- Boost - lists.boost.org

Template Instantiation Profiler with Callgrind Output (KCacheGrind)
by Mikael Persson 29 Jan '15

29 Jan '15

Hi boost developers and users, I just wanted to announce a recent development in my little pet-project, which is Templight: a template instantiation profiling tool based on Clang. I think that many of you would be interested in experimenting with it because many boost libraries are great examples of template-heavy libraries and many maintainers, developers and users might be interested in gathering data and diagnosing the real compilation costs (in time and memory) of various components of Boost. Templight is a Clang-based profiler that can act as a drop-in replacement (e.g., in cmake) for the clang compiler, with the addition of generating a complete trace of the template instantiation history with associated time and memory (optional) costs. This allows you to see not only how deeply recursive certain template instantiations are, but also identify which instantiations the compiler spends the most time or memory on. It is basically to meta-programming what a run-time profiler is to ordinary code. The templight profiler itself has reached a fairly stable state. It works at least under Linux and Windows (through "templight-cl.exe", which is a drop-in replacement for "clang-cl.exe", which is compatible with MSVC "cl.exe"), but probably works wherever clang works. Which means that you are welcome to try it. I have also used it successfully to build a few cmake-based projects (by setting CC and CXX environment variables to point to templight). Templight is currently also used in the back-end of the "metashell" project as well (an interactive shell interpreter and debugger for C++ template meta-programs). You can build the templight profiler against the clang source (with a patch + some added code), as it instructs on the github readme: https://github.com/mikael-s-persson/templight But most recently, I have created a separate set of tools to convert templight trace files to other formats. The most awesome of these outputs is the "callgrind" output, which produces a "meta-"call-graph, that is similar to (and in the same format as) call-graphs generated by run-time profiling tools like callgrind, but the profiling data is, of course, related to the compilation costs (time / memory). That conversion tool is available here: https://github.com/mikael-s-persson/templight-tools With that, you can open the converted templight trace files in the visualization tool KCacheGrind, which is great (works both on unix-like systems and Windows, thanks to Qt5). Here is a screenshot of that for one of Boost.Container's example programs: https://raw.githubusercontent.com/mikael-s-persson/templight-tools/master/i… I really hope that some of you will give this a try. I'm really hoping that it will be useful to you, and that you can provide feedback and feature requests. I couldn't think of any better target audience for this tool. Cheers, Mikael. -- Sven Mikael Persson, M.Sc.(Tech.) PhD Candidate and Vanier CGS Scholar, Department of Mechanical Engineering, McGill University,

2 3

[config] [detail] Macro BOOST_WORKAROUND
by Andrzej Krzemienski 28 Jan '15

28 Jan '15

Hi Everyone, I have a couple of concerns regarding macro BOOST_WORKAROUND. 1. Is usage thereof recommended? 2. Why is it documented in Boost.Config but defined in Boost.Detail? 3. Should it not belong to Boost.Config? 4. Why should I use it rather than manual if (BOOST_INTEL < 551) ? Now, that it is in Boost.Detail, I have to make a dependency on Boost.Detail, and indirectly on Boost.MPL because other components in Boost.Detail do. Regards, &rzej

4 4

[test] Who had the emulation of Test written using gtest?
by Niall Douglas 28 Jan '15

28 Jan '15

Sometime during the past few weeks someone mentioned they had written out a set of Boost.Test macros which converted Boost.Test unit test suites into Google Test without needing to modify the Boost.Test using unit test suite. Can whoever that was make themselves known to me please? It would save having to replicate your effort here ... thanks in advance. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

1 0

Re: [boost] [filesystem, regex, asio] These three libraries successfully ported to Windows Phone 8.1
by Moshe Rubin 28 Jan '15

