- Boost-users - lists.boost.org

[boost-users][proto] Using proto with nvcc
by Manjunath Kudlur 27 Apr '10

27 Apr '10

I noticed that fix for ticket #3919 has been checked-in. Thanks JoelF. I checked out boost trunk, and am trying to compile the following simple program using nvcc. #include <boost/proto/proto.hpp> namespace proto=boost::proto; int main() { return 0; } I was getting "ISO C++ forbids declaration of ‘expr_param’ with no type" error in couple of places in proto/traits.hpp and proto/transform/pass_through.hpp. The code used to look like typename pass_through_impl::expr_param e I added a typedef as follows transform_impl<Expr, State, Data> base_type; and changed the code to typename base_type::expr_param e Now the compiler error goes away. I have 2 questions : 1) Is what I am doing the right fix for the problem, or is the nvcc compiler really broken when handling this? 2) If my fix is OK, how do I make it permanent? The other error I am seeing is probably not proto's problem, but I will mention it here anyway : boost/boost-trunk/boost/mpl/size_t_fwd.hpp:23: error: expected unqualified-id before ‘[’ token boost/boost-trunk/boost/mpl/size_t_fwd.hpp:26: error: ‘mpl_::size_t’ has not been declared boost/boost-trunk/boost/mpl/aux_/integral_wrapper.hpp:42: error: expected unqualified-id before ‘[’ token Any ideas on this? Thanks, Manjunath http://nonchalantlytyped.net/blog/musings/

3 7

Build errors with boost 1.42.0 under MinGW
by Michael Chisholm 27 Apr '10

27 Apr '10

I am having problems building boost 1.42.0 using MinGW. The bootstrap.sh script only works if I specify 'mingw' as the toolset. The resulting bjam tool doesn't work though. It doesn't seem to support mingw as a toolset, but I am unable to force it to use something else. I am using WinXP SP3, MinGW GCC 3.4.5. Just running ./bootstrap.sh yields: $ ./bootstrap.sh Building Boost.Jam with toolset gcc... Failed to build Boost.Jam Consult 'bootstrap.log' for more details The log contents are at the bottom of this email. If I instead specify mingw as the toolset, it works: $ ./bootstrap.sh --with-toolset=mingw Building Boost.Jam with toolset mingw... tools/jam/src/bin.ntx86/bjam Unicode/ICU support for Boost.Regex?... not found. Backing up existing Boost.Build configuration in project-config.jam.3 Generating Boost.Build configuration in project-config.jam... Bootstrapping is done. To build, run: ./bjam etc... Now if I run ./bjam.exe, I get the following error: $ ./bjam.exe mingw.jam: No such file or directory c:/Programming/C/boost_1_42_0/tools/build/v2/build\toolset.jam:38: in toolset.using rule mingw.init unknown in module toolset. c:/Programming/C/boost_1_42_0/tools/build/v2/build\project.jam:884: in using project-config.jam:12: in modules.load c:/Programming/C/boost_1_42_0/tools/build/v2\build-system.jam:241: in load-config c:/Programming/C/boost_1_42_0/tools/build/v2\build-system.jam:407: in load-configuration-files c:/Programming/C/boost_1_42_0/tools/build/v2\build-system.jam:538: in load c:\Programming\C\boost_1_42_0\tools\build\v2/kernel\modules.jam:283: in import c:\Programming\C\boost_1_42_0\tools\build\v2\kernel\bootstrap.jam:142: in boost-build c:\Programming\C\boost_1_42_0\boost-build.jam:17: in module scope If I try to specify gcc as the toolset via: $ ./bjam.exe toolset=gcc I get the exact same error. It's still trying to find 'mingw.jam'. The "Getting Started" documentation, as far as providing troubleshooting help, seems to assume the bjam command will work, if provided the proper toolset. So I'm stuck now. Has anyone out there got the latest boost libs built under MinGW? Andy Failed bootstrap.log contents: ### ### Using 'gcc' toolset. ### rm -rf bootstrap mkdir bootstrap gcc -o bootstrap/jam0 command.c compile.c debug.c expand.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 output.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 class.c native.c w32_getreg.c modules/set.c modules/path.c modules/regex.c modules/property-set.c modules/sequence.c modules/order.c execunix.c fileunix.c execunix.c:17:26: sys/resource.h: No such file or directory execunix.c:18:23: sys/times.h: No such file or directory execunix.c:19:22: sys/wait.h: No such file or directory execunix.c: In function `exec_cmd': execunix.c:193: error: `F_SETFL' undeclared (first use in this function) execunix.c:193: error: (Each undeclared identifier is reported only once execunix.c:193: error: for each function it appears in.) execunix.c:193: error: `O_NONBLOCK' undeclared (first use in this function) execunix.c:212: error: storage size of 'buf' isn't known execunix.c:216: error: `_SC_CLK_TCK' undeclared (first use in this function) execunix.c:243: error: storage size of 'r_limit' isn't known execunix.c:246: error: `RLIMIT_CPU' undeclared (first use in this function) execunix.c:249: warning: passing arg 2 of `execvp' from incompatible pointer type execunix.c: At top level: execunix.c:385: error: syntax error before "fd_set" execunix.c: In function `populate_file_descriptors': execunix.c:388: error: storage size of 'buf' isn't known execunix.c:393: error: `fds' undeclared (first use in this function) execunix.c:418: error: `SIGKILL' undeclared (first use in this function) execunix.c:423: error: `fmax' undeclared (first use in this function) execunix.c: In function `exec_wait': execunix.c:441: error: `fd_set' undeclared (first use in this function) execunix.c:441: error: syntax error before "fds" execunix.c:442: error: storage size of 'new_time' isn't known execunix.c:453: error: `fds' undeclared (first use in this function) execunix.c:459: error: invalid use of undefined type `struct timeval' execunix.c:460: error: invalid use of undefined type `struct timeval' execunix.c:519: error: invalid use of undefined type `struct tms' execunix.c:520: error: invalid use of undefined type `struct tms' execunix.c:524: error: `old_time' has an incomplete type execunix.c: At top level: execunix.c:69: error: storage size of `tv' isn't known execunix.c:73: error: storage size of `old_time' isn't known fileunix.c:97:17: ar.h: No such file or directory fileunix.c: In function `file_archscan': fileunix.c:273: error: storage size of 'ar_hdr' isn't known fileunix.c:282: error: `SARMAG' undeclared (first use in this function) fileunix.c:282: error: (Each undeclared identifier is reported only once fileunix.c:282: error: for each function it appears in.) fileunix.c:283: error: `ARMAG' undeclared (first use in this function) fileunix.c:294: error: invalid application of `sizeof' to incomplete type `ar_hdr' fileunix.c:294: error: invalid application of `sizeof' to incomplete type `ar_hdr' fileunix.c:295: error: `ARFMAG' undeclared (first use in this function) fileunix.c:326: error: invalid application of `sizeof' to incomplete type `ar_hdr' fileunix.c:356: error: invalid application of `sizeof' to incomplete type `ar_hdr'

