Re: [capsicum] wait() and pdfork()



On 5 Mar 2014, at 15:19, David Drysdale <drysdale at google.com> wrote:

> On a specific detail -- what do we expect to happen for waitpid(pdforked_child_pid,...) ?  This works at the moment, but I could be persuaded either way as to whether it should or not.

I would expect it to work, because anything that is allowed to call waitpid() has access to the global pid namespace and so should be allowed to wait for any pid to terminate.  

I would, however, expect that *all* pdwait4() calls with the same process descriptor succeed until the last process descriptor to the process is closed and that there be no possible race (as there is with waitpid() if the pid is recycled before the waitpid() call starts, which causes problems for debuggers).  The kernel should preserve the exit status and rusage values for the process for as long as there exists a process descriptor to it.  

David





This archive was generated by a fusion of Pipermail (Mailman edition) and MHonArc.