October 2015 - Boost - lists.preview.boost.org

Re: [boost] [interprocess] Interprocess mutex not working / throwing exception when trying to write lock
by nav999 19 Oct '15

19 Oct '15

> Hi, > > 1. Your lockWithWriteLock and lockWithReadLock create a lock, which locks > on construction and unlocks on destruction, hence when the function > returns no lock is held on the mutex. > > 2. The fact that you put a scope here makes me think that your intention > is for the lock to be unlocked at the end of this scope (which it won't, > as it is unlocked way before, as per #1). > > { > std::cout<<"Program 1: Before first locking > -------------------------- 1 v\n"; > temp.sharedMutex->lockWithWriteLock(); > const unsigned int SLEEP_TIME_IN_SECOND = 60; > std::cout<<"Program 1: sleep for "<< SLEEP_TIME_IN_SECOND << "\n"; > sleep(SLEEP_TIME_IN_SECOND); > std::cout<<"Program 1: Finished sleeping\n"; > } > > In this case though, you would perform the wait in bold while still > holding the lock, and hence preventing the 2nd program from locking it. > Once out of the scope you destroy the object right away. > > NOTE: both the issue in #1 and #2 lead to your first program possibly > deleting the mutex before the 2nd program is done with it. Which leads to > the issue you are experiencing. > > 3. If you need to have the 1st program wait for the 2nd program to be > done, then do not use sleeps, use a condition_variable. Thanks. I didn't know the lock unlocks on destruction. I had taken a cue from a post on StackOverflow: http://stackoverflow.com/questions/12439099/interprocess-reader-writer-lock… I'm using sleep only for testing purpose. The dilemma still is, how to make the mutex visible to both processes? This thread got rejected by the admin who suggested it could be posted on the users subforum, so kindly reply to my post there to continue the discussion: http://boost.2283326.n4.nabble.com/interprocess-Interprocess-mutex-not-work… -- View this message in context: http://boost.2283326.n4.nabble.com/interprocess-Interprocess-mutex-not-work… Sent from the Boost - Dev mailing list archive at Nabble.com.

1 1

[interprocess] How to know how much memory is taken up by an object allocated in boost interprocess shared memory?
by nav999 19 Oct '15

19 Oct '15

Re-posting, as the boost mailing list rejected the first post <http://boost.2283326.n4.nabble.com/How-to-know-how-much-memory-is-taken-up-…> mentioning that I had to add "[interprocess]" to the subject line. I have asked this question on stackoverflow <http://stackoverflow.com/questions/30642585/why-does-an-object-allocated-in…> , but got no response. For this program: #include <iostream> #include <boost/interprocess/mapped_region.hpp> #include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/containers/list.hpp> #include <iostream> #define SHARED_MEMORY_NAME "SO12439099-MySharedMemory" #define DATAOUTPUT "OutputFromObject" #define INITIAL_MEM 650000 #define STATE_MATRIX_SIZE 4 using namespace std; namespace bip = boost::interprocess; class SharedObject { public: unsigned int tNumber; bool pRcvdFlag; bool sRcvdFlag; unsigned long lTimeStamp; }; typedef bip::allocator<SharedObject, bip::managed_shared_memory::segment_manager> ShmemAllocator; typedef bip::list<SharedObject, ShmemAllocator> SharedMemData; int main() { bip::managed_shared_memory* seg; SharedMemData *sharedMemOutputList; bip::shared_memory_object::remove(DATAOUTPUT); seg = new bip::managed_shared_memory(bip::create_only, DATAOUTPUT, INITIAL_MEM); const ShmemAllocator alloc_inst(seg->get_segment_manager()); sharedMemOutputList = seg->construct<SharedMemData>("TrackOutput")(alloc_inst); std::size_t beforeAllocation = seg->get_free_memory(); std::cout<<"\nBefore allocation = "<< beforeAllocation <<"\n"; SharedObject temp; sharedMemOutputList->push_back(temp); std::size_t afterAllocation = seg->get_free_memory(); std::cout<<"After allocation = "<< afterAllocation <<"\n"; std::cout<<"Difference = "<< beforeAllocation - afterAllocation <<"\n"; std::cout<<"Size of SharedObject = "<< sizeof(SharedObject) <<"\n"; std::cout<<"Size of SharedObject's temp instance = "<< sizeof(temp) <<"\n"; }//main The output is: Before allocation = 649680 After allocation = 649632 Difference = 48 Size of SharedObject = 16 Size of SharedObject's temp instance = 16 If the size of SharedObject and it's instance is 16 bytes, then how can the difference in allocation be 48? Even if padding had automatically been done, it's still too much to account for 3 times the size (for larger structures it goes to 1.33 times the size). Because of this, I'm unable to allocate and dynamically grow the shared memory reliably. If SharedObject contains a list which grows dynamically, that could add to the uncertainty of space allocation even more. How can these situations be safely handled? How can I know exactly how much of memory to allocate? -- View this message in context: http://boost.2283326.n4.nabble.com/interprocess-How-to-know-how-much-memory… Sent from the Boost - Dev mailing list archive at Nabble.com.

