[PATCH 53-55] (Was: Q: what user_enable_single_step() actually means?)

[Oleg Nesterov]

On 09/25, Roland McGrath wrote:
>
> > > it being after the syscall insn, but (in vanilla ptrace, and at the
> > > time of report_clone callbacks at utrace level) without the return
> > > value register having been written yet.
> >
> > And how we can implement this?
> >
> > regs->ax is updated right after "call *sys_call_table[]", and we
> > report PTRACE_EVENT_FORK or PTRACE_EVENT_EXEC much later.
>
> I didn't intend to.  This is a change I think is more desireable than the
> bug-compatibility.  In today's ptrace these extra stops are the sole
> exceptions where the register state you fiddle with ptrace gets clobbered
> later.  I've never thought that was useful.

Ah, good! I misunderstood you.

Let's ignore this minor incompatibility then.

> > The current logic which delays the stop adds so many complications...
>
> Of course, everything is open for debate.  IMHO it is the old ptrace ABI's
> totally wrong model that introduces complexity.

Yes, agreed.

Oleg.