- Boost - lists.preview.boost.org

[asio] inefficiency in the polling
by Tatsuyuki Ishi 29 Sep '16

29 Sep '16

The current run/poll_one does polling regardless if there's task in the queue. If there's task in the queue, polling is just a waste of system call, since there's already a job to do. Also, my common use case is "wait and consume all work", which is not covered efficiently by current design. (run_one+poll takes two system calls) Side note: on Windows asio performs even worse, it only retrieve one event per syscall... Not sure if it's intended. Friendly note: please use "Reply to All" when replying to me.

4 3

Population count procedure in dynamic_set
by Wojciech Muła 28 Sep '16

28 Sep '16

Hello! The library dynamic_bitset contains a slow population count procedure: https://github.com/boostorg/dynamic_bitset/blob/340822f979371973fe4a8436362… There are several better, faster and (more or less) portable algorithms. Daniel Lemire, Nathan Kurz and I work on a library which provides fastest possible implementations for CPU having AVX2 instruction set, you can check out its current state: https://github.com/CountOnes/hamming_weight. I also did an extensive comparison of different algorithms on different machines: https://github.com/WojciechMula/sse-popcount, one example https://github.com/WojciechMula/sse-popcount/blob/master/results/skylake/sk… As I mentioned these algorithms are different, the fastest use SIMD instructions, other depends on a CPU instruction. However, some approaches are portable, machine-independent, like this https://github.com/WojciechMula/sse-popcount/blob/master/popcnt-bit-paralle… It's possible to provide better solution at small cost, a lot of work has already been done. Would you be interested in including such boosted procedures as part of dynamic_bitset (or a new library)? Popcount itself is used in number of fields: I've found it in DNA matching (Jaccard index), computer vision, databases, and Daniel recently reported application in machine learning. best regards Wojciech

2 1

[container] Trie data structure
by Зайцев Александр 28 Sep '16

28 Sep '16

Hi! My implementation of Aho-Corasick uses my own implementation of trie data structure. But i want something to use from Boost. Is there trie in Boost? If not, can I start work on trie for Boost? -- Best regards, Alexander Zaitsev

6 9

Re: [boost] [Boost-users] [release] Boost 1.62.0 Release Candidate 1
by Marshall Clow 28 Sep '16

28 Sep '16

On Tue, Sep 27, 2016 at 7:14 PM, Edward Diener <eldiener(a)tropicsoft.com> wrote: > On 9/27/2016 9:24 PM, Marshall Clow wrote: > >> >> >> On Mon, Sep 26, 2016 at 5:52 PM, Marshall Clow <mclow.lists(a)gmail.com >> <mailto:mclow.lists@gmail.com>> wrote: >> >> On Mon, Sep 26, 2016 at 7:36 AM, Rene Rivera <grafikrobot(a)gmail.com >> <mailto:grafikrobot@gmail.com>> wrote: >> >> The release candidates for the 1.62.0 release are now available >> at: >> >> >> [snip] >> >> >> >> Now to run some tests.... >> >> -- Marshall >> >> >> The test "vmd/test/test_doc_modifiers_return_type.cpp" hangs clang. >> Reported as https://llvm.org/bugs/show_bug.cgi?id=30546 >> > > Evidently clang runs out of memory. Note that with VC++ and gcc the test > passes. > > I don't know why clang would run out ot memory while VC++ and gcc do not > but I have not queried the clang mailing list further about it. An educated > guess is that clang calculates some fixed size for the preprocessor and > when macro expansions cause that fixed size to overrun you get an out of > memory condition. I can hardly believe that the preprocessor is really > running out of virtual memory, no matter how complicated the preprocessor > code gets in VMD. > > When I was discussing the problem with Richard, he opined that that might be a problem, if the preprocessed source exceeded 1 GB in size - which it doesn't. (GCC expands this test to about 700K) So I suspect it is "just a bug" -- Marshall

1 0

[release] Boost 1.62.0 Release Candidate 1
by Rene Rivera 28 Sep '16

28 Sep '16

The release candidates for the 1.62.0 release are now available at: https://sourceforge.net/projects/boost/files/boost/1.62.0.rc.1/boost_1_62_0… https://sourceforge.net/projects/boost/files/boost/1.62.0.rc.1/boost_1_62_0… https://sourceforge.net/projects/boost/files/boost/1.62.0.rc.1/boost_1_62_0… https://sourceforge.net/projects/boost/files/boost/1.62.0.rc.1/boost_1_62_0… The SHA256 checksums are as follows: b91c2cda8bee73ea613130e19e72c9589e9ef0357c4c5cc5f7523de82cce11f7 boost_1_62_0_rc1.7z 36c96b0f6155c98404091d8ceb48319a28279ca0333fba1ad8611eb90afb2ca0 boost_1_62_0_rc1.tar.bz2 440a59f8bc4023dbe6285c9998b0f7fa288468b889746b1ef00e8b36c559dce1 boost_1_62_0_rc1.tar.gz 084b2e0638bbe0975a9e43e21bc9ceae33ef11377aecab3268a57cf41e405d4e boost_1_62_0_rc1.zip As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. Thanks, the release managers. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