3 4

Do you know of standard implementation of async() that blocks on the future-destructor?
by Vicente J. Botet Escriba 18 Oct '15

18 Oct '15

Hi, all is in the title. Best, Vicente

3 5

[Math] Help needed with FreeBSD failures
by John Maddock 18 Oct '15

18 Oct '15

Can anyone familiar with FreeBSD help me figure out what's causing a large number of runtime failures from Boost.Math, for example http://www.boost.org/development/tests/develop/developer/output/Flast-FreeB… Strangely gcc-5.1.0 is fine, but 4.8.5, 5.2.0 and 6.0 all fail. Many thanks in advance, John.

2 4

[serialization] Massive failures in develop?
by John Maddock 18 Oct '15

18 Oct '15

Local testing seems to have badly failed: http://www.boost.org/development/tests/develop/developer/serialization.html I've only looked at the msvc failures, and there's an issue in the iostream class destructor - which is to say after the archive classes have been destructed, the iostream class they were using is left in a non-sane state. Beyond that I haven't been able to get anything sensible out of the debugger (looks like memory corruption actually, but I can't be sure). Any ideas? Thanks, John.

2 6

[thread] Do we can a non backward compatible version with non-blocking futures?
by Vicente J. Botet Escriba 18 Oct '15

18 Oct '15

Hi, in Boost 1.58 I fixed some issues as the result of the async and future<T>:then() was not a blocking future as the documentation described. However the implementation is horrible and I suspect that it is subject to deadlocks. BTW, I added already some fixes on the develop branch. Sometime ago I started in a branch an implementation that don't use blocking futures [1]. Well the implementation is also horrible after making futures non-blocking, but less ;-) The problem is that now I have some issues with the executors :( Anyway, Gor has told me that with the blocking implementation he is unable to define await_suspend (using resumable functions) >It is just if .then returns a blocking future, I cannot subscribe a callback to .then in await_suspend. >await_suspend(f, cb) { > f.then(cb); >} &will block here preventing coroutine from suspending I don't know yet if he will reach to configure boost::future<T> on the non-blocking branch. My question is if the Boost community will accept (or even wants) a breaking change on a near release (when I would reach to fix the executors issues) including non-blocking futures instead of the current blocking futures. Best, Vicente [1] https://github.com/boostorg/thread/tree/feature/non_blocking_futures

3 13

[intrusive]: determine the hook type of intrusive::list<>?
by Oliver Kowalke 18 Oct '15

18 Oct '15

I'd like to determine the hook type of an intrusive container (intrusive::list): template< typename List > void do_link( List & lst) { static_assert(std::is_same< typename List::???, my_hook_type >::value); lst.push_back( * this); } In the documentation I didn't found a hint. Oliver

1 0

[regression] New runners for old compilers MSVC-7.1 and GCC 3.4
by Ion Gaztañaga 17 Oct '15

17 Oct '15

Hi, The last compiler update two years ago in Boost 1.55 left MSVC-7.1 and GCC 3.3 as two of the oldest supported compilers: http://www.boost.org/users/history/version_1_55_0.html I have no access to GCC 3.3 but managed to get a Mingw-based GCC 3.4. I will run regressions for for both mingw-GCC-3.4 and msvc-7.1 in master and develop branches in case anyone is interested. I plan to maintain support for them in Move, Intrusive, Container and Interprocess until we reach consensus on a new compiler update. Best, Ion

6 6

[python] Merging pull request #25
by Edward Diener 17 Oct '15

17 Oct '15

The PR is necessary so that user's of Python to not see a deprecated warning coming from the updated type_traits whenever they use Boost python. Would the maintainer of Boost Python please take a look at this simple PR so we can get this in the upcoming release. Thanks !

1 0

[Build] Specify compiler on the command line
by Louis Dionne 17 Oct '15

17 Oct '15

Hi, Still attempting to setup Boost.Build for Hana, I find myself unable to specify the compiler to be used by `b2` on the command line. The system compiler is then used by default, which does not work because I need to test on recent compilers. This can be handled by writing a project-config.jam file with the following in it: import feature ; if ! clang in [ feature.values <toolset> ] { using clang : : /path/to/clang ; } However, this is not very script-friendly, and I wonder whether there's a way to specify it on the command-line instead. Regards, Louis

4 6