Hi,
The code in copy_thread() which sets the ESP is like
this:
childregs = ((struct pt_regs *) (THREAD_SIZE +
unsigned long) p)) - 1;
...
...
p->thread.esp = (unsigned long) childregs;
The value of THREAD_SIZE is 8192. So, effectively the
ESP is set to (p + 8191), which is top of the
allocated kernel stack.
From what I understand, the esp value passed in
pt_regs is the task's user-mode stack pointer
(probably used when switching back to user-mode).
-Ravi.
> The value for the new task's ESP is somehow
> passed
> into the sys_fork() call in the pt_regs struct. I
> don't
> understand how that makes sense; it seems
> self-evident that
> the kernel must set the new task's ESP to the top of
> the
> allocated kernel stack, not to some apparently
> arbitrary
> value passed in by the caller of sys_fork() (which
> of course
> is, ultimately, a user task in most cases).
>
> Thanks,
>
> -- Joe
__________________________________________________
Do You Yahoo!?
Send your FREE holiday greetings online!
http://greetings.yahoo.com
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
IRC Channel: irc.openprojects.net / #kernelnewbies
Web Page: http://www.kernelnewbies.org/
