Re: [boost] [thread] Reducing the number of open tickets.

Subject: Re: [boost] [thread] Reducing the number of open tickets.
From: Vicente J. Botet Escriba (vicente.botet_at_[hidden])
Date: 2011-12-10 07:07:50

Le 05/12/11 00:56, Vicente J. Botet Escriba a écrit :
> Le 03/12/11 18:23, Vicente J. Botet Escriba a écrit :
>> Hi,
>>
>> I have committed the following tickets in trunk.
>>
>> Fixed Bugs:
>>
>> * [@http://svn.boost.org/trac/boost/ticket/2639 #2639] documentation 
>> should be extended(defer_lock, try_to_lock, ...).
>> * [@http://svn.boost.org/trac/boost/ticket/3762 #3762] Thread can't 
>> be compiled with winscw (Codewarrior by Nokia).
>> * [@http://svn.boost.org/trac/boost/ticket/3885 #3885] document about 
>> mix usage of boost.thread and native thread api.
>> * [@http://svn.boost.org/trac/boost/ticket/3975 #3975] Incorrect 
>> precondition for promise::set_wait_callback().
>> * [@http://svn.boost.org/trac/boost/ticket/4480 #4480] OpenVMS 
>> patches for compiler issues workarounds.
>> * [@http://svn.boost.org/trac/boost/ticket/4819 #4819] boost.thread's 
>> documentation misprints.
>> * [@http://svn.boost.org/trac/boost/ticket/5423 #5423] thread issues 
>> with C++0x.
>> * [@http://svn.boost.org/trac/boost/ticket/5739 #5739] 
>> set-but-not-used warnings with gcc-4.6
>> * [@http://svn.boost.org/trac/boost/ticket/6168 #6168] 
>> recursive_mutex is using wrong config symbol (possible typo).
>> * [@http://svn.boost.org/trac/boost/ticket/6200 #6200] patch to have 
>> condition_variable and mutex error better handle EINTR.
> More easy to solve tickets committed in trunk
>
> * [@http://svn.boost.org/trac/boost/ticket/6208 #2309] Lack of g++ 
> symbol visibility support in Boost.Thread.
> * [@http://svn.boost.org/trac/boost/ticket/4315 #4315] gcc 4.4 
> Warning: inline ... declared as dllimport: attribute ignored.
> * [@http://svn.boost.org/trac/boost/ticket/5859 #5859] win32 
> shared_mutex constructor leaks on exceptions.
> * [@http://svn.boost.org/trac/boost/ticket/6208 #6208] 
> try_lock_wrapper swap compiler error on clang 3.0 c++11.
> * [@http://svn.boost.org/trac/boost/ticket/6207 #6207] shared_lock 
> swap compiler error on clang 3.0 c++11.
More, some of them where updated already last week but I forget them in 
the previous list.

* [@http://svn.boost.org/trac/boost/ticket/3639 #3639] Boost.Thread 
doesn't build with Sun-5.9 on Linux.
* [@http://svn.boost.org/trac/boost/ticket/4048 #4048] thread::id 
formatting involves locale
* [@http://svn.boost.org/trac/boost/ticket/5617 #5617] boost::thread::id 
copy ctor.
* [@http://svn.boost.org/trac/boost/ticket/5826 #5826] threads.cpp: 
resource leak on threads creation failure.
* [@http://svn.boost.org/trac/boost/ticket/5839 #5839] thread.cpp: 
ThreadProxy leaks on exceptions.
* [@http://svn.boost.org/trac/boost/ticket/6100 #6100] Compute 
hardware_concurrency() using get_nprocs() on GLIBC systems.
* [@http://svn.boost.org/trac/boost/ticket/6175 #6175] Compile error 
with SunStudio.

> Anthony, please take a look at them and merge whatever you consider is 
> good after the testers have the time to take them in account (not 
> before the 12/12/2011)
>
> I have also closed some tickets that were invalid or duplicated:
>
> #1826     Auto detection of threadapi and target-os broken under 
> cygwin gcc     thread     closed     Bugs         5 hours
> #4648     boost::lock_error Thrown While Using Upgrade Locks     
> thread     closed     Support Requests         5 hours
> #4367     Proposed fix for test_thread_move_return.cpp     thread     
> closed     Bugs         6 hours
> #6130     condition_variable::timed_wait wakes up too early     thread 
>     closed     Bugs     To Be Determined     9 hours
> #5936     Missing prototype for pthread_mutex_timedlock causes 
> Oracle/SUN CC to fail on Linux     thread     closed     Bugs     To 
> Be Determined     14 hours
> #5373     Boost Thread fails on SunStudio/OpenSolaris     thread     
> closed     Bugs     To Be Determined     23 hours
> #4894     segmentation fault on thread_group::join_all depending on 
> the machine used     thread     closed     Support Requests     To Be 
> Determined     23 hours
> #6170     boost::move conflicts in thread upgrade mutexes (1.48.0)     
> thread     closed     Bugs     To Be Determined     2 days
> #5344     condition_variable.hpp warning     thread     closed     
> Patches     To Be Determined     3 days
> #4773     boost::this_thread::sleep(Microsecond Resolution)
More closed tickets either duplicate, invalid or already committed 
previously.
#5727 <https://svn.boost.org/trac/boost/ticket/5727> race condition 
between ~basic_condition_variable() and notify_all() 
<https://svn.boost.org/trac/boost/ticket/5727>
#5699 <https://svn.boost.org/trac/boost/ticket/5699> 
boost::thread_specific_ptr will reuse a value if allocated with the same 
this pointer <https://svn.boost.org/trac/boost/ticket/5699>
#4849 <https://svn.boost.org/trac/boost/ticket/4849> MinGW-64: 
detail/interlocked.hpp chooses wrong prototypes for 
InterlockedCompareExchange et al 
<https://svn.boost.org/trac/boost/ticket/4849>
#5150 <https://svn.boost.org/trac/boost/ticket/5150> boost::thread does 
not print or return the error value when creating a thread fails. 
<https://svn.boost.org/trac/boost/ticket/5150>

Should these duplicated or invalid closed tickets appear in the change list?
<https://svn.boost.org/trac/boost/timeline?from=2011-12-06T19%3A11%3A16-05%3A00&precision=second> 

>
> 69 Bugs (1 with patch)
> 17 Feature requests (1 with patch)
> 3 support requests
> 1 task
> 90 TOTAL
52 Bugs (1 with patch)
16 Feature requests
13 support requests
10 task
91 TOTAL
+
22 to be merged and closed

As you see the total number of tickets is not decreasing yet, there were 
some news since. However, I moved some to Support request when I 
believed that there weren't bugs and the resolution needed to be 
clarified. I expect most of them will be closed on a week.

Hoping I'm not introducing regressions ;-)

Best,
Vicente