Re: [PATCH 2/2 v3] lib/vsprintf: Initialize vsprintf's pointer hash once the random core is ready.

From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
To: "Jason A. Donenfeld" <Jason@zx2c4.com>
Cc: linux-kernel@vger.kernel.org,
	Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
	John Ogness <john.ogness@linutronix.de>,
	Mike Galbraith <efault@gmx.de>, Petr Mladek <pmladek@suse.com>,
	Rasmus Villemoes <linux@rasmusvillemoes.dk>,
	Sergey Senozhatsky <senozhatsky@chromium.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Theodore Ts'o <tytso@mit.edu>,
	Thomas Gleixner <tglx@linutronix.de>
Subject: Re: [PATCH 2/2 v3] lib/vsprintf: Initialize vsprintf's pointer hash once the random core is ready.
Date: Mon, 1 Aug 2022 14:41:06 +0200	[thread overview]
Message-ID: <YufJ4vQyAMkU9HBx@linutronix.de> (raw)
In-Reply-To: <YufHN7fF0dROMETj@zx2c4.com>

On 2022-08-01 14:36:48 [+0200], Jason A. Donenfeld wrote:
> > --- a/lib/vsprintf.c
> > +++ b/lib/vsprintf.c
> > @@ -751,31 +751,39 @@ static int __init debug_boot_weak_hash_e
> >  early_param("debug_boot_weak_hash", debug_boot_weak_hash_enable);
> >  
> >  static bool filled_random_ptr_key;
> > +static siphash_key_t ptr_key __read_mostly;
> > +static void fill_ptr_key_workfn(struct work_struct *work);
> > +static DECLARE_DELAYED_WORK(fill_ptr_key_work, fill_ptr_key_workfn);
> > +
> > +static void fill_ptr_key_workfn(struct work_struct *work)
> > +{
> > +	if (!rng_is_initialized()) {
> > +		queue_delayed_work(system_unbound_wq, &fill_ptr_key_work, HZ  * 2);
> > +		return;
> > +	}
> 
> This seems kind of crazy and over-complex. You only need this on RT,
> right? Because of the raw lock issue? So just schedule it for the right
> time on RT and not elsewhere.

Then we have two code paths for RT and !RT and I would to avoid that
especially if it ends up as a user visible change.
And then there is lockdep which might yell on !RT if you acquire a
spinlock_t from context which won't work on RT.

> I'll send a more basic patch.

oki.

> Jason

Sebastian