6 9

[dll] Problem building documentation on Windows
by Vinnie Falco 27 Sep '16

27 Sep '16

I am trying to build the Boost documentation on Windows and I am encountering this error bash-4.3$ pwd /d/lib/boost_1_62_0/libs/dll/doc bash-4.3$ bjam Jamfile.v2:66: in modules.load *** argument error * rule doxygen ( target : sources + : requirements * : default-build * : usage-requirements * ) * called with: ( autodoc_shared_library_core : : <doxygen:param>HIDE_UNDOC_MEMBERS=YES <doxygen:param>HIDE_UNDOC_CLASSES=YES <doxygen:param>HIDE_SCOPE_NAMES=YES <doxygen:param>EXTRACT_ALL=NO <doxygen:param>EXTRACT_PRIVATE=NO <doxygen:param>BUILTIN_STL_SUPPORT=YES <doxygen:param>ENABLE_PREPROCESSING=YES <doxygen:param>MACRO_EXPANSION=YES <doxygen:param>ALIASES= \ "forcedlink{1}=\xmlonly<link linkend='boost.dll.\1'>boost::dll::\1</link>\endxmlonly" \ "forcedmacrolink{1}=\xmlonly<link linkend='\1'>\1</link>\endxmlonly" <doxygen:param>PREDEFINED= \ "BOOST_RV_REF(T)=T&&" \ "BOOST_RV_REF(shared_library)=shared_library&&" \ "BOOST_COPY_ASSIGN_REF(shared_library)=const shared_library&" \ "BOOST_MOVABLE_BUT_NOT_COPYABLE(shared_library)= \ shared_library(const shared_library&) = delete; \ shared_library& operator=(const shared_library&) = delete; " \ "BOOST_DLL_IMPORT_RESULT_TYPE=result_type" \ "BOOST_DLL_MANGLED_IMPORT_RESULT_TYPE=result_type" \ "BOOST_EXPLICIT_OPERATOR_BOOL()=explicit operator bool() const noexcept;" \ "BOOST_DLL_DOXYGEN" <xsl:param>boost.doxygen.reftitle=Shared Library Reference ) * missing argument sources D:/lib/boost_1_62_0/tools/build/src/tools\doxygen.jam:622:see definition of rule 'doxygen' being called D:/lib/boost_1_62_0/tools/build/src/build\project.jam:325: in load-jamfile D:/lib/boost_1_62_0/tools/build/src/build\project.jam:64: in load D:/lib/boost_1_62_0/tools/build/src/build\project.jam:145: in project.find D:/lib/boost_1_62_0/tools/build/src\build-system.jam:535: in load D:\lib\boost_1_62_0\tools\build\src/kernel\modules.jam:295: in import D:\lib\boost_1_62_0\tools\build\src/kernel/bootstrap.jam:139: in boost-build D:\lib\boost_1_62_0\boost-build.jam:17: in module scope I get the same error when I try to build the full documentation from boost_1_62_0/doc/ My doxygen is at version 1.8.12 -- Follow me on Github: https://github.com/vinniefalco

2 5

[fiber] design fault with multithreading
by Tatsuyuki Ishi 27 Sep '16

27 Sep '16

The current design doesn't allow a work-stealing algorithm to distribute work perfectly. Unfortunately, the use of sleep queue and remote ready queue made it insufficient for schedulers like shared_work. Instead here, I'd like to propose it as non-pure virtual function in the algorithm class. This allows schedulers to override the queues, and enables the use of underlying implementation's timer manager, which is included in Asio. Friendly notice: please use "Reply All" when replying to me.

