Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20220405163931.1108442-1-jannh@google.com> <CAHmME9pW6heXtPrfCP7J6ODgSc8sotsv6E3dnJoVBaPi+Ph=HA@mail.gmail.com>
 <YqywapDM7NPC/X+E@linutronix.de> <Yq0EcG9GsAkoiF3H@zx2c4.com> <YrAlPKeOowD5qv/B@linutronix.de>
In-Reply-To: <YrAlPKeOowD5qv/B@linutronix.de>
From:   Linus Torvalds <torvalds@linux-foundation.org>
Date:   Mon, 20 Jun 2022 14:00:18 -0500
Message-ID: <CAHk-=wg5jqTdjZrwbSsMsd=NUSi_acBHoLQmRDgNArMvXWj8bw@mail.gmail.com>
Subject: Re: [PATCH] random: Fix signal_pending() usage
To:     Sebastian Siewior <bigeasy@linutronix.de>
Cc:     "Jason A. Donenfeld" <Jason@zx2c4.com>,
        Jann Horn <jannh@google.com>, "Theodore Ts'o" <tytso@mit.edu>,
        LKML <linux-kernel@vger.kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Jun 20, 2022 at 2:44 AM Sebastian Siewior <bigeasy@linutronix.de> wrote:
>
> Based on that, I don't see a problem dropping that signal check
> especially that requests larger than 4KiB are most likely exotic.

Why would we do that?

Anybody who doesn't handle -EINTR is a clown, not a security issue.

Your "6s isn't that bad" is ridiculous, since

 (a) 6 seconds is forever

 (b) there are issues like "oops, we're out of memory, you got a
signal because root is trying to kill your annoying stupid program
using top"

and the fact is, anybody who asks for more than a few kilo-*bits* from
the kernel is already doing something questionable to begin with, and
there is no reason to bend over backwards and try to make such a crazy
use suddenly act differently from ALL OTHER character devices.

Handling signals is the *default* behavior. It is only regular files
where that doesn't happen. This is not a regular file, and the "it's
about security" is not an argument.

As mentioned, expecting an uninterruptible read is not "security". It's garbage.

              Linus