28 Jan '15

On January 27, 2015 09:55 GMT, Paul A. Bristow wrote: >>> And thank you very much for the improved WinRT support. >> >> My pleasure. It was an education for me (in the positive sense of the >> word!) participating in porting Boost. I will only rest on my laurels >> once the port authorized by the crew and back in the repository <g>. > > I'm pleased to see that Windows Phone developers will now be able to use these Boost libraries. When we went looking for Boost ports for Windows Phone 8, we had a sinking feeling because no one had ported any of the libraries to WinRT. Steven Gates of Microsoft took the first step, and was a great encouragement to me and my company. I'm glad to be able to return the favor to the Boost project, as we have benefited very much from this excellent collection of libraries. > I'm also curious to know if you have tried to port the Boost.Math library? > > Although it contains a lot of esoteric mathy stuff that are unlikely to be mainstream on a phone > app, there are also some very simple-to-use statistics stuff that will/should be widely applicable. > > Have you tried this yet? Are there any obstacles that you can foresee? We did not need Boost.Math in our product so I did not look into building it. I'm happy to say that I just now built it for Windows Phone 8.1 on a Visual Studio 12 2013 system and it compiles and creates the resulting libs like a pro: <command> b2 toolset=msvc-12.0 windows-api=phone architecture=arm link=static --with-math variant=debug debug-store=database debug-symbols=on pch=off cxxflags=-ZW stage </command> Libraries created: (*) libboost_math_tr1-vc120-mt-gd-1_58.lib (*) libboost_math_tr1f-vc120-mt-gd-1_58.lib (*) libboost_math_c99-vc120-mt-gd-1_58.lib (*) libboost_math_c99f-vc120-mt-gd-1_58.lib The only warning message received, and there are many of these, is LNK4264: <snippet> msvc.archive bin.v2\libs\math\build\...\libboost_math_tr1-vc120-mt-gd-1_58.lib bin.v2\libs\math\build\...\sph_neumann.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata bin.v2\libs\math\build\...\sph_legendre.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata bin.v2\libs\math\build\...\sph_bessel.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata . . . bin.v2\libs\math\build\...\beta.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata bin.v2\libs\math\build\...\assoc_legendre.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata bin.v2\libs\math\build\...\assoc_laguerre.obj : warning LNK4264: archiving object file compiled with /ZW into a static library; note that when authoring Windows Runtime types it is not recommended to link with a static library that contains Windows Runtime metadata common.copy stage\lib\libboost_math_tr1-vc120-mt-gd-1_58.lib bin.v2\libs\math\build\...\libboost_math_tr1-vc120-mt-gd-1_58.lib 1 file(s) copied. </snippet> I encountered many of these warnings when building my ported libraries. The online documentation seems to indicate that this is a recommendation. I can tell you that the ported libraries, with these warnings, all function correctly with no problems. In short, it looks like Boost.Math compiles as-is for Windows Phone 8.1. Moshe

2 1

Re: [boost] [filesystem, regex, asio] These three libraries successfully ported to Windows Phone 8.1
by Moshe Rubin 28 Jan '15

28 Jan '15

On 28 Jan 2015 at 02:04:22 GMT, Niall Douglas wrote: > And once again, my apologies for earlier. It is no excuse, but that > particular day I had the misfortune of combining codeine and a lot of > cough medicine due to a heavy flu. This affected my mind. I was quite > appalled with myself when I read what I had written to you the next > day, very embarrassing, I was clearly off in a grim imagination land > and had completely underestimated its effects on my wits. > > As a token of making good on my earlier behaviours, do feel free to > email me off list to review your pull requests for any problems, > either before or after you submit them. And thank you once again for > making the effort, it is hugely appreciated here. Rest assured, I harbor no ill will at all. Quite the contrary, notwithstanding the text written, it was obvious you were a battle-seasoned professional and knowledgeable software engineer, one worth listening to. As a newbie to the Boost mailing list, I have gained invaluable information from you and others that will stand me in good stead. The ones to feel sorry for are those involved in the Linux kernel -- Linus Torvalds most certainly does not apologize for his comments! <g> I appreciate your candor, a welcome trait in the professional world. I'm currently working on verifying the changes using the regression test suite, and cleaning up the changes based on the comments you posted the other day. Thank you for your kind offer to eyeball my pull requests before or after submitting them. I will certainly feel free to take you up on your offer. Moshe

1 0

[filesystem, regex, asio] These three libraries successfully ported to Windows Phone 8.1
by Moshe Rubin 28 Jan '15

