Messages in this thread | | | Date | Thu, 19 May 2011 19:13:27 +0200 | From | Oleg Nesterov <> | Subject | Re: [PATCH 10/10] ptrace: implement group stop notification for ptracer |
| |
On 05/19, Tejun Heo wrote: > > Hey, > > On Thu, May 19, 2011 at 06:32:46PM +0200, Oleg Nesterov wrote: > > > +static void ptrace_trap_notify(struct task_struct *t) > > > +{ > > > + siginfo_t *si = t->last_siginfo; > > > + > > > + WARN_ON_ONCE(!(t->ptrace & PT_SEIZED)); > > > + assert_spin_locked(&t->sighand->siglock); > > > + > > > + /* > > > + * @t is being ptraced and new SEIZE behavior is in effect. > > > + * Schedule sticky trap which will clear on the next GETSIGINFO. > > > + */ > > > + t->jobctl |= JOBCTL_TRAP_NOTIFY; > > > > This is also set by do_signal_stop(). Cleared by PTRACE_GETSIGINFO. > > > > How can this work? Doesn't this mean PTRACE_GETSIGINFO becomes mandatory > > before PTRACE_CONT? IOW, unless the tracee does PTRACE_GETSIGINFO to clear > > this bit, PTRACE_CONT just leads to another trap, no? > > Yes, group stop state change raises a sticky trap condition which is > cleared by GETSIGINFO.
Hmm. At least now I understand the meaining what "sticky" means in this discussion ;) I was confused.
> > > + if (task_is_traced(t) && si && si->si_code == PTRACE_STOP_SI_CODE) { > > > > OK, this PTRACE_STOP_SI_CODE check is clear. But the same check in > > ptrace_check_attach() looks confusing, why can't we set BLOCK_NOTIFY > > unconditionally? > > It's an optimization. If we set the flag, we'll have to acquire > siglock
OK, I see.
> > > + t->jobctl |= JOBCTL_TRAPPING; > > > + if (!(t->jobctl & JOBCTL_BLOCK_NOTIFY)) > > > + signal_wake_up(t, true); > > > > Could you please remind me why we can't avoid the awful ptrace_wait_trapping() > > in do_wait() paths? Assuming that ptrace_check_attach() does this. I got lost > > a bit. > > Please consider the following scenario. > > 1. Tracee is in group stop and stops at TRAP_STOP notifying the > tracer. > > 2. Tracer does WNOWAIT wait(2) and determines that the tracee is > trapped in TRAP_STOP. > > 3. Something generates SIGCONT which finishes the group stop and > triggers the notification re-trapping. > > 4. While tracee is re-trapping, tracer issues WNOHANG
OK. I still hope we can avoid this somehow. May be play with exit_code so that do_wait() can succeed even if the JOBCTL_TRAPPING tracee is running. Perhaps.
If only we could notify the tracer from ptrace_trap_notify... IIUC, this is the only problem? I mean, apart from this there is no need to wake up the tracee.
Oleg.
| |