On Tue, Oct 29, 2019 at 07:58:16AM -0700, James Bottomley wrote: > On Tue, 2019-10-29 at 10:42 +0200, Jarkko Sakkinen wrote: > > On Mon, Oct 21, 2019 at 02:39:39PM +0300, Jarkko Sakkinen wrote: > > > On Thu, Oct 17, 2019 at 09:04:40PM +0300, Jarkko Sakkinen wrote: > > > > On Wed, Oct 16, 2019 at 03:10:29PM -0400, James Bottomley wrote: > > > > > On Wed, 2019-10-16 at 19:25 +0300, Jarkko Sakkinen wrote: > > > > > > On Wed, Oct 16, 2019 at 08:34:12AM -0400, James Bottomley > > > > > > wrote: > > > > > > > reversible ciphers are generally frowned upon in random > > > > > > > number > > > > > > > generation, that's why the krng uses chacha20. In general > > > > > > > I think > > > > > > > we shouldn't try to code our own mixing and instead should > > > > > > > get the > > > > > > > krng to do it for us using whatever the algorithm du jour > > > > > > > that the > > > > > > > crypto guys have blessed is. That's why I proposed adding > > > > > > > the TPM > > > > > > > output to the krng as entropy input and then taking the > > > > > > > output of > > > > > > > the krng. > > > > > > > > > > > > It is already registered as hwrng. What else? > > > > > > > > > > It only contributes entropy once at start of OS. > > > > > > > > Ok. > > > > > > > > > > Was the issue that it is only used as seed when the rng is > > > > > > init'd > > > > > > first? I haven't at this point gone to the internals of krng. > > > > > > > > > > Basically it was similar to your xor patch except I got the > > > > > kernel rng > > > > > to do the mixing, so it would use the chacha20 cipher at the > > > > > moment > > > > > until they decide that's unsafe and change it to something > > > > > else: > > > > > > > > > > https://lore.kernel.org/linux-crypto/1570227068.17537.4.camel@H > > > > > ansenPartnership.com/ > > > > > > > > > > It uses add_hwgenerator_randomness() to do the mixing. It also > > > > > has an > > > > > unmixed source so that read of the TPM hwrng device works as > > > > > expected. > > > > > > > > Thinking that could this potentially racy? I.e. between the calls > > > > something else could eat the entropy added? > > > > > > Also, what is wrong just taking one value from krng and mixing > > > it with a value from TPM RNG where needed? That would be non-racy > > > too. > > > > I guess we can move forward with this? > > Sure I suppose; can we can figure out how to get the mixing function du > jour exposed? Maybe it is best to reflect the whole issue in the context of the Sumit's 2nd patch set, which adds ARM TEE support in order to move forward. /Jarkko
