other subgraph, B, and matched edges are directed the other way. Over the course of growing the matching, a simple way to add and remove edges from the candidate set is just to reverse the edge, thus the need for some way to reverse the edge within BGL. I don't think I can just remove and then re-add the edge in the other direction, primarily because doing so will (I believe) invalidate any edge_descriptors referring to that edge. In particular, there are property maps keeping track of the weight of edges, the predecessors of certain nodes, etc. that use edge descriptors as keys. I've spent a fair bit of time reading the documentation and looking through the implementation of adjacency_lists (incl. edge_list_impl), and I'm not really sure of a way to either move an edge directly within a graph, or even to create a new edge from another s.t. they have the same edge_descriptor (in general a bad thing, probably). Anyone have any thoughts on this? Thanks, -Josh

4 10

Graph Library: Ordered Vertex List?
by dmatt001 01 Aug '03

01 Aug '03

Hi Guys, I would like to use an ordered vertex list (ie based on property type) in an implementation that I am working on such that duplication can be avoided. I have read and understand how to implement an ordered edge list thanks to the previous posts but am finding it hard to get started! Perhaps I have missed the key example? It appears that from the way config::VertList is constructed that it may be a similar process in nature? adj_list_impl::add_vertex appears to discard the boolean return value from the associative insertion. Would such a construction allow me to avoid the duplication I seek to avoid? I am guessing so, but some advice would give me a confidence "boost". ( <-- It's a lame pun: I really do need help ;-) ) My first steps are below rehashing the excellent ordered edges example with a simple search/replace and maching my target container base. I'll keep plugging along with it anyway and it may just compile one day. Suggestions/Ideas/Refs/Ptrs are appreciated. Matt --- first guess follows --- template <class VertDesc> struct order_by_name : public std::binary_function<VertDesc,VertDesc,bool> { bool operator()(const VertDesc& v1, const VertDesc& v2) const { // Order by vertex name. return boost::get(boost::vertex_name, v1) < boost::get(boost::vertex_name, v2); } }; #if !defined BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION struct ordered_set_by_nameS { }; namespace boost { template <class ValueType> struct container_gen<ordered_set_by_nameS, ValueType> { typedef std::set<ValueType, order_by_name<ValueType> > type; }; } #else struct ordered_set_by_nameS { template <class T> struct bind_ { typedef std::multiset<T, order_by_name<T> > type; }; }; namespace boost { template <> struct container_selector<ordered_set_by_nameS> { typedef ordered_set_by_nameS type; }; } #endif

3 3

Inconsistent regexp matching when using quantifiers?
by Dean 13 May '03

13 May '03

Hi all, I'm using regexp from boost-1.28 and experience the following behavior. Consider the following (somewhat artificial) regexp pattern: a{1}b As I expected, that pattern is found in "ab" but not "aab". To my suprise however, the same pattern *is* found in "aaab", "aaaaab", and any other string consisting of an odd number of "a"s followed by a "b". It is not found in strings consisting of an even number of "a"s followed by a "b". This seems odd (no pun intended). I see the same sort of behavior with quantifiers other than "{1}" and where the quantified expression matches other single characters. (Oddly enough, the behavior changes when using a quantified expression that matches multiple characters. "(ab){1}c" is found in "abc", "ababc", "abababc", and any other string containing "abc".) FWIW, I first observed the behavior when trying to find social security numbers with the following pattern: \d{3}-\d{2}-\d{4} As expected, that pattern was found in "123-12-1234" but not in "1234- 12-1234". However it *was* found in "1234567-12-1234". Is this behavior by design or is it a bug? If it's a bug, has it been fixed in a subsequent boost release? Also what is the correct behavior? Should "a{1}b" be found in "aab" (albeit starting at the second character)? FWIW, it's easy enough for me to workaround the current behavior with a pattern like this: (^|[^a])a{1}b Thanks, --Dean