28 Jan '15

After the helpful advice I received in this mailing list and other forums, I'm happy to report that the three Boost libraries, filesystem, regex, and asio have been successfully ported to Windows Phone 8.1. Our corporate product is back up and running on Windows Phone 8.1 devices. I assume there are others who will be happy to download the new Boost code for use in their WP 8.1 apps. These libraries will hopefully join those already ported by Steven Gates of Microsoft [http://blogs.msdn.com/b/vcblog/archive/2014/07/18/using-boost-libraries-in-…] I am therefore beginning the process of verifying the port using the Boost regression test suite (I have to decide how to get them to run under Windows Phone 8.1). When the code has passed the regression suite I will issue a pull request against the GitHub repository. Thanks again to everyone in this list for their helpful comments and suggestions, Moshe Rubin Technology Manager Hyperlync Technologies Ltd. Jerusalem, Israel

4 5

Re: [boost] [filesystem, regex, asio] These three libraries successfully ported to Windows Phone 8.1
by Moshe Rubin 28 Jan '15

28 Jan '15

On 26 Jan 2015 15:51, Niall Douglas wrote: >> When the code has passed the regression suite I will issue a pull >> request against the GitHub repository. > Some tips for creating a successful pull request: > > 1. Avoid #ifdef spam. Write thunk emulations of missing APIs where possible. Here's an example for GetTickCount64(): > https://github.com/boostorg/thread/blob/master/include/boost/thread/wi > n32/thread_primitives.hpp#L290. Remember Boost still supports Windows XP and exclusively targets an XP level Win32 API. Use GetProcAddress for newer APIs. > > 2. Avoid using macros to implement functionality. > > 3. Don't forget to modify the documentation where needed to support the changes. > > 4. It might sound obvious, but make sure the unit test suite still passes on Windows XP in default build settings on all of MSVC-8, > mingw32 and mingw-w64. You'd be surprised how often pull requests don't because they use some Vista or later API, or an API not supported by mingw32. If a pull request doesn't instantly go all green on all supported platforms it tends to not be merged and get put > off till much later (sometimes forever). > > Right now cygwin support is broken at least in thread, but other libraries are probably a lot better, so test that too. > > 5. Consider additional unit tests where appropriate. In this situation I'd doubt you'd need them. These are excellent points, obviously the product of your own "battles" with modifying the Boost source base. Many thanks for sharing them, and now is the perfect time to learn them. > And thank you very much for the improved WinRT support. My pleasure. It was an education for me (in the positive sense of the word!) participating in porting Boost. I will only rest on my laurels once the port authorized by the crew and back in the repository <g>. Moshe

3 2

[release] Release schedule changes..
by Rene Rivera 27 Jan '15

27 Jan '15

After some discussion in the Boost Release Team we've decided on an ongoing schedule for releases. Taking into account the realities of how long it takes for authors and the release team to work on a release we've decided to move to *three* releases a year. The targets for the releases will now be: * Spring release on the second Wednesday of April. * Summer release on the second Wednesday of August. * Fall/Winter release on the second Wednesday of December. There will be no January release. This will allow for the fact that most people are on extended holidays during Winter. It will also allow for people to work on larger changes that require more testing time during the longer Winter period. We are also moving to Wednesday release to allow the release managers non-weekend time to work before and after the release. Please consult the Boost calendar for specific dates and merge schedule for the next release. Thanks, the Boost Release Team. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

2 1

Boost modular: Build without version number suffixes
by Robert Dailey 27 Jan '15

27 Jan '15

When I build modular boost on Linux Mint 17, I get SO files named like so: libboost_date_time.so.1.57.0 with a corresponding symlink without the version number: libboost_date_time.so I'd like it to build without the version number because when I rename the base file to remove the version, code links fine but during runtime it tries to load the SO with the version number at the end. I don't know how to fix that, so I assume the fix is to build boost without the version numbers. Any help is appreciated.

5 8

[doc][build]
by Damian Vicino 27 Jan '15

27 Jan '15

Hi, I noticed the boost.build site links to documentations are broken in http://www.boost.org/doc/libs/1_57_0/tools/build/index.html <http://www.boost.org/doc/libs/1_57_0/tools/build/index.html>: http://www.boost.org/doc/libs/doc/html/bbv2.html <http://www.boost.org/doc/libs/doc/html/bbv2.html> http://www.boost.org/doc/libs/1_57_0/tools/build/doc/userman.pdf <http://www.boost.org/doc/libs/1_57_0/tools/build/doc/userman.pdf> both return 404. I got there following links in installation documentation.

1 0