Subj : Re: pthread function behavior in C++ program
To   : comp.programming.threads
From : Peter Dimov
Date : Sun Jun 12 2005 04:43 am

Torsten Robitzki wrote:
> Peter Dimov wrote:
>> Alexander Terekhov wrote:
>
>>> Well, it can be caught using catch(...), but not finalized. Under
>>> NPTL, catch(...) of "forced unwind" is supposed to exit with
>>> rethrow, otherwise they terminate the entire process. It's not C++
>>> anymore.
>>
>> Well... that's much better than not running destructors or entering
>> catch(...) blocks at all. When C++ people complain loudly and
>> persistently, the NPTL folks will just stop terminating and all will
>> be well. It's only a matter of time. Even the most stubborn people
>> admit their mistakes given a year or two of the world sadly shaking
>> its collective head.
>
> There are cases, where one realy don't care if an error occures or not
> and this can be expressed with an exception handler who catches all
> exceptions without rethrowing them. Such former legal code would now
> cause every now and then the program to terminate if a thread get
> canceled? Than not runnung destructors and exception handlers at all
> seems to be more predictable to me.

It's better because it's only one step from the "right thing". They only 
need to re-cancel (not rethrow!) instead of terminating. Not running 
destructors and exception handlers will "only" leak and deadlock, but such 
an implementation would be much more difficult to fix.

.