4 9

Compiling smart_ptr library
by l＠lo 08 May '03

08 May '03

Hi everybody! I'm trying to compile the smart_ptr library using the Jamfile in the test directory. Some of the test *.cpp files are compiled successfully, but some of them drops the following error: sunpro-Link-action ../../../libs/smart_ptr/test/bin/shared_ptr_alloc_ test.test/sunpro/debug/runtime-link-dynamic/shared_ptr_alloc_test Undefined first referenced symbol in file boost::__RTTI__1nFboostPchecked_deleter4nBZ___ /rest/proj/users/kercsoj/boost/boost_1_30_0/libs/s mart_ptr/test/bin/shared_ptr_alloc_test.test/sunpr o/debug/runtime-link-dynamic/ SunWS_cache/CC_obj__/_iANwWXwaw2nbnjBybyL.o [Hint: static member boost::__RTTI__1nFboostPchecked_deleter4nBZ___ must be defined in the program] It's seems, it is a common error, because it appears several times in the error output of the compiler. Has anybody any ideea, why this error message comes, and how could be corrected? Thanks, lalo

4 5

bjam and windows command line problem.
by Fernando Cacciola 07 May '03

07 May '03

Hi, I'm having a hard time trying to use bjam to build boost 1.29.0. I have Win98 in Spanish, and I use Borland C++ Builder 5.0 (its bcc32.exe corresponds to is BCC5.5.1) I have boost 1.29.0 installed at: "C:\Libraries\boost\Repository\boost" I've downloaded bjam for Microsoft Windows: >bjam -v Boost.Jam Version 3.1.0. OS=NT. Copyright 1993-2002 Christopher Seiwald and Perforce Software, Inc. Copyright 2001 David Turner. Copyright 2001-2002 David Abrahams. I've temporarily modified the root 'jamfile' so that it builds only the Test library. and I've done: subst d: "c:\archivos de programa\borland\cbuilder5" set BCCROOT=d: (you'll see why in a moment) If I try: c:\>cd C:\Libraries\boost\Repository\boost" C:\Libraries\boost\Repository\boost> bjam -sTOOLS=borland I receive messages of the form: borland-C++-action libs\test\build\bin\libpem.lib\borland\debug\runtime-link-static\threading-s ingle\execution_monitor.obj "d:\bin\bcc32" -j5 -q -c -P -w -Ve -Vx -a8 -v -Od -vi- -tWC -WM- -w-8001 -I"libs\test\build" -I"C:\Libraries\boost\Repository\boost" -I"d:\includ e" -o"libs\test\build\bin\libpem.lib\borland\debug\runtime-link-static\thre ading-single\execution_moni Error E2266: No file names given tor.obj" "libs\test\build\../src\execution_monitor.cpp" C:\WINDOWS\TEMP\JAMTMP00.BAT [2] Unknown command "TOR.OBJ" "LIBS\TEST\BUILD\.." "d:\bin\bcc32" -j5 -q -c -P -w -Ve -Vx -a8 -v -Od -vi- -tWC -WM- -w-8001 -I"libs\test\build" -I"C:\Libraries\boost\Repository\boost" -I"d:\includ e" -o"libs\test\build\bin\libpem.lib\borland\debug\runtime-link-static\thre ading-single\execution_monitor.obj" "libs\test\build\../src\execution_monitor.cpp" ...failed borland-C++-action libs\test\build\bin\libpem.lib\borland\debug\runtime-link-static\threading-s ingle\execution_monitor.obj... If I'm interpreting this output correctly, the problem is that the command line feed to bcc32 exceeds 256 characters. What can be done about this? Is there a way for 'borland-tools.jam' to put the bcc32 parameters in a tmp response file and use: bcc32.exe @tmprsp.rsp? Any help will be appreciated. TIA, -- Fernando Cacciola

