From: Vinod Subject: Re: [PATCH 3/3] hwrng: msm - Add support for prng v2 Date: Thu, 21 Jun 2018 09:47:51 +0530 Message-ID: <20180621041751.GC27187@vkoul-mobl> References: <20180618141259.23141-1-vkoul@kernel.org> <20180618141259.23141-4-vkoul@kernel.org> <20180619142853.wgi5easw4zv6ttrb@gondor.apana.org.au> <20180620053204.GG13316@vkoul-mobl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Herbert Xu , "open list:HARDWARE RANDOM NUMBER GENERATOR CORE" , open list , Matt Mackall , Arnd Bergmann , Greg Kroah-Hartman , linux-arm-msm@vger.kernel.org To: PrasannaKumar Muralidharan Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-crypto.vger.kernel.org Hi PrasannaKumar, On 20-06-18, 23:15, PrasannaKumar Muralidharan wrote: > Hi Vinod, > > On 20 June 2018 at 11:02, Vinod wrote: > > On 19-06-18, 22:28, Herbert Xu wrote: > >> On Mon, Jun 18, 2018 at 07:42:59PM +0530, Vinod Koul wrote: > >> > Qcom 8996 and later chips support prng v2 where we need to only > >> > implement .read callback for hwrng. > >> > > >> > Add a new table for v2 which supports this and get version required for > >> > driver data. > >> > > >> > Signed-off-by: Vinod Koul > >> > >> Is this really a pseudo-RNG? If so it needs to be moved over to > >> the algif_rng interface. > >> > >> hwrng is for true hardware RNGs only, because it may be directly > >> fed into /dev/random. > > > > I am trying to find how how much random output this hardware generates. > > > > Meanwhile, can you please point out examples/Documentation for algif_rng > > and if any test tools for this? > > As Herbert suggested exynos rng is a good example. It was wrongly > using hwrng framework then switched to using crypto framework. > > I have a patch for msm rng using crypto framework, only compile tested > though. You can find the patch at > https://pastebin.ubuntu.com/p/44r6BJgBkN/. I am having a hard time > sending patches via git send-email form a gmail account so I have put > it in pastebin. Feel free to use it if that is useful. Ah, I already started based on msm downstream implementation (they had one using crypto APIs too). Ideally you should have posted it and tested it. Thanks -- ~Vinod