March 2009 - Boost-users - lists.boost.org

Re: [Boost-users] Formal Review: Boost.RangeEx
by Alan M. Carroll 03 Mar '09

03 Mar '09

1 0

RE [Q] any successes in building 1.38+GCC-4.3.2+cygwin
by Vial, Florent 03 Mar '09

03 Mar '09

>>> ./boost/config/requires_threads.hpp:29:4: error: #error "Threading >>> support unavaliable: it has been explicitly disabled with BOOST_DISABLE_THREADS" >> This may be an issue with 4.3.2 as it works fine with 3.4.4. >AFAIR, there is no pthreads in gcc-4 on cygwin. I agree with that. I tried it too for a couple of weeks, but the current gcc 4.3.2 version installed in cygwin was built without pthread. Try a "$gcc-4.exe --v". On my machine it returns: Using built-in specs. Target: i686-pc-cygwin Configured with: /gnu/gcc/release/gcc4-4.3.2-1/src/gcc-4.3.2/configure --srcdir=/gnu/gcc/release/gcc4-4.3.2-1/src/gcc-4.3.2 --prefix=/usr --exec-prefix=/us --bindir=/usr/bin --sbindir=/usr/sbin --libexecdir=/usr/sbin --localstatedir=/var --sysconfdir=/etc --datadir=/usr/share --infodir=/usr/share/info --mandi =/usr/share/man -v --with-gmp=/usr --with-mpfr=/usr --enable-bootstrap --enable-version-specific-runtime-libs --with-slibdir=/usr/bin --libexecdir=/usr/lib --enable-static --enable-shared --enable-shared-libgcc --enable-__cxa_atexit --with-gnu-ld --with-gnu-as --with-dwarf2 --disable-sjlj-exceptions --enable-l nguages=ada,c,c++,fortran,java,objc,obj-c++ --disable-symvers --enable-libjava --program-suffix=-4 Thread model: single gcc version 4.3.2 20080827 (alpha-testing) 1 (GCC) The line saying "Thread model: single" says it all... HTH, Florent Arnold & Richter Cine Technik GmbH & Co. Betriebs KG Sitz: Munchen - Registergericht: Amtsgericht Munchen - Handelsregisternummer: HRA 57918 Personlich haftender Gesellschafter: Arnold & Richter Cine Technik GmbH Sitz: Munchen - Registergericht: Amtsgericht Munchen - Handelsregisternummer: HRB 54477 Geschaftsfuhrer: Prof. Dr. Dejan Ilic; Franz Kraus; Jens Schafer

1 0

problems with 32-bit Boost libs on Linux 64-bit with multilib gcc
by Damien Hocking 03 Mar '09

03 Mar '09

Hi all, I'm having some problems with building a 32-bit Boost on a 64-bit Linux. I'm building 1.38 and I need 64 and 32-bit version of the libraries. The compiler is gcc 4.3.2 with the multilibs installed (Ubuntu 8.10). My builds go through successfully, but when I link my 32-bit version of my application to what should be 32-bit Boost libraries, I get the "skipping incompatible" warning from the linker, which means that the address model is wrong. For example: /usr/bin/ld: skipping incompatible 3rdparty/boost/linux/x86/debug/libboost_serialization-gcc43-mt-d-1_38.so The 64-bit libraries are fine and the application links and runs correctly. I noticed in gcc.jam there's a comment about not being able to build a clean 32-bit binary with a 64-bit compiler. This is my bjam command for a 32-bit Boost build: bjam --build-dir=./build32 address-model=32 --build-type=complete --stagedir=./stage32 stage Do I need a -m32 in a config file or something extra on that command line? Damien

1 1

Timing of BoostPro 1.38 Installer?
by Scott Meyers 03 Mar '09

03 Mar '09

Does anybody happen to know when an installer will be available for Boost 1.38? Thanks, Scott

5 5

[Q] any successes in building 1.38+GCC-4.3.2+cygwin
by V S P 03 Mar '09

03 Mar '09

I was trying to test my code on gcc and since do not have access to linux, I used cygwin gcc v 4.3.2 with boost 1.38. I first built bjam (it went Ok). copied it to /bin then created symlink g++ to gcc-4 (cygwin has gcc 4.3.2 as gcc-4) and then with this command bjam toolset=gcc variant=debug release link=shared threading=multi runtime-link=shared stage started the build thread library built (although I am not sure if pthreads are supported on cygwin, may be it was just a dummy library). Many link errors. So wanted to ask if anybody has tried - or if you guys would recommend to try the same with gcc-3.4 thank you in advance -- V S P toreason(a)fastmail.fm -- http://www.fastmail.fm - One of many happy users: http://www.fastmail.fm/docs/quotes.html