4 22

BGL leaky?
by Martin Okrslar 06 May '03

06 May '03

Dear Boost users, is the BGL leaky? For example the code added below produces memory leaks (as seen from also added valgrind output). When I comment out the line marked by /* THIS LINE DOES IT */, there are no leaks. I am running BGL 1.29 and gcc 3.2. Any ideas for a work-around? cheers Martin #include <iostream> // for std::cout #include <utility> // for std::pair #include <algorithm> // for std::for_each #include <boost/graph/graph_traits.hpp> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/dijkstra_shortest_paths.hpp> using namespace boost; int main(int,char*[]) { // create a typedef for the Graph type typedef adjacency_list<vecS, vecS, bidirectionalS> Graph; // Make convenient labels for the vertices enum { A, B, C, D, E, N }; const int num_vertices = N; // writing out the edges in the graph typedef std::pair<int, int> Edge; Edge edge_array[] = { Edge(A,B), Edge(A,D), Edge(C,A), Edge(D,C), Edge(C,E), Edge(B,D), Edge(D,E), }; int num_edges = sizeof(edge_array)/sizeof(edge_array[0]); // declare a graph object Graph* g = new Graph(num_vertices); // add the edges to the graph object for (int i = 0; i < num_edges; ++i) add_edge(edge_array[i].first, edge_array[i].second, *g); /* THIS LINE DOES IT */ delete g; return 0; } okrslar@liebrich:~/molgen/domclust/src/test> valgrind --leak-check=yes --show-reachable=yes ./domclust ==23473== valgrind-1.0.4, a memory error detector for x86 GNU/Linux. ==23473== Copyright (C) 2000-2002, and GNU GPL'd, by Julian Seward. ==23473== Estimated CPU clock rate is 2412 MHz ==23473== For more details, rerun with: -v ==23473== ==23473== ==23473== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0) ==23473== malloc/free: in use at exit: 384 bytes in 1 blocks. ==23473== malloc/free: 3 allocs, 2 frees, 544 bytes allocated. ==23473== For counts of detected errors, rerun with: -v ==23473== searching for pointers to 1 not-freed blocks. ==23473== checked 6470036 bytes. ==23473== ==23473== definitely lost: 0 bytes in 0 blocks. ==23473== possibly lost: 0 bytes in 0 blocks. ==23473== still reachable: 384 bytes in 1 blocks. ==23473== ==23473== 384 bytes in 1 blocks are still reachable in loss record 1 of 1 ==23473== at 0x4004A3A8: __builtin_new (vg_clientfuncs.c:125) ==23473== by 0x4004A3F6: operator new(unsigned) (vg_clientfuncs.c:138) ==23473== by 0x4035E4F1: ??? (bits/stl_alloc.h:482) ==23473== by 0x4035E40D: ??? (bits/stl_alloc.h:532) ==23473== ==23473== LEAK SUMMARY: ==23473== definitely lost: 0 bytes in 0 blocks. ==23473== possibly lost: 0 bytes in 0 blocks. ==23473== still reachable: 384 bytes in 1 blocks. ==23473== -- ----------------------------------------------------------------------------- Martin Okrslar MPI for Molecular Genetics phone: ++ 49 + 30 / 8413-1166 Computational Molecular Biology Fax: ++ 49 + 30 / 8413-1152 Ihnestrasse 73 email: okrslar(a)molgen.mpg.de D-14195 Berlin URL: http://cmb.molgen.mpg.de -----------------------------------------------------------------------------

2 1

More Mac Compiler Woes - Boost 1.30.0 - Thread and type_traits
by Bobby Thomale 05 May '03

05 May '03

Hey guys, I am having some more problems getting Boost 1.30 to compile on the Mac. I am using CW Pro 8.3, compiling a TARGET_CARBON project. I am having trouble getting the MP implementation of threads to compile. It seems to be having trouble with function pointers where the syntax function<type> is used. I remember at one point it was mentioned that: function<type, type...> was deprecated and we were to use either the more universally supported: functionN<type,type...> or the cleaner: function<void myfuncSig()> instead if our compiler supported it. (Mine doesn't support that last one yet I don't think.) I am guessing that the oldest syntax was removed and no one has bothered to update the Mac carbon thread implementaton to compile again since then? That one seems pretty easy to fix - just update to a supported syntax? (I could do this, update to the functionN<type,type...> syntax since that is what my compiler supports. If someone wants me to make and submit the changes to them for inclusion I'd be happy to, if I knew who to send them to...) Or is there a better/easier workaround? Secondly, the Signals library is not compiling. It's actually not a Signals problem, I think it is a type_traits problem. I am getting a bunch of strange compile errors, in a boost::type_traits file called: is_base_and_derived.hpp It looks like there is a template near the top that is the first thing that fails. There is a define that goes into an alternate implementation for certain broken compilers. Perhaps CodeWarrior needs to be added to the list of compilers that compiles the "broken compiler" version of the stuff in this header? Thanks in advance for your help! -- Bobby --------------------------------------------------------------------- Bobby Thomale Senior Software Developer Inoveon Corporation http://www.inoveon.com/ ---------------------------------------------------------------------

5 6

RE: [Boost-Users] How To Build Boost 1.30 With Dev-CPP?
by bjorn.karlsson＠readsoft.com 04 May '03

04 May '03

> From: firingme [mailto:yg-boost-users@m.gmane.org] > E:\LY\boost_1_30_0>bjam sTOOS=mingw This should read sTOOLS, not sTOOS. Bjorn

