January 2007 - Boost - lists.boost.org

[Report] 25 regressions on RC_1_34_0
by Douglas Gregor 14 Jan '07

14 Jan '07

Boost Regression test failures Report time: 2007-01-14T08:21:02Z This report lists all regression test failures on release platforms. Detailed report: http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/i… 25 failures in 6 libraries concept_check (15) date_time (1) regex (1) test (2) utility (1) wave (5) |concept_check| class_concept_fail_expected: borland-5.6.4 borland-5.8.2 borland-5.8.2 msvc-6.5 msvc-6.5 msvc-7.0 stl_concept_covering: borland-5.8.2 borland-5.8.2 gcc-3.3.6 gcc-3.4.5_linux gcc-3.4.5_linux_x86_64 gcc-4.0.3_linux gcc-4.1.0_linux gcc-4.1.0_linux_x86_64 gcc-4.1.1_sunos_i86pc |date_time| testgenerators: qcc-3.3.5_gpp |regex| regex_regress_threaded: msvc-6.5 |test| prg_exec_fail2: darwin-4.0.1 prg_exec_fail3: cw-9.4 |utility| operators_test: gcc-3.4.5_linux_x86_64 |wave| testwave: darwin-4.0.1 gcc-4.1.1_sunos_i86pc testwave_dll: darwin-4.0.1 gcc-3.3.6 gcc-4.1.1_sunos_i86pc

2 1

[shared_ptr] how does quick_allocator free its memory pool?
by gchen 14 Jan '07

14 Jan '07

I am trying to enable BOOST_SP_USE_QUICK_ALLOCATOR while using shared_ptr, so to make shared_ptr to use quick_allocator to alloc/dealloc memory, But it seems that quick_allocator never free the memory pool (blocks) after my program exits, do I miss something? Or do I need extra step to use quick_allocator? I am using boost cvs head, vs2005, and disabled BOOST_HAS_THREADS (yes, for speed issue, same reason for using quick_allocator), vs2005.

2 2

[Threads] uncoditional scoped lock?
by Sohail Somani 14 Jan '07

14 Jan '07

Hi, Somewhat longer-than-normal-noise follows :) Consider the following pseudo-code which has a race condition: condition g_some_cond; mutex g_condition_mtx; void thread_fn() { { ... // get lock g_some_cond.wait(condition_lock); } // do work } void main_thread() { thread_group grp; // Start all threads and have them wait on the condition for(i in [0,N)) grp.create_thread(&thread_fn); { ... // get lock g_some_cond.notify_all(); // let threads continue } } The race condition is of course that it is possible that not all threads actually start waiting on the condition before notify_all() is called. In this case, it is possible to solve the race condition in two ways: 1) Have the last thread that reaches the condition wait, signal to the main thread 2) Have the main thread checking the condition before notify_all() I don't really like the first case because it requires bi-directional signalling between the threads. Though in my ignorance, I really do consider it the correct solution. The second solution would look something like this: while(true) { scoped_lock cond_lock(g_condition_mtx); // .num_waiting_threads() is made up if(N_EXPECTED_THREADS==g_some_cond.num_waiting_threads()){break;} // release lock } Should the second pattern be encapsulated in the condition object somehow? It is almost like timed_wait(mtx,xt,pred) but there is no timeout. So maybe untimed_wait()? Is there another way to do this? Thanks for your eyes! Sohail

4 6

[Boost-bugs] [ boost-Support Requests-1635211 ] Crash on RH machine with gcc < 3.4
by SourceForge.net 14 Jan '07

14 Jan '07

Support Requests item #1635211, was opened at 2007-01-14 07:01 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=207586&aid=1635211&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: regex Group: None Status: Open Resolution: None Priority: 5 Private: No Submitted By: Nobody/Anonymous (nobody) Assigned to: John Maddock (johnmaddock) Summary: Crash on RH machine with gcc < 3.4 Initial Comment: The following code crshes with Seg.Fault while running on RH machine (kernel 7.1, compiler gcc-3.3.2) #include <boost/regex.hpp> #include <string> #include <stdlib.h> int main() { boost::regex expression("abc"); boost::match_results<std::string::const_iterator> what; boost::match_flag_type flags = boost::match_default; std::string str("abcabc"); if (regex_search( str, what, expression, flags) == true) { printf ("FOUND\n"); } return 0; } Is it a known issue? Thanks, Genia ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=207586&aid=1635211&group_… ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[GIL] Update on GIL
by Lubomir Bourdev 14 Jan '07

14 Jan '07

I just wanted to give the list an update regarding GIL. We are still waiting for the review summary from Tom. In the meantime I am making some fairly significant updates and enhancements to GIL, in part based on suggestions during the review. After those updates are complete (within a couple of months) I will upload the latest GIL to the Boost vault. In the mean time we may post incremental updates on GIL's web page at http://opensource.adobe.com/gil We would like to continue using this web page for people who want to get the latest GIL between Boost releases. I am hoping GIL will be included in the release after 1.34.0 (is that 1.35.0 targeted around the end of 2007?) Thanks, Lubomir

4 3

[GIL] Update on GIL
by Tom Brinkman 14 Jan '07

14 Jan '07

I started a small project using GIL and will relay my experiences in my "review summary" shortly. GIL is a very large library with many innovative features for graphics programmers -- its taken me a few months to get my "head" around it -- and I still have much to learn. >> Why GIL cannot be added to the next release 1.34.0 since it has been voted >> in? I was thinking that once a library has been accepted into Boost it is >> good to go. Do you need to add stuff to the docs or do further changes? GIL needs lots of work to make it "boost" friendly, including the documentation, tests, and examples. As GIL is also part of the "Adobe Open Source" library, which has its own documentation standards, hopefully some compromise can be made between the needs of both projects. I've discussed this with Lubomir before "of-list". My opinion is that GIL has the potential to be a very popular library, but the documentation and examples needs to signifigently expanded to include "a getting started guide" and many more samples, similar to boost::spirit, which is the "gold-standard" for boost::libraries in these areas.

