linux-kernel - Re: [PATCH v4 2/6] perf: Enqueue SIGTRAP always via task

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20240702090128.GC11386@noisy.programming.kicks-ass.net>
Date: Tue, 2 Jul 2024 11:01:28 +0200
From: Peter Zijlstra <peterz@...radead.org>
To: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
Cc: linux-perf-users@...r.kernel.org, linux-kernel@...r.kernel.org,
	Adrian Hunter <adrian.hunter@...el.com>,
	Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
	Arnaldo Carvalho de Melo <acme@...nel.org>,
	Daniel Bristot de Oliveira <bristot@...nel.org>,
	Frederic Weisbecker <frederic@...nel.org>,
	Ian Rogers <irogers@...gle.com>, Ingo Molnar <mingo@...hat.com>,
	Jiri Olsa <jolsa@...nel.org>, Kan Liang <kan.liang@...ux.intel.com>,
	Marco Elver <elver@...gle.com>, Mark Rutland <mark.rutland@....com>,
	Namhyung Kim <namhyung@...nel.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Arnaldo Carvalho de Melo <acme@...hat.com>
Subject: Re: [PATCH v4 2/6] perf: Enqueue SIGTRAP always via task_work.

On Mon, Jul 01, 2024 at 03:27:25PM +0200, Sebastian Andrzej Siewior wrote:
> On 2024-07-01 14:28:29 [+0200], Peter Zijlstra wrote:
> > On Mon, Jun 24, 2024 at 05:15:15PM +0200, Sebastian Andrzej Siewior wrote:
> > > @@ -9735,18 +9719,28 @@ static int __perf_event_overflow(struct perf_event *event,
> > >  
> > >  		if (regs)
> > >  			pending_id = hash32_ptr((void *)instruction_pointer(regs)) ?: 1;
> > > -		if (!event->pending_sigtrap) {
> > > -			event->pending_sigtrap = pending_id;
> > > +
> > > +		if (!event->pending_work &&
> > > +		    !task_work_add(current, &event->pending_task, TWA_RESUME)) {
> > > +			event->pending_work = pending_id;
> > >  			local_inc(&event->ctx->nr_pending);
> > 
> > task_work_add() isn't NMI safe, at the very least the
> > kasan_record_aux_stack() thing needs to go. But the whole
> > set_notify_resume() thing also needs an audit.
> 
> oh. I just peeked and set_notify_resume() wouldn't survive the audit.
> Would you mind adding task_work_add_nmi() which is task_work_add(,
> TWA_NONE) without kasan_record_aux_stack()? The signal part would need
> to be moved the irq_work() below in the NMI case.

Perhaps add TWA_NMI_CURRENT or somesuch. That would be descriptive and
the value can be used to elide the call to kasan*() and possibly also
include that self-IPI for the NMI-from-user case where we don't have a
return-to-user path.

It can also verify task == current, which ensures people don't try and
use it for other things.