5 5

const shared_ptr<>
by Hossein Haeri 02 May '03

02 May '03

Dear all, I'm a newbie in boost! I want to know what the following means: int a = 5; const boost::shared_ptr<int> sp(&a); Does it mean that the following line is an error? *sp = 7; If so, what are the guarantees? Is there any documented guarantee? (I couldn't find anything in the documentation. Excuseme if I'm careless!) Thanks, --Hossein __________________________________ Do you Yahoo!? The New Yahoo! Search - Faster. Easier. Bingo. http://search.yahoo.com

3 11

regexp match question
by Simon Brown 01 May '03

01 May '03

Hi All, Here's a question I can't answer. I'm using regexp to dig out strings from data received via Ham Radio, so the txt often is corrupted. What I want to match is two consecutive callsigns separated by whitespace, for example 'hb9drv hb9drv' but not 'hb9drv hb9drX'. I can define the match condition for a callsign, but can I use the match option {2,3} somehow to match the same callsign two or three times? Effectively match the same word two or three times? Help received with much thanks, Simon Brown, Casa Bergenia, CH-7031 Laax www.hb9drv.ch [Non-text portions of this message have been removed]

3 5

Licence terms
by Ben Hutchings 30 Apr '03

30 Apr '03

Where should I post to ask questions and make suggestions about licence terms for Boost libraries? I am specifically concerned about the large number of slightly different wordings of some of the licences, which I think would all have to be included in documentation of a program using the covered libraries.

1 0

RE: [Boost-Users] const shared_ptr<>
by Tom Matelich 30 Apr '03

30 Apr '03

> From: Hossein Haeri [mailto:powerprogman@yahoo.com] > Sent: Wednesday, April 30, 2003 8:29 AM > > I want to know what the following means: > > int a = 5; > const boost::shared_ptr<int> sp(&a); > I assume you know this, but this code is in error. When the last shared_ptr referencing &a goes away, it will try to delete a, which is illegal. There is a workaround defined at http://www.boost.org/libs/smart_ptr/sp_techniques.html#static , though. ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

1 0

ptime and time_duration
by Andre du Toit 30 Apr '03

30 Apr '03

Hi. Since time_duration uses <long> to represent each of the fields (H:M:S:FS), I should be able to add any hour offset to a ptime, as long as the resultant ptime is inside the date/time range representable by boost? With the date ranges of boost being 1400-Jan-1 00:00:00 -> 10000-Jan-1 00:00:00, the number of hours between these dates is (very roughly!) 7.534E7. This is less than the <long> limit on my system, which is +/- 2.147E9. That is good. So; doing the following: ptime t_max( date(10000,Jan,1) ); ptime t_min( date(1400,Jan,1) ); time_duration td = t_max - t_min; std::cout << to_simple_string( td ) << std::endl; should produce something like : 7.534E7:0:0 or if you want to write it out: 7533600:0:0 not? Instead; the output I get is : 224112:19:12. This corresponds to a timespan of roughly 25 years which means I just ran into some overflow/underflow problem somewhere. What gives? What time range is supported by time_duration if it does not use the <long> limits? I have difficulty deducing this information from the header files, the source for the lib is not too helpfull either! What am I missing here? Greetings, Andre.

2 1

running regression tests
by Vaclav Vesely 30 Apr '03

30 Apr '03

It's possible to execute all regression tests so it's possible to execute one test suite: bjam test or bjam filesystem But it's impossible to execute several (but not all) test suits at once: bjam config filesystem Only the last (filesystem) suit is executed. It's possible to improve testing.jam to allow it? Can someone give a bit of advice? Thanks and regards, w

4 5

Re: Re: running regression tests
by Dave Abrahams 30 Apr '03

30 Apr '03

Peter Dimov wrote: > David Abrahams wrote: >> BTW, Vaclav, the place to post about Boost.Build issues is >> jamboost(a)yahoogroups.com. >> >> http://groups.yahoo.com/group/jamboost/ > > ... and the place to announce mailing lists is > > http://www.boost.org/more/mailing_lists.htm Updated.

