- Boost - lists.boost.org

[release] Boost 1.79.0 Beta 1 Release Candidate 1 is available
by Marshall Clow 11 Mar '22

11 Mar '22

are now available at: <https://boostorg.jfrog.io/ui/native/main/beta/1.79.0.beta1/source <https://boostorg.jfrog.io/ui/native/main/beta/1.79.0.beta1/source>> The SHA256 checksums are as follows: 8985105623f48d55be01076d98ca4e7a8cb548a8d7f9ad0e1fc3c710d2c75fcf boost_1_79_0_b1_rc1.tar.bz2 4b58e4d3b6ff1176b3c6cb9eec496c1f6d2535270e97ec140da533d68766ca39 boost_1_79_0_b1_rc1.tar.gz 768875ad5c77763445e8b119072cfa1e64854b071a353dd44c3c73ff46df910b boost_1_79_0_b1_rc1.zip 4d05b458cd039afe86003cf7e1e6c3a358b5b2d031892e620eedc344ab200246 boost_1_79_0_b1_rc1.7z 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. -- The Release Managers

4 8

Master branch now closed for 1.79.0 beta release
by Marshall Clow 10 Mar '22

10 Mar '22

Please do not make any commits to the master branch. — Marshall

1 0

[Graph] Attempt to resolve open issue
by Anne van de Graaf 10 Mar '22

10 Mar '22

Hello all, I am trying to get an open issue in Boost.Graph resolved. Since I haven't received any response at the corresponding GitHub page nor from the Boost Users mailing list (where I posted the below message last Monday), I am reaching out to you. My aim is to get in touch with the maintainer(s) of Boost.Graph in order to discuss what is the preferred/most efficient way to fix the issue. Since this is my first post to this mailing list, please accept my apologies if you feel this is not the right way to get in touch with the maintainers. In that case, I would appreciate it very much if you could kindly explain what I should have done instead. Please find below a copy of my message that I sent last Monday to the Boost Users mailing list. Thank you very much for your attention and any replies. Original message: When I remove a vertex from a graph, I occasionally experience a crash (on a system running Red Hat Enterprise Linux 8.5). I have been able to replicate the behavior with the following test program: #include <vector> #include <boost/graph/adjacency_list.hpp> int main() { using MyGraph = boost::adjacency_list<boost::setS, boost::vecS, boost::undirectedS>; using MyVertexDescriptor = boost::graph_traits<MyGraph>::vertex_descriptor; using MyVertexDescriptorVec = std::vector<MyVertexDescriptor>; MyGraph g; MyVertexDescriptorVec vertices{ boost::add_vertex( g ), boost::add_vertex( g ), boost::add_vertex( g ) }; boost::add_edge( vertices[0], vertices[1], g ); boost::add_edge( vertices[1], vertices[2], g ); boost::clear_vertex( vertices[0], g ); boost::remove_vertex( vertices[0], g ); return 0; } Also on a Windows platform, I observe similar behavior. When I build the above code with Microsoft Visual Studio 2019 (MSVC 14.27.29110) and run it with a debugger, a debug assertion is triggered ("cannot increment value-initialized map/set iterator"). In both cases (on Linux and Windows platforms), the debugger reveals that the problem is at line 2071 of boost/graph/detail/adjacency_list.hpp. For clarity, I have copied the involved template function below: template < class EdgeList, class vertex_descriptor > inline void reindex_edge_list( EdgeList& el, vertex_descriptor u, boost::disallow_parallel_edge_tag) { for (typename EdgeList::iterator ei = el.begin(); ei != el.end(); ++ei) // line 2071: increment operator fails { if (ei->get_target() > u) { typename EdgeList::value_type ce = *ei; el.erase(ce); --ce.get_target(); el.insert(ce); } } } While I was searching for a proper place to post my question, I found the following issue at GitHub (which was opened on July 1 of this year): https://github.com/boostorg/graph/issues/268. This appears to be the exact same issue. So last week, I added a comment saying that I am facing the exact same problem. Unfortunately, so far no one has responded yet. To me, it seems that the problem is well-understood. Also, some have proposed a bug fix to resolve the issue (including a pull request). Nevertheless, no progress has been made over the past months. So I was just wondering what is the best way to proceed if we want to fix the issue in Boost itself? Is there anything that I can do (apart from starting this thread) to help in resolving the issue? Any feedback is greatly appreciated. Kind regards, Anne van de Graaf

