Re: [PATCH] uprobes: Improve scalability by reducing the contention on siglock

Re: [PATCH] uprobes: Improve scalability by reducing the contention on siglock

[Liao, Chang]

在 2024/8/8 18:28, Oleg Nesterov 写道:
> On 08/08, Liao, Chang wrote:
>>
>>   - pre_ssout() resets the deny signal flag
>>
>>   - uprobe_deny_signal() sets the deny signal flag when TIF_SIGPENDING is cleared.
>>
>>   - handle_singlestep() check the deny signal flag and restore TIF_SIGPENDING if necessary.
>>
>> Does this approach look correct to you,do do you have any other way to implement the "flag"?
> 
> Yes. But I don't think pre_ssout() needs to clear this flag. handle_singlestep() resets/clears
> state, active_uprobe, frees insn slot. So I guess we only need
> 
> 
> --- x/kernel/events/uprobes.c
> +++ x/kernel/events/uprobes.c
> @@ -2308,9 +2308,10 @@ static void handle_singlestep(struct upr
>  	utask->state = UTASK_RUNNING;
>  	xol_free_insn_slot(current);
>  
> -	spin_lock_irq(&current->sighand->siglock);
> -	recalc_sigpending(); /* see uprobe_deny_signal() */
> -	spin_unlock_irq(&current->sighand->siglock);
> +	if (utask->xxx) {
> +		set_thread_flag(TIF_SIGPENDING);
> +		utask->xxx = 0;
> +	}

Agree, if no more discussion about this flag, I will just send v2 today.

Thanks.

>  
>  	if (unlikely(err)) {
>  		uprobe_warn(current, "execute the probed insn, sending SIGILL.");
> 
> and that is all.
> 
> Oleg.
> 
> 

-- 
BR
Liao, Chang

Re: [PATCH] uprobes: Improve scalability by reducing the contention on siglock

[Oleg Nesterov]

On 08/08, Liao, Chang wrote:
>
>
> 在 2024/8/8 18:28, Oleg Nesterov 写道:
> > --- x/kernel/events/uprobes.c
> > +++ x/kernel/events/uprobes.c
> > @@ -2308,9 +2308,10 @@ static void handle_singlestep(struct upr
> >  	utask->state = UTASK_RUNNING;
> >  	xol_free_insn_slot(current);
> >
> > -	spin_lock_irq(&current->sighand->siglock);
> > -	recalc_sigpending(); /* see uprobe_deny_signal() */
> > -	spin_unlock_irq(&current->sighand->siglock);
> > +	if (utask->xxx) {
> > +		set_thread_flag(TIF_SIGPENDING);
> > +		utask->xxx = 0;
> > +	}
>
> Agree, if no more discussion about this flag, I will just send v2 today.

Please also resend the previous patch a 1/2, this one as 2/2.

Oleg.