1 0

boost::spirit multiline comment parsing - how?
by Markus Werle 30 Apr '03

30 Apr '03

Hi! While playing around with the examples of the spirit library I feel like I missed some essential points. The far goal I have is to parse a file containing C/C++ comments and assignments of the form name = value; where I drop the comments and store the assignments in a map<string, string>. If You have done something similar, please share Your knowledge with the rest of us. Below is a modifed version of comments.cpp. Here I only try to output any comment found while dropping the other stuff. Why does the parser drop stuff after the 1st assignment? the output of the code is Parsed C-comment successfully! Matched (27) characters: "//test2 /* test1 */// test Any help appreciated. Also I would like not to gather all commenst in one string ... -------- #include "boost/spirit.hpp" #include <iterator> #include <string> #include <iostream> #include <cassert> #include <fstream> #include <map> #include <cassert> /////////////////////////////////////////////////////////////////////////////// // actor called after successfully matching a single character class actor_string { public: actor_string(std::string &rstr) : matched(rstr) { } void operator() (const char *pbegin, const char *pend) const { matched += std::string(pbegin, pend-pbegin); } private: std::string &matched; }; /////////////////////////////////////////////////////////////////////////////// // actor called after successfully matching a C++-comment void actor_cpp (const char *pfirst, const char *plast) { std::cerr << "Parsing C++-comment" << std::endl; std::cerr << "Matched (" << plast-pfirst << ") characters: "; char cbbuffer[128]; strncpy(cbbuffer, pfirst, plast-pfirst); cbbuffer[plast-pfirst] = '\0'; std::cerr << "\"" << cbbuffer << "\"" << std::endl; } /////////////////////////////////////////////////////////////////////////////// // main entry point int main () { using namespace boost::spirit; ///////////////////////////////////////////////////////////////////////////// char const* pCComment = "//test2\n/* test1 */\n\n// test\nalpha = 3.0;\n\n// beta is soo important\nbeta=5.6;\n\n/* \n\nSometimes comments are\ngood since\nbeta = 4.5;\nis a good choice, too\n\n*/"; typedef skipper_iteration_policy<> iter_policy_t; typedef scanner_policies<iter_policy_t> scanner_policies_t; typedef scanner<char const*, scanner_policies_t> phrase_scanner_t; rule<phrase_scanner_t> cpp_comment, ignored, all; cpp_comment = comment_p("/*", "*/") // rule for C-comments | comment_p("//") // rule for C++ comments ; std::string comment_c; all = ( *cpp_comment[actor_string(comment_c)] | *anychar_p ) ; ignored = (space_p | chlit<char>('\n')); parse_info<> result; result = parse (pCComment, all, space_p); if (result.hit) { std::cerr << "Parsed C-comment successfully!" << std::endl; std::cerr << "Matched (" << comment_c.size() << ") characters: "; std::cerr << "\"" << comment_c << "\"" << std::endl; } else { std::cerr << "Failed to parse C/C++-comment!" << std::endl; } std::cerr << std::endl; } Markus -- Compile time analytic differentiation? Yes, at http://daixtrose.sourceforge.net/

2 1

bjam, targets from another projects
by Vaclav Vesely 30 Apr '03

30 Apr '03

Hi, It's possible to specify dependency on a target from another project tree? I want to write to my project's "Jamfile" something like this: --- Jamfile --- exe test : test.cpp <dll>$(BOOST_ROOT)/libs/threads/build/boost_thread : <include>$(BOOST_ROOT) ; --- Thanks and regards, w

2 1

bjam - several projects and ALL_LOCATE_TARGET
by Vaclav Vesely 30 Apr '03

30 Apr '03

Hi, I would like to ask for help with using bjam. I want to have more project directory trees, for example: C:\src\boost C:\src\my_projects\first C:\src\my_projects\first and I want to compile binaries to separate directories: C:\build\boost C:\build\my_projects\first C:\build\my_projects\second So I have to set different values of the ALL_LOCATE_TARGET environment variable for each project. It's unconfortable to set that value each time I want to inovoke the bjam, so I want to set it in the Jamfile of each project. I mean somethin like this: --- C:\src\my_first_project\Jamfile --- project-root ; ALL_LOCATE_TARGET = C:\build\my_projects\first --- Of course, this doesn't work. Can someone propose the solustion? BTW it would be handy to use the "project-root" directive for this reason: --- C:\src\my_first_project\Jamfile --- project-root my_projects/first ; --- The specified path is ment as relative to the ALL_LOCATE_TARGET. Thanks and regards, w

2 1

tokenizer: finding the last element. is it possible?
by Vincent LaBella 29 Apr '03