3 6

Windows Build Issue for Boost Wave
by Tom Kent 09 Mar '22

09 Mar '22

In preparation for the lockdown today, I did a dry run of the windows build based on the latest snapshot. During this build I encountered errors with wave on the msvc-12.0 compiler (both 32 & 64 bit). Roughly the following is repeated several times during the build: compile-c-c++ D:\RB\bin.v2\boost\bin.v2\libs\wave\build\msvc-12.0\debug\address-model-32\threading-multi\token_ids.obj token_ids.cpp D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(335) : error C2146: syntax error : missing ';' before identifier 'token_id' D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(335) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(341) : error C2146: syntax error : missing ';' before identifier 'token_id' D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(341) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(341) : error C2086: 'int boost::wave::constexpr' : redefinition D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(335) : see declaration of 'boost::wave::constexpr' D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(347) : error C2144: syntax error : 'bool' should be preceded by ';' D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(347) : error C4430: missing type specifier - int assumed. Note: C++ does not support default-int D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(347) : error C2086: 'int boost::wave::constexpr' : redefinition D:\RB\boost_1_79_0\boost/wave/token_ids.hpp(335) : see declaration of 'boost::wave::constexpr' call "D:\RB\bin.v2\boost\bin.v2\standalone\msvc\msvc-12.0\address-model-32\architecture-x86\msvc-setup.bat" x86 >nul cl /Zm800 -nologo "libs\wave\src\token_ids.cpp" -c -Fo"D:\RB\bin.v2\boost\bin.v2\libs\wave\build\msvc-12.0\debug\address-model-32\threading-multi\token_ids.obj" -TP /wd4675 /EHs /GR /Z7 /Od /Ob0 /W3 /MDd /Zc:forScope /Zc:wchar_t /Zc:inline -DBOOST_ALL_DYN_LINK=1 -DBOOST_ALL_NO_LIB=1 -D_CRT_SECURE_NO_DEPRECATE -D_SCL_SECURE_NO_DEPRECATE "-I." ...failed compile-c-c++ D:\RB\bin.v2\boost\bin.v2\libs\wave\build\msvc-12.0\debug\address-model-32\threading-multi\token_ids.obj... Hope this helps, if not fixed it will be blocking the beta rc when it gets rolled. Tom

1 0

Boost 1.79.0 closed for breaking changes and new libraries today
by Glen Fernandes 02 Mar '22

02 Mar '22

We are closed for new libraries and breaking changes today. We will close for all major changes this Friday (2022-03-04). And will close for the beta the following Wednesday (2022-03-09). The Release Managers

1 0

std::numeric_limits<__float128>::signaling_NaN + __float128-Type
by Gero Peterhoff 26 Feb '22

26 Feb '22

Hello, please see https://godbolt.org/z/Ksshxvhaj 1) std::numeric_limits<__float128>::signaling_NaN always returns 0 but gcc-__builtin_nansq doesn't. I think __builtin_nansq is correct as it is also distinct from std::numeric_limits<__float128>::quiet_NaN (__builtin_nanq). 2) The types __float128, __extension__ __float128 (gcc) and _Quad (intel) are identical, but there are different definitions in boost/config/detail/suffix.hpp and boost/math/cstdfloat/cstdfloat_types.hpp. I think it would make sense to have just one definition or is there a reason why it was done that way? If yes, which one is it? You could always use __float128 directly and/or adjust boost/config/detail/suffix.hpp so that it also works with clang/intel. Which way would be better? I would like to help consolidate this but without having to deal with different typedef's or long #define's. thx Gero

