Messages in this thread | | | Date | Fri, 27 Jul 2012 08:08:28 -0700 | From | "H. Peter Anvin" <> | Subject | Re: [GIT PULL] New TPM driver and features |
| |
On 07/27/2012 07:29 AM, Kent Yoder wrote: > Hi, > > On Thu, Jul 26, 2012 at 02:46:25PM -0700, H. Peter Anvin wrote: >> On 07/26/2012 07:31 AM, Kent Yoder wrote: >>> hw_random: add support for the TPM chip as a hardware RNG source >> >> Could you clarify this? rngd (the user of /dev/hw_random) already >> has support for the TPM... is this fundamentally different (e.g. can >> it coexist with tcsd being in use at the same time)? > > These can coexist at the same time since /dev/tpm0 is only opened once > from user-space (from tcsd) and this hwrng driver comes in through the > kernel. There would be some amount of contention with the tpm, but its > not likely to be noticable - grabbing rng data doesn't require any state > in the tpm, like an auth session or key loading. > > Since the rngd support for the tpm requires trousers, this patch will > be useful in places where you don't want to have to deploy the full TSS, > like embedded and early boot. >
Actually the rngd support for the TPM *conflicts* with trousers... I was looking at adding trousers support when I stumbled over your patch. Your patch is better, because it solves the handover problem (rngd should normally be started as early as possible.)
-hpa
-- H. Peter Anvin, Intel Open Source Technology Center I work for Intel. I don't speak on their behalf.
| |