29 Apr '03

Hello, I'm using boost::tokenizer and want to goto the last token in the sequence, so I tried Tok::iterator last = tok.end(); last--; // error decrement not defined last = tok.begin(); advance(last,tok.size()-1); // error size() not defined. Is it possible to add decrement or a back() reference to tokenizer? Thanks Vince

3 4

scoped_ptr(auto_ptr) constructor
by Daniel Franklin 29 Apr '03

29 Apr '03

Is the constructor for scoped_ptr that takes an auto_ptr argument part of the proposed standard? The boost documentation makes no mention of it, yet it exists in the implementation as: explicit scoped_ptr(std::auto_ptr<T> p): ptr(p.release()) // never throws { } Personally I think it should be included in the proposed standard if it is not. I can't see any reason why it shouldn't be, as it doesn't seem to break the fundamental concepts behind auto_ptrs (ownership transferal) or scoped_ptrs (sole ownership). Ownership of the object is transfered from the auto_ptr to the scoped_ptr, and thereafter retained solely by that scoped_ptr. Regards, Daniel Franklin

1 0

[PP] Getting the size of a tuple
by Sam Partington 29 Apr '03

29 Apr '03

Hello, I've just started using the preprocessor library, for a fairly simple task. Essentially I have the have a macro called DECLARE_TESTS, which I use like the following DECLARE_TESTS(2, (Test1, Test2) ) which expands into the following void Test1(); void Test2(); suite() { ADDTEST(Test1); ADDTEST(Test2); } Now the problem: it is inconvenient to have to specify the size of the the tuple in the first macro. Is there any way to deduce the size? The macros I'm using are below. Thanks in advance for any help (positive or otherwise) Sam PS Before you suggest using Boost.Test, we have looked at it, and feel that it is not quite what we're looking for, and I don't want to get into that discussion! :-) #define DECLARE_TESTS(test_count, test_list) \ BOOST_PP_LIST_FOR_EACH(UNITTESTS_DECLARE_TEST, _, BOOST_PP_TUPLE_TO_LIST(test_count, test_list)) \ public: \ ClassToTest() \ { \ BOOST_PP_LIST_FOR_EACH(UNITTESTS_ADD_TEST, _, BOOST_PP_TUPLE_TO_LIST(test_count, test_list)) \ } // implementation macros #define UNITTESTS_DECLARE_TEST(R, _, name) void name(); #define UNITTESTS_ADD_TEST(R, _, name) ADDTEST(name)

1 0

uBLAS compile fail on sparse matrix debug code
by Zhang Le 29 Apr '03

29 Apr '03

Hi all, I just write a simple program to test the efficiency of sparse matrix operation in uBLAS and get the following error message on a FreeBSD with gcc 2.95 : !gmake 2>&1| tee /tmp/v446688/2g++ -c test.cpp -I. -I/usr/local/include -Wall -g -O2 -march=i686 -pipe -ftemplate-depth-30 -g3 -O0 -o test.o g++ test.o -o a.out test.o: In function `pair<unsigned int, unsigned int> * __copy_backward<pair<unsigned int, unsigned int> *, pair<unsigned int, unsigned int> *, int>(pair<unsigned int, unsigned int> *, pair<unsigned int, unsigned int> *, pair<unsigned int, unsigned int> *, random_access_iterator_tag, int *)': /usr/include/g++/stl_algobase.h(.gnu.linkonce.t.norm_inf__Q45boost7numeric5ublast11type_traits1ZUiRCUi+0x15): undefined reference to `boost::numeric::ublas::type_traits<unsigned int>::imag(unsigned int const &)' /usr/include/g++/stl_algobase.h(.gnu.linkonce.t.norm_inf__Q45boost7numeric5ublast11type_traits1ZUiRCUi+0x24): undefined reference to `boost::numeric::ublas::type_traits<unsigned int>::abs(unsigned int const &)' /usr/include/g++/stl_algobase.h(.gnu.linkonce.t.norm_inf__Q45boost7numeric5ublast11type_traits1ZUiRCUi+0x3a): undefined reference to `boost::numeric::ublas::type_traits<unsigned int>::real(unsigned int const &)' /usr/include/g++/stl_algobase.h(.gnu.linkonce.t.norm_inf__Q45boost7numeric5ublast11type_traits1ZUiRCUi+0x49): undefined reference to `boost::numeric::ublas::type_traits<unsigned int>::abs(unsigned int const &)' /usr/include/g++/stl_algobase.h(.gnu.linkonce.t.equals__H2ZQ45boost7numeric5ublast6vector2ZUiZQ45boost7numeric5ublast15unbounded_array1ZUiZQ45boost7numeric5ublast6vector2ZUiZQ45boost7numeric5ublast15unbounded_array1ZUi_Q35boost7numeric5ublasRCQ45boost7numeric5ublast17vector_expression1ZX01RCQ45boost7numeric5ublast17vector_expression1ZX11_b+0x2e): undefined reference to `boost::numeric::ublas::type_traits<unsigned int>::sqrt(unsigned int const &)' If I comment out the second prod() or use -DNDEBUG everything is ok. So I suspect the error is caused by debug code in sparse matrix prod operation. here is the main code block: int main(int argc,char* argv[]) { try { namespace ublas = boost::numeric::ublas; using namespace ublas; size_t d = 7000; ublas::vector<unsigned> v(d); ublas::matrix<unsigned> m (d,d); ublas::sparse_matrix<unsigned> ms (d,d); for (size_t i = 0;i < 10000;++i) { size_t r = (size_t)(drand48() * d); size_t c = (size_t)(drand48() * d); ++m(r,c); ++ms(r,c); ++v(r); ++v(c); } { progress_timer t; std::cout << "matrix * vector" << endl; ublas::prod(m,v); } { progress_timer t; std::cout << "sparse matrix * vector" << endl; ublas::prod(ms,v); } }

