s390 && user_enable_single_step() (Was: odd utrace testing results on s390x)

[Martin Schwidefsky]

On Thu, 7 Jan 2010 19:16:32 +0100
Oleg Nesterov <oleg at redhat.com> wrote:

> On 01/07, Martin Schwidefsky wrote:
> >
> > On Wed,  6 Jan 2010 13:08:12 -0800 (PST)
> > Roland McGrath <roland at redhat.com> wrote:
> >
> > > That's what tracehook_signal_handler is for.  You're both doing it yourself
> > > in the arch code (by setting TIF_SINGLE_STEP), and then telling the generic
> > > code to do it (by passing stepping=1 to tracehook_signal_handler).
> >
> > Ok, so with the full utrace the semantics of tracehook_signal_handler
> > is more than just causing a SIGTRAP. It is an indication for a signal
> > AND a SIGTRAP if single-stepping is active. To make both cases work we
> > should stop setting TIF_SINGLE_STEP in do_signal and pass
> > current->thread.per_info.single_step to tracehook_signal_handler
> > instead of test_thread_flag(TIF_SINGLE_STEP).
> 
> Can't understand why do we need TIF_SINGLE_STEP at all.
> 
> Just pass current->thread.per_info.single_step to
> tracehook_signal_handler() ?
> 
> Oleg.
> 
> --- a/arch/s390/kernel/signal.c
> +++ b/arch/s390/kernel/signal.c
> @@ -504,14 +504,8 @@ void do_signal(struct pt_regs *regs)
>  			 * for a normal instruction, act like we took
>  			 * one for the handler setup.
>  			 */
> -			if (current->thread.per_info.single_step)
> -				set_thread_flag(TIF_SINGLE_STEP);
> -
> -			/*
> -			 * Let tracing know that we've done the handler setup.
> -			 */
>  			tracehook_signal_handler(signr, &info, &ka, regs,
> -					 test_thread_flag(TIF_SINGLE_STEP));
> +					 current->thread.per_info.single_step);
>  		}
>  		return;
>  	}
> 

That is what I meant in the other mail. The patch on my local disk
looks almost the same but it removes the comment prior to the
TIF_SINGLE_STEP if statement as well.

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.