> > > > > > Nir Livni (nirl@cyber-ark.com) wrote: > > > > Hi all, > > > > I would like to know where can I read about the do_fork > algorithm > > > > (detailed explaination). > > > > > > > > In addition , maybe you could tell me How do I debug > using UML the > > > > initialization of a forked process. I would like to debug the > > > > initialization of a forked (CHILD) process from the > moment it is > > > > created, until the return of fork() which returns 0. > > > > > > http://lxr.linux.no/source/kernel/fork.c#L577 > > > > > > hope this help. > > > -- > > > Christophe Lucas -- Iomeda SA (clucas@iomeda.fr) > > > > I would still appreciate a bit more info. > > Where is the code that initializes the child process ? > > I have a problem that causes the child to exit immediately after I > > fork() it, So it exits before the fork() returns with 0. > > (The parent receives a normal pid number return code, but > the child simply > > isn't there anymore) > > do_fork() make the initialisation by calling lot of > functions, the child return in ret_from_fork in arch/i386/entry.S > > see > http://lxr.linux.no/source/arch/i386/kernel/process.c#L581 > http://lxr.linux.no/source/arch/i386/kernel/entry.S#L186 > OK... This is getting a bit more complicated than I thought. As I mentioned before, my problem is that a child process exists immediately after its creation (I suppose it is created, because fork() returns well on parent). If I run strace -f -ff on the parent process, the child DOES NOT exit, and continues as planned. I can also tell you that the parent creates few files before fork()ing the child. If I use O_CREAT | O_SYNC - the child DOES NOT exit. If I use O_CREATE without sync - the child exists unexpectedly. According to /proc/self/fd these files are already closed when fork()ing. And I have no Idea where to begin... -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
