Re: [Boost-testing] Martin Wille's Intel 8.1 Linux tests

From: Douglas Gregor (doug.gregor_at_[hidden])
Date: 2005-05-01 19:23:36

On May 1, 2005, at 12:17 PM, Hartmut Kaiser wrote:
> The Wave test suite fails with a segmentation fault during Martin's  
> test
> runs as well. I'd appreciate some help on this since I've no way to  
> test it
> on my own.
>
> Douglas, does the Wave tests fail for you on the intel-linux toolset  
> too?

The Wave tests compile okay, but they hang when they run. Here's a  
backtrace from gdb, which seems to indicate that something is going  
awry in the Boost.Pool library:

#0  0x0819d345 in __pthread_sigsuspend ()
#1  0x0819c5f5 in __pthread_wait_for_restart_signal ()
#2  0x0819e895 in __pthread_alt_lock ()
#3  0x0819feba in pthread_mutex_lock ()
#4  0x081e6b0a in _cxa_guard_get_lock ()
#5  0x081e6b5f in __cxa_guard_acquire ()
#6  0x080afa20 in  
boost::details::pool::singleton_default<boost::singleton_pool<boost:: 
fast_pool_allocator_tag, 28, boost::default_user_allocator_new_delete,  
boost::details::pool::null_mutex, 32>::pool_type>::instance() ()
     at singleton.hpp:83
#7  0x080afa90 in  
_ZN5boost7details4pool17singleton_defaultINS_14singleton_poolINS_23fast_ 
pool_allocator_tagELj28ENS_33default_user_allocator_new_deleteENS1_10nul 
l_mutexELj32EE9pool_typeEE14object_creatorC9Ev (this=0x841127c)
     at singleton.hpp:66
#8  0x080afaa3 in  
boost::details::pool::singleton_default<boost::singleton_pool<boost:: 
fast_pool_allocator_tag, 28, boost::default_user_allocator_new_delete,  
boost::details::pool::null_mutex,  
32>::pool_type>::object_creator::object_creator() (this=0x841127c) at  
singleton.hpp:65536
#9  0x08140226 in __sti__$E () at singleton.hpp:65536
#10 0x082456e5 in __do_global_ctors_aux ()
#11 0x080480ca in _init ()
#12 0x081e6d83 in __libc_start_main ()

Unfortunately, I don't have time to debug this at the moment.

        Doug