[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YmaB8TifJ7EWW1Xi@zx2c4.com>
Date: Mon, 25 Apr 2022 13:11:45 +0200
From: "Jason A. Donenfeld" <Jason@...c4.com>
To: Guenter Roeck <linux@...ck-us.net>
Cc: LKML <linux-kernel@...r.kernel.org>,
Linux Crypto Mailing List <linux-crypto@...r.kernel.org>,
linux-arch <linux-arch@...r.kernel.org>,
Dinh Nguyen <dinguyen@...nel.org>,
Nick Hu <nickhu@...estech.com>,
Max Filippov <jcmvbkbc@...il.com>,
Palmer Dabbelt <palmer@...belt.com>,
"David S . Miller" <davem@...emloft.net>,
Yoshinori Sato <ysato@...rs.sourceforge.jp>,
Michal Simek <monstr@...str.eu>,
Borislav Petkov <bp@...en8.de>, Guo Ren <guoren@...nel.org>,
Geert Uytterhoeven <geert@...ux-m68k.org>,
Joshua Kinard <kumba@...too.org>,
David Laight <David.Laight@...lab.com>,
Dominik Brodowski <linux@...inikbrodowski.net>,
Eric Biggers <ebiggers@...gle.com>,
Ard Biesheuvel <ardb@...nel.org>,
Arnd Bergmann <arnd@...db.de>,
Thomas Gleixner <tglx@...utronix.de>,
Andy Lutomirski <luto@...nel.org>,
Kees Cook <keescook@...omium.org>,
Lennart Poettering <mzxreary@...inter.de>,
Konstantin Ryabitsev <konstantin@...uxfoundation.org>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Theodore Ts'o <tytso@....edu>
Subject: Re: [PATCH v1] random: block in /dev/urandom
Hi Guenter,
On Sun, Apr 24, 2022 at 06:54:10PM -0700, Guenter Roeck wrote:
> On 4/24/22 17:12, Jason A. Donenfeld wrote:
> > Hi Guenter,
> >
> > On Sat, Apr 23, 2022 at 07:04:26PM -0700, Guenter Roeck wrote:
> >> I'll run another test tonight.
> >
> > Super, thanks. Looking forward to learning what transpires. Hopefully
> > all pass this time through...
> >
>
> Build results:
> total: 147 pass: 146 fail: 1
> Failed builds:
> m68k:allmodconfig
> Qemu test results:
> total: 489 pass: 489 fail: 0
>
> The failure is inherited from mainline, so all looks good.
That is excellent news! Thanks again for testing.
So what this means is: the rationale for reverting the /dev/random +
/dev/urandom unification has now been fixed. That's some real tangible
progress.
Now, I don't want to rush into trying the unification again too soon. I
think if anything, the lesson from the first attempt wasn't simply, "I
should fix a few of Guenter's test cases," but rather that the problem
is fairly nuanced and will take a lot wider testing and research.
However, the fact that the initial thing, across multiple platforms,
that lead to the revert has been fixed gives me a decent amount of
optimism that at /some point/ down the road, we'll be able to try this
again. One step at a time.
Jason
Powered by blists - more mailing lists