Selon Guillaume Thouvenin <guillaume.thouvenin@polymtl.ca>: > I changed function elsa_copy_parent_bank(struct task_struct *p) to > elsa_copy_parent_bank(struct task_struct *from, struct task_struct *to) > where > "from" is the parent and "to" is the child. I also added a printk to see pid > of > processus implied in the crash. I can see that crash occured when: > "elsa_copy_parent_bank: from pid#0 to pid#1" > > AFAIK, process 0 is the idle task and process 1 is "init". My question is, > where > are process 0 fields initilized? Others processes are initialized during > the > fork (right?) but how/where process 0 starts? I think that I found the problem but I'm not sure that's why I'd like to have a confirmation. I think that the initialization of the process that I put in kernel/fork.c inside copy_process() is not enough. It seems that by doing that, process 0 is not initialized (I don't know why). So, when calling elsa_copy_parent_bank(0,1), it causes an oops because current->bank_head is not initialized. To correct that I add another initialization in fork_init()..... I understand that fork_init() is called when kernel starts to initialize information used during fork so, this function is just called once, right? Is copy_process() called with current->pid == 0? I'm not very happy with my implementation because I must initialize the process 0 and other process, is it possible to do it in the same appropriate place for all processes? Thanks, Guillaume -- Kernelnewbies: Help each other learn about the Linux kernel. Archive: http://mail.nl.linux.org/kernelnewbies/ FAQ: http://kernelnewbies.org/faq/