1 0

[asio] misc_ecat is undefined on Windows
by Pavel Kalugin 13 Jan '07

13 Jan '07

There is an error in boost/asio/error.hpp in HEAD - definition of boost::asio::detail::error_base<T>::misc_ecat is unavailable on Windows build. The solution is to move #endif from line 103 to line 98. The following patch works for me: Index: error.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/asio/error.hpp,v retrieving revision 1.11 diff -u -r1.11 error.hpp --- error.hpp 9 Jan 2007 13:54:51 -0000 1.11 +++ error.hpp 13 Jan 2007 04:18:33 -0000 @@ -95,12 +95,12 @@ static std::string addrinfo_md(const boost::system::error_code& ec); static boost::system::wstring_t addrinfo_wmd( const boost::system::error_code& ec); +#endif // !defined(BOOST_WINDOWS) && !defined(__CYGWIN__) static boost::system::error_category misc_ecat; static int misc_ed(const boost::system::error_code& ec); static std::string misc_md(const boost::system::error_code& ec); static boost::system::wstring_t misc_wmd(const boost::system::error_code& ec); -#endif // !defined(BOOST_WINDOWS) && !defined(__CYGWIN__) static boost::system::error_category ssl_ecat; static int ssl_ed(const boost::system::error_code& ec); -- Pavel

2 2

Re: [boost] Final Getting Started Guide
by David Abrahams 13 Jan '07

13 Jan '07

Roland Schwarz <roland.schwarz(a)chello.at> writes: > Hallo David, > > I already posted the below to the list, but got no response. Possibly > it slipped through your attention, No, it's just down in my priority queue. > since I answered to an old post. > > Ah, yes, the link actually is > boost_root/doc/html/thread/build.html#thread.build Thanks... > (without the last ".building") > > Roland > > > David Abrahams wrote: >> > I refactored the getting started guide into separate documents. >> > Please see >> > >> > http://www.boost-consulting.com/boost/more/getting_started.html >> > >> > Library authors have not posted any working URLS of library-specific >> > build instructions, so library names other than Boost.Python all link >> > to the librarys' top-level pages. > > For the thread lib this is: > boost_root/doc/html/thread/build.html#thread.build.building > > This link is available after the docs have been built. No stub available. > > While trying to update this piece of the doc I found the following: > > 1) The getting started guide uses " --toolset= " while the bbv2 manual > states " toolset= ". Obviously both versions work. Which one should go > into the documentation? I think most users who don't know more about BBv2 will be less confused by --toolset=. Of course, they may be confused later when they learn more. Tradeoff. I don't find it's a big priority to change, and since I'm busy... :) > 2) The Boost.Build ref link from the windows.html getting started points > to tools/build/v2. This might be fine when viewed over a web server but > opens a directory when viewed from the filesystem. I suggest correcting > this link to use the index.html file. If I find the time. > 3) The only place I have found where toolset names are explicitely named > is in the getting started guide. I intend to link to > /more/getting_started/windows.html#identify-your-toolset. Questions are > if this is a stable link Should be > and how this will work in a pdf scenario? A "pdf scenario?" I suppose you mean when the Boost docs are generated as PDF? Links to web pages from PDFs just work. That page won't be part of the fraction of Boost docs that are generated as PDF as long as we're going the strict DocBook -> FOP route. > 4) I tried to build the getting_started doc on my local machine. But no > luck. What do I need beyond using docutils ; in my user-config.jam? That's all I have. What are your errors? -- Dave Abrahams Boost Consulting www.boost-consulting.com

2 3

[Report] 20 regressions on RC_1_34_0
by Douglas Gregor 13 Jan '07

13 Jan '07

Boost Regression test failures Report time: 2007-01-11T07:17:04Z This report lists all regression test failures on release platforms. Detailed report: http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/i… 20 failures in 5 libraries concept_check (15) date_time (1) regex (1) test (2) utility (1) |concept_check| class_concept_fail_expected: borland-5.6.4 borland-5.8.2 borland-5.8.2 msvc-6.5 msvc-6.5 msvc-7.0 stl_concept_covering: borland-5.8.2 borland-5.8.2 gcc-3.3.6 gcc-3.4.5_linux gcc-3.4.5_linux_x86_64 gcc-4.0.3_linux gcc-4.1.0_linux gcc-4.1.0_linux_x86_64 gcc-4.1.1_sunos_i86pc |date_time| testgenerators: qcc-3.3.5_gpp |regex| regex_regress_threaded: msvc-6.5 |test| prg_exec_fail2: darwin-4.0.1 prg_exec_fail3: cw-9.4 |utility| operators_test: gcc-3.4.5_linux_x86_64

6 11

Re: [boost] Transition to SVN from CVS? (was: Web access to CVS)
by Gregory Dai 13 Jan '07

13 Jan '07

Hi there, In the recent months, I experienced a transition from using CVS to using SVN with the same project that I've been working on for the past few years in my company. SVN is much faster and much easier to use. I'd suggest that boost make such a transition ASAP.

3 3