1 0

Boost should not be saying "_FPOSOFF"
by Billy O'Neal (VC LIBS) 27 Sep '16

27 Sep '16

Hello, Boost folks. :) I tried to remove this undocumented internal macro from our STL because we used it in only one place, but this broke Boost: #define _FPOSOFF(fp) ((long long)(fp)) ...patience... ...found 722 targets... ...updating 18 targets... common.mkdir build\boost\bin.v2\libs common.mkdir build\boost\bin.v2\libs\iostreams common.mkdir build\boost\bin.v2\libs\iostreams\build common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99 common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\debug common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\file_descriptor.obj file_descriptor.cpp d:\src\vc\wcfb01\qa\vc\FE\3rdpartycomp\boost_1_60_0\boost/iostreams/positioning.hpp(107): error C3861: '_FPOSOFF': identifier not found libs\iostreams\src\file_descriptor.cpp(265): warning C4244: 'argument': conversion from 'std::streamsize' to 'DWORD', possible loss of data libs\iostreams\src\file_descriptor.cpp(289): warning C4244: 'argument': conversion from 'std::streamsize' to 'DWORD', possible loss of data cl /Zm800 -nologo @"build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\file_descriptor.obj.rsp" ...failed compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\file_descriptor.obj... compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\mapped_file.obj mapped_file.cpp d:\src\vc\wcfb01\qa\vc\FE\3rdpartycomp\boost_1_60_0\boost/iostreams/positioning.hpp(107): error C3861: '_FPOSOFF': identifier not found libs\iostreams\src\mapped_file.cpp(59): warning C4244: 'return': conversion from 'const boost::iostreams::stream_offset' to '::size_t', possible loss of data libs\iostreams\src\mapped_file.cpp(325): warning C4244: 'argument': conversion from 'boost::iostreams::stream_offset' to 'SIZE_T', possible loss of data cl /Zm800 -nologo @"build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\mapped_file.obj.rsp" ...failed compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi\mapped_file.obj... ...skipped <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi>libboost_iostreams-vc-mt-gd-1_60.lib for lack of <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi>file_descriptor.obj... ...skipped <pstage\lib>libboost_iostreams-vc-mt-gd-1_60.lib for lack of <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\debug\link-static\threading-multi>libboost_iostreams-vc-mt-gd-1_60.lib... common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\release common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static common.mkdir build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\file_descriptor.obj file_descriptor.cpp .\boost/iostreams/positioning.hpp(107): error C3861: '_FPOSOFF': identifier not found libs\iostreams\src\file_descriptor.cpp(265): warning C4244: 'argument': conversion from 'std::streamsize' to 'DWORD', possible loss of data libs\iostreams\src\file_descriptor.cpp(289): warning C4244: 'argument': conversion from 'std::streamsize' to 'DWORD', possible loss of data cl /Zm800 -nologo @"build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\file_descriptor.obj.rsp" ...failed compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\file_descriptor.obj... compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\mapped_file.obj mapped_file.cpp .\boost/iostreams/positioning.hpp(107): error C3861: '_FPOSOFF': identifier not found libs\iostreams\src\mapped_file.cpp(59): warning C4244: 'return': conversion from 'const boost::iostreams::stream_offset' to '::size_t', possible loss of data libs\iostreams\src\mapped_file.cpp(325): warning C4244: 'argument': conversion from 'boost::iostreams::stream_offset' to 'SIZE_T', possible loss of data cl /Zm800 -nologo @"build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\mapped_file.obj.rsp" ...failed compile-c-c++ build\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi\mapped_file.obj... ...skipped <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi>libboost_iostreams-vc-mt-1_60.lib for lack of <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi>file_descriptor.obj... ...skipped <pstage\lib>libboost_iostreams-vc-mt-1_60.lib for lack of <pbuild\boost\bin.v2\libs\iostreams\build\msvc-99\release\link-static\threading-multi>libboost_iostreams-vc-mt-1_60.lib... ...failed updating 4 targets... ...skipped 4 targets... ...updated 10 targets... failed Billy3

8 18

[iterator] Is there any interest in Null Iterator Adaptor ?
by Rahman Salim Zengin 27 Sep '16

27 Sep '16

Hi! It would be pleasure for me to hear about your precious thoughts. https://github.com/rszengin/null_iterator Best regards.

3 5

Re: [boost] [fiber] Suggestions regarding asio
by Tatsuyuki Ishi 27 Sep '16

27 Sep '16

It would be better if you used "Reply to All", since I disabled the delivery from the list. > When considering using multiple threads with an io_service passed to a custom Boost.Fiber scheduler, I keep bumping into the problem that a given fiber scheduler instance must always run on its original thread -- but I know of no way to give an Asio handler "thread affinity." How do we get processor cycles to the scheduler on the thread whose fibers it is managing? Since we only poll on one master thread, we can easily get affinity. Maybe pseudocode is easier: * We have two class, one for main polling, others waiting for job. * The first one, polls from asio using the same way as the example (poll, or run_one). while(true) { if(!run_one()) stop_other_threads(); // No more work in asio, give up waiting poll(); // flush the queue // Now we should have some fibers in the queue. yield(); // continue execution } * Other threads, uses condvar to wait on the ready queue. awakened(){queue.push_back(f); cv.notify_[one,all](); iosvc.post([]{});} pick_next(){while(!jobs_available){cv.wait();} return job;}

5 7