2 1

Trouble building 1.30.0 on Linux
by Garen Parham 29 Apr '03

29 Apr '03

When I try to build boost with this method: ./build.sh gcc || die "couldn't build bjam" cd $srcdir ./tools/build/jam_src/bin.linux${ARCH}/bjam -j2 \ -sBOOST_ROOT=${S} \ -sPYTHON_ROOT=/usr \ -sPYTHON_VERSION=2.2 \ -sTOOLS=gcc This happens: ...failed updating 6 targets... ...updated 539 targets... On the other hand, if I restart the build it seems to..magically work: ### ### Using 'gcc' toolset. ### rm -rf bootstrap.gcc mkdir bootstrap.gcc gcc -o bootstrap.gcc/jam0 command.c compile.c execnt.c execunix.c execvms.c expand.c filent.c fileos2.c fileunix.c filevms.c glob.c hash.c hdrmacro.c headers.c jam.c jambase.c jamgram.c lists.c make.c make1.c newstr.c option.c parse.c pathunix.c pathvms.c regexp.c rules.c scan.c search.c subst.c timestamp.c variable.c modules.c strings.c filesys.c builtins.c pwd.c ./bootstrap.gcc/jam0 -f build.jam --toolset=gcc --toolset-root= ...found 44 targets... ...found 3407 targets... ...updating 6 targets... FileClone libs/python/build/bin-stage/libboost_python_debug.so.1.30.0 FileClone libs/python/build/bin-stage/libboost_python.so.1.30.0 FileClone libs/regex/build/bin-stage/libboost_regex_debug.so.1.30.0 FileClone libs/regex/build/bin-stage/libboost_regex.so.1.30.0 FileClone libs/thread/build/bin-stage/libboost_threadd.so.1.30.0 FileClone libs/thread/build/bin-stage/libboost_thread.so.1.30.0 ...updated 6 targets... That didn't happen with 1.29. Is there a cleaner way not documented to get it to build nicely? I had less luck trying to build with icc.

1 0

bug in testing.jam
by Vaclav Vesely 29 Apr '03

29 Apr '03

There is missing enclosing parenthesis. Regards, w Index: tools/build/testing.jam =================================================================== RCS file: /cvsroot/boost/boost/tools/build/testing.jam,v retrieving revision 1.16 diff -u -r1.16 testing.jam --- tools/build/testing.jam 3 Mar 2003 19:59:45 -0000 1.16 +++ tools/build/testing.jam 26 Apr 2003 23:09:40 -0000 @@ -99,7 +99,7 @@ if $(files) { # make relative to the project root instead of "." - local file = $(files[1]:R=$(SUBDIR_TOKENS:J=/) ; + local file = $(files[1]:R=$(SUBDIR_TOKENS:J=/)) ; # try to undo absolute paths source-files += [ relative-path $(file) ] ; begin 666 testing.jam.patch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`` ` end

2 1