From: Herbert Xu Subject: Re: [PATCH 3/5] hwrng: core - Do not register device opportunistically Date: Fri, 26 Dec 2014 12:00:56 +1100 Message-ID: <20141226010056.GB20702@gondor.apana.org.au> References: <20141223053909.GA28001@gondor.apana.org.au> <87mw6egd22.fsf@rustcorp.com.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Fengguang Wu , LKP , linux-kernel@vger.kernel.org, Linux Crypto Mailing List , Amos Kong , m@bues.ch, mpm@selenic.com, amit.shah@redhat.com To: Rusty Russell Return-path: Received: from helcar.apana.org.au ([209.40.204.226]:47952 "EHLO helcar.apana.org.au" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750725AbaLZBBF (ORCPT ); Thu, 25 Dec 2014 20:01:05 -0500 Content-Disposition: inline In-Reply-To: <87mw6egd22.fsf@rustcorp.com.au> Sender: linux-crypto-owner@vger.kernel.org List-ID: On Wed, Dec 24, 2014 at 09:59:41AM +1030, Rusty Russell wrote: > > Does this break userspace by creating a device which will just return > -ENODEV on read? Sure, callers *should* handle it... If somebody complains we could easily fix this by making open fail. In fact, if such applications exist then they're likely to want to hold onto a reference to the RNG after then open... Cheers, -- Email: Herbert Xu Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt