Hi,
I sent an e-mail a couple of weeks ago regarding the proper usage of
apr_pool_note_subprocess(). After looking in the code, I believe that
there may be a bug somewhere. I call apr_pool_note_subprocess with
s->process->pool as the memory pool (I'm writing an Apache module), the
process structure of type apr_proc_t of a process I just created, and
APR_KILL_AFTER_TIMEOUT as arguments. By the time the pool is in the
process of being destroyed and free_proc_chain() is called, the
subprocesses that I registered no longer exist in the process_chain linked
list.
I believe that the problem is either inadequate knowledge on my part or an
actual bug in the library. I would appreciate any help you can provide.
Regards,
Erik Lotspeich