1 0

[GSoC] Student Proposals
by Andrew Sutton 27 Apr '10

27 Apr '10

Hi All, I'd like to start by thanking all of the students who are interested in working on the Boost C++ Libraries and submitted proposals. Unfortunately, we can't fund every proposal we received (about 85), and the competition for available funding slots (only 10) is very competitive. We carefully reviewed every proposal based on the merits and feasibility of the work described. If your proposal was not accepted, that does not mean that it was without merit. We encourage all students to continue participating with the Boost community as users and developers, regardless of the results of this program, and look forward to seeing new proposals next year. If you are interested in feedback about your proposal's review and how you can improve the competitiveness of your submissions, feel free to contact me directly. Having said that, congratulations to the students whose projects were accepted. I will be sending out instructions soon on getting started and up to speed with the Boost development infrastructure, Subversion, mailing lists, etc. I'm will be posting the instructions to the Boost developers mailing list (not users!). If you are a student whose project was accepted, you need to subscribe to the Boost developer's mailing list [1]. I'd also like to thank all of the mentors for helping with the ranking and assignment of student projects. Thanks again and best regards, Andrew Sutton andrew.n.sutton(a)gmail.com [1] http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

Multi-array non-preserving resize?
by lindleyf＠cox.net 27 Apr '10

27 Apr '10

I'd like to use a multi_array<double, 2> inside a loop, but the dimensions may be different each time. Now, I could create a new one on each iteration, but I'd prefer to avoid the extra allocations/frees if I can. I also could create it outside the loop and use resize(), but that promises to preserve elements if possible, and that's extra copying I don't need (presumably; I'm assuming here that an underlying 1D memory layout is used, which means that preserving elements after the first row would require moving things around). Is there any simple way to ask a multi_array to resize itself without preserving elements?

2 1

[Random] Good Truly-Random Seed
by Hossein Haeri 26 Apr '10

26 Apr '10

Dear all, May I please know what do you guys use as you "good truly-random seed"? TIA, --Hossein

