On Apr 20 11:18, Ernie Rael wrote: > On 4/20/2016 10:22 AM, Canham, Timothy K (348C) wrote: > >I think I understand what happened now. The call to pthread_attr_init() is > >contained within a function we use to start threads. It is called successive > >times and it fails on the second call. It would appear that the stack entry > >gets laid out the exact same way as the previous call, so > >pthread_attr_init() sees the previous pthread_attr_t. I added a memset() of > >att and the issue went away. It looks like that is a good practice anyway. > > Best practice might be using pthread_attr_destroy(attr).
Good point. Not calling pthread_attr_destroy on abandoned pthread_attr_t also results in memory leaks on Cygwin. Corinna -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Maintainer cygwin AT cygwin DOT com Red Hat
signature.asc
Description: PGP signature