3 13

Boost.URL Proposal
by Alan de Freitas 21 Feb '22

21 Feb '22

Hi! My name is Alan Freitas. Vinnie Falco and I are looking for a review manager for the library that we hope will become Boost.URL. I've taken over the project and in addition to introducing myself I would like to update you guys as to where we are on this in terms of progress: - Path segments can be represented as views of the underlying path - Query parameters can be represented as views of the underlying query - Percent-decoding methods return lazy constant strings with small buffer optimization and shared ownership - The library implements the specification as in RFC3986 <https://datatracker.ietf.org/doc/html/rfc3986> + its errata <https://www.rfc-editor.org/errata/rfc3986> - Non-ASCII characters are not considered valid URLs - Other specifications such as RFC3492 <https://datatracker.ietf.org/doc/html/rfc3492>, WhatWG <https://url.spec.whatwg.org/>, or any other form of relaxed parsing are out of the scope of the library Both Vinnie and I are available to answer questions about the library on the list, on the GitHub repository page <https://github.com/CPPAlliance/url>, via email <alandefreitas(a)gmail.com>, or on Slack (in #boost-url). -- Alan Freitas https://github.com/alandefreitas

5 7

Understanding the code base.
by Rahul Sharma 21 Feb '22

21 Feb '22

So far I have been reading multiple documentations and reading the projects from past year GSoC and exploring some modules given on the GitHub and I came to understand now that in order to contribute to this organization the knowledge of only C++ and data structures is not enough. I have some questions which are halting my progress so far, it would be very generous of you if someone could address these issues I am facing…… 1. I am unable to understand the code base in the repositories on GitHub. 2. In order to contribute or to understand things in a better way do I need to learn how Boost libraries work. 3. What are some packages I need to install. It would be very kind of you if you could provide some useful resources for the same and also for the things you think could be of use to understand things clearly and for making some PRs. When I read the code base of some repositories and some documentations I got excited as I think there are many things and skills I need to learn, the skills, it takes to be a decent developer. Sent from [1]Mail for Windows References 1. https://go.microsoft.com/fwlink/?LinkId=550986

2 1

[PATCH] futex: fix build on 32-bit architectures using 64-bit time_t
by Khem Raj 21 Feb '22

21 Feb '22

Fix the following build failure on 32-bit architectures using 64-bit time_t (e.g. riscv32): | ./boost/fiber/detail/futex.hpp:31:23: error: use of undeclared identifier 'SYS_futex' | return ::syscall( SYS_futex, addr, op, x, nullptr, nullptr, 0); | ^ | 1 error generated. Signed-off-by: Khem Raj <raj.khem(a)gmail.com> --- boost/fiber/detail/futex.hpp | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/boost/fiber/detail/futex.hpp b/boost/fiber/detail/futex.hpp index e64bd5990..d0353e9e5 100644 --- a/boost/fiber/detail/futex.hpp +++ b/boost/fiber/detail/futex.hpp @@ -12,6 +12,10 @@ #include <boost/fiber/detail/config.hpp> +#ifndef SYS_futex +#define SYS_futex SYS_futex_time64 +#endif + #if BOOST_OS_LINUX extern "C" { #include <linux/futex.h> -- 2.35.1

1 0

unordered_map implementation
by Peter Dimov 19 Feb '22

19 Feb '22

I've been helping Christian Mazakas maintain Unordered, and we've been looking into trying to improve its performance on the synthetic benchmarks we have for it (https://github.com/boostorg/unordered/tree/develop/benchmark) One of the things that are hurting us is that we need to do a double indirection in order to find an element, because Unordered uses a singly-linked list for all its nodes, with the bucket array pointing at the element before the first one (in order to be able to support erasing.) There's also the option of using a doubly-linked list, which however increases the memory footprint. I'm looking at the iterator invalidation rules for std::unordered_map, and iterators are invalidated on every rehash. This makes me wonder why is it necessary to have all the elements in a single list, as every implementation seems to do. It seems to me that it's possible to just have singly- linked lists for each bucket, and then have a "fat" iterator that contains the bucket index and the pointer to the element. On increment, when the end of the bucket list is reached, the iterator can just go to the next non-empty bucket. Is there something I'm missing in the requirements that renders this implementation non-conforming?

2 2