Re: [PATCH] proc: Disable /proc/$pid/wchan

From: Mark Rutland <mark.rutland@arm.com>
To: Kees Cook <keescook@chromium.org>
Cc: "Vito Caputo" <vcaputo@pengaru.com>,
	"Jann Horn" <jannh@google.com>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Josh Poimboeuf" <jpoimboe@redhat.com>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Borislav Petkov" <bp@alien8.de>,
	"H. Peter Anvin" <hpa@zytor.com>, "Jens Axboe" <axboe@kernel.dk>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Stefan Metzmacher" <metze@samba.org>,
	"Andy Lutomirski" <luto@kernel.org>,
	"Lai Jiangshan" <laijs@linux.alibaba.com>,
	"Christian Brauner" <christian.brauner@ubuntu.com>,
	"Andrew Morton" <akpm@linux-foundation.org>,
	"Kenta.Tada@sony.com" <Kenta.Tada@sony.com>,
	"Daniel Bristot de Oliveira" <bristot@redhat.com>,
	"Michael Weiß" <michael.weiss@aisec.fraunhofer.de>,
	"Anand K Mistry" <amistry@google.com>,
	"Alexey Gladkov" <legion@kernel.org>,
	"Michal Hocko" <mhocko@suse.com>, "Helge Deller" <deller@gmx.de>,
	"Dave Hansen" <dave.hansen@linux.intel.com>,
	"Andrea Righi" <andrea.righi@canonical.com>,
	"Ohhoon Kwon" <ohoono.kwon@samsung.com>,
	"Kalesh Singh" <kaleshsingh@google.com>,
	"YiFei Zhu" <yifeifz2@illinois.edu>,
	"Eric W. Biederman" <ebiederm@xmission.com>,
	linux-kernel@vger.kernel.org, x86@kernel.org,
	linux-fsdevel@vger.kernel.org, linux-hardening@vger.kernel.org
Subject: Re: [PATCH] proc: Disable /proc/$pid/wchan
Date: Mon, 27 Sep 2021 10:03:51 +0100	[thread overview]
Message-ID: <20210927090337.GB1131@C02TD0UTHF1T.local> (raw)
In-Reply-To: <202109240716.A0792BE46@keescook>

On Fri, Sep 24, 2021 at 07:26:22AM -0700, Kees Cook wrote:
> On Fri, Sep 24, 2021 at 02:54:24PM +0100, Mark Rutland wrote:
> > On Thu, Sep 23, 2021 at 06:16:16PM -0700, Kees Cook wrote:
> > > On Thu, Sep 23, 2021 at 05:22:30PM -0700, Vito Caputo wrote:
> > > > Instead of unwinding stacks maybe the kernel should be sticking an
> > > > entrypoint address in the current task struct for get_wchan() to
> > > > access, whenever userspace enters the kernel?
> > > 
> > > wchan is supposed to show where the kernel is at the instant the
> > > get_wchan() happens. (i.e. recording it at syscall entry would just
> > > always show syscall entry.)
> > 
> > It's supposed to show where a blocked task is blocked; the "wait
> > channel".
> > 
> > I'd wanted to remove get_wchan since it requires cross-task stack
> > walking, which is generally painful.
> 
> Right -- this is the "fragile" part I'm worried about.
> 
> > We could instead have the scheduler entrypoints snapshot their caller
> > into a field in task_struct. If there are sufficiently few callers, that
> > could be an inline wrapper that passes a __func__ string. Otherwise, we
> > still need to symbolize.
> 
> Hmm. Does PREEMPT break this?

Within the core scheduler functions interrupts should be disabled, and
as long as we only update task_struct there we shouldn't have a race.

> Can we actually use __builtin_return_address(0) in __schedule?

We'd need to do this in a few entry points above __schedule, since the
currently get_wchan walks until !in_sched_functions(). It should be
possible, though we might need to make sure those the nexus points
aren't inlined.

Thanks,
Mark.