12 30

enable-if overloading out the wazoo
by John Dlugosz 26 Apr '10

26 Apr '10

I tried something like this: First, the general form in one header known by everything: template <typename T1, typename T2> dim_s foo (const T1& left, const T2& right); Then, in a header that gets more concrete, provide its own form that is meant to be used when objects of those types are involved: template<typename T1, typename T2> typename enable_if_c< is_base_of<genericbase,T1>::value && is_base_of<genericbase,T2>::value, dim_s>::type foo (const T1& left, const T2& right); Here, the template defined in this header derives from genericbase, so I have a hook to say "all types derived from any specialization of this template". Now, this actually causes an ambiguity. The latter is not a specialization but a peer overloading, and when both are present, both are templates that form exact matches and template arguments at the least restrictive way to refer to the type. This function is already internal plumbing, so I don't mind it being a little messy to use. What I'm really doing is multimethod dispatch: various concrete classes, generic versions for making such, and special combinations all have their own efficient logic, with a fallback (slow) that always works. And I want all the plumbing to disappear at compile time so it efficiently calls the efficient function. So if I'm really guilty of the "X for Y" problem, that is what I'm really trying to do, and if there are any suggestions on doing that better in general, please let'em rip. As for the road I'm on now, the simplest thing is to somehow give each form a priority. I did that by adding a 3rd parameter. The general form declares a void*, and the specialized form declares an int. The caller always gives a zero. I'm wondering what is the longest chain of different priorities I can use with this method, using only compiler primitive types that the optimizer should be able to toss when the function is inlined. I know I could use dummy classes with any number of derivations, but I'm thinking I can get several levels just using primitive types. Any suggestions? Thanks in advance, --John (watch the footer) TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

4 5

[boost] [asio] how to handle "boost::asio::placeholders::error" in "boost::function"?
by 曹亚文 26 Apr '10

26 Apr '10

Hi all, I met a problem which can't be solved. Can any kind people help me? The code looks like the below: class transfor_data { public: void one_fun( const boost::system::error_code& e, connection conn ) { ... } void another_fun() { { //compile error because of the placeholder boost::function<void(void)> fun = boost::bind(&transfor_file_push::write_file, this, boost::asio::placeholders::error, conn); } { //fine without placeholder boost::system::error_code e; boost::function<void(void)> fun = boost::bind(&transfor_file_push::write_file, this, e, conn); } } }

2 1

Boost Interval Arithmetic Library: msvc_rounding_control.hpp broken in 64 bit msvc?
by Iman Brouwer 26 Apr '10

26 Apr '10

Hello, When using the boost interval library we get a run-time assertion failure in debug with 64 bit Visual Studio 2005 and up. In file msvc_rounding_control.hpp, the function set_rounding_mode(..) calls _control87(..) which in debug will cause an assert to fail. According to microsoft documentation: "On the x64 architecture, changing the floating-point precision is not supported. If the precision control mask is used on that platform, the invalid parameter handler is invoked, as described in Parameter Validation." Run-time library reference: http://msdn.microsoft.com/en-us/library/e9b52ceh%28VS.80%29.aspx Visual Studio 2010 documentation: http://msdn.microsoft.com/en-us/library/c9676k6h.aspx In release the invalid parameter handler is not invoked so the code runs without complaining, but the rounding-mode is of course not changed. Did anybody else run into this problem? Iman

1 0

[filesystem] remove_filename( ) - is this an error?
by Robert Jones 26 Apr '10

26 Apr '10

Hi Guys This bit of code... #include <iostream> #include <boost/filesystem.hpp> int main( ) { const char * bits[] = { "first", "second", "third", "fourth", 0 }; boost::filesystem::path path; for ( const char * * bit = bits; * bit; ++ bit ) { path /= * bit; } while ( path.has_parent_path( ) ) { std::cout << path << std::endl; path.remove_filename( ); } } produces first/second/third/fourth first/second/third/ first/second/third first/second/ first/second first/ which surprised me, as the documentation says basic_path& remove_filename(); *Effects:* If has_parent_path() then remove the last *filename* from the stored path. If that leaves the stored path with one or more trailing *slash * elements not representing *root-directory*, remove them. Seems to me that remove_filename() does NOT remove trailing slashes as stated. - Rob.

1 0

Boost.Function __stdcall, __fastcall, etc
by Brian Rowlett 26 Apr '10

26 Apr '10

I am writing a DLL that gets injected into another process, and creates pointers to the target processes functions. The target process uses all 3 calling conventions, __stdcall, __fastcall and __cdecl. I have created a class whos constructor gets the address of the function, and passes it to the base class boost::function: > template <typename _Signature, enum LibraryId _LibraryId, int _Offset> > class FunctionPointer: protected OffsetPointer, public > boost::function<_Signature> > { > public: > FunctionPointer(void) > : boost::function<_Signature>( OffsetPointer::_getOffset(_LibraryId, > _Offset ) > { return; } > }; > I instantiate this with: > FunctionPointer<return_type__stdcall( argument_types ), LibraryId, Offset> > Function; FunctionPointer<return_type __cdecl( argument_types ), LibraryId, Offset> > Function; Whatever calling convention Visual Studio is setup to use (__cdecl by default), works just fine. But trying to instantiate it with another calling convention gives the following error: > Error 1 error C2504: 'boost::function<Signature>' : base class undefined It seems weird that it would work for some calling conventions, and not others.. Any ideas? -- + Brian J. Rowlett

2 2