4 5

Taucs installation
by Chaman Singh Verma 03 Mar '09

03 Mar '09

Hello, I downloaded "Taucs" from the CGAL site, but it seems that installation is not clean. Can someone tell what I must do ? Thanks. csv csv610@blackhole:~/Downloads/taucs_full$ export OSTYPE=linux csv610@blackhole:~/Downloads/taucs_full$ ./configure ./configure: 61: [[: not found cc -o configurator/configurator configurator/taucs_config.c rm -rf lib/linux/* rm -rf obj/linux/* rm -rf bin/linux/* rm -rf lib/linux/*.* rm -rf obj/linux/*.* rm -rf bin/linux/*.* rm -rf build/linux/taucs_config_tests.h rmdir obj/linux/ rmdir: failed to remove `obj/linux/': No such file or directory make: [clean] Error 1 (ignored) rmdir bin/linux/ rmdir: failed to remove `bin/linux/': No such file or directory make: [clean] Error 1 (ignored) rmdir lib/linux/ rmdir: failed to remove `lib/linux/': No such file or directory make: [clean] Error 1 (ignored) csv610@blackhole:~/Downloads/taucs_full$

2 2

Visual C++ 9.0 + WinCE
by Hernan Di Pietro 03 Mar '09

03 Mar '09

I'm using Boost 1.37 and I've found that using the headers to develop WinCE applications with WinCE causes Boost header config/compiler/visualc.hpp to emit the EVC++ version not recognized. I've patched the *visualc.hpp* header to recognize the WinCE/VC9 configuration as follows: (line 150): # if defined(UNDER_CE) # if _MSC_VER < 1200 // Note: these are so far off, they are not really supported # elif _MSC_VER < 1300 // eVC++ 4 comes with 1200-1202 # define BOOST_COMPILER_VERSION evc4.0 # elif _MSC_VER == 1400 # define BOOST_COMPILER_VERSION evc8 *# elif _MSC_VER == 1500 # define BOOST_COMPILER_VERSION evc9* # else # if defined(BOOST_ASSERT_CONFIG) # error "Unknown EVC++ compiler version - please run the configure tests and report the results" # else # pragma message("Unknown EVC++ compiler version - please run the configure tests and report the results") # endif # endif Thanks, I dont if there's any workaround already available. Hernán.

2 1

[Function] Small function optimization
by Robert Jones 03 Mar '09

03 Mar '09

A while back I wrote some code which made heavy use of small boost::functions, using Boost 1.33.1. The performance hit was pretty grim, and in the end I had to take thme out and use non generic hand rolled solutions. However, we have recently upgraded to Boost 1.37, and I notive from all the Boost version issue docs that a small function optimization was added in n Boost 1.35. Can anyone tell me in general terms what was done, and possibly if I'm likely to find the performance significantly better now. Many Thanks, Rob.

1 0

can't debug into BOOST_FOREACH?
by Stephan Menzel 03 Mar '09

03 Mar '09

Hi there, I just noticed there seems to be no way of debugging into a BOOST_FOREACH loop... Is that true? I use g++ 4.3 / gdb 6.8 here and neither would the debugger stop any any breakpoint inside such loops nor can I step into them from outside. Is this supposed to be like that? puzzled... Stephan

5 7

New and updated serialization performance results
by Brian Wood 03 Mar '09

03 Mar '09

Recently I updated this comparsion, http://webEbenezer.net/comparison.html, of Ebenezer Enterprises and the Boost Serialization library. Here's a link to the previous version of that file, http://webEbenezer.net/compprev.html. The previous tests used Boost 1.34.1 and the updated versions use Boost 1.38. I didn't detect any improvement performance- wise in the Boost Serialization library in this round of testing. In addition to the updated results, I've added some new tests that test loading/receiving data. Although Boost Serialization comes closer to Ebenezer times when loading/receiving data than it does when saving/sending data, there still is a gap between the two. That gap is greatest when (multi)set, (multi)map, or rb_tree are used. This is probably because Boost Serialization doesn't use the insert with a hint function available with these containers. I believe that multi_index uses the hinted verison of insert in it's marshalling code, so in my opinion it makes sense to consider this in Boost Serialization also. Additionally, we've added support for flex_string, http://erdani.org/code/main.html, to the site. On that page it says Harmut Kaizer reported that dropping flex_string into the Wave Boost Preprocessor improved it's speed by 5-10%, depending on input. I've done a few performance tests comparing the two strings and also found flex_string to be more efficient than std::string. Brian Wood Ebenezer Enterprises www.webEbenezer.net "Then Samuel took a rock and set it up between Mizpah and Shen. He named it Ebenezer [Rock of Help] and said, 'Until now the LORD has helped us.'"

2 1