Re: arc4random - are you sure we want these?

From: Yann Droneaud <ydroneaud@opteya.com>
To: Florian Weimer <fweimer@redhat.com>,
	"Jason A. Donenfeld via Libc-alpha" <libc-alpha@sourceware.org>
Cc: "Jason A. Donenfeld" <Jason@zx2c4.com>,
	Michael@phoronix.com, linux-crypto@vger.kernel.org,
	jann@thejh.net
Subject: Re: arc4random - are you sure we want these?
Date: Tue, 26 Jul 2022 16:35:40 +0200	[thread overview]
Message-ID: <df5c0926-22f8-af79-4972-ff41488ef945@opteya.com> (raw)
In-Reply-To: <87v8rlqscj.fsf@oldenburg.str.redhat.com>

Hi,

Le 25/07/2022 à 14:39, Florian Weimer a écrit :
> * Jason A. Donenfeld via Libc-alpha:
>>> The performance numbers suggest that we benefit from buffering in user
>>> space.
>> The question is whether it's safe and advisable to buffer this way in
>> userspace. Does userspace have the right information now of when to
>> discard the buffer and get a new one? I suspect it does not.
> Not completely, no, but we can cover many cases.  I do not currently see
> a way around that if we want to promote arc4random_uniform(limit) as a
> replacement for random() % limit.

+1

That the reason I've reviewed the implementation positively: for me 
arc4random is not about generating secret keys but small integers.
I want to be able to divert developers from
     srand(time(NULL))
     identifier = rand() % 33
to
     identifier = arc4random_uniform(33)

Safe, fast, and reasonably secure.

Regards.

-- 
Yann Droneaud
OPTEYA