Re: [PATCH] tpm: fix a race condition tpm2_unseal_trusted()

[Jason Gunthorpe]

On Tue, Aug 09, 2016 at 01:36:29PM +0300, Jarkko Sakkinen wrote:

> Functionally my patch should not break anything. I understand the need
> for clean up of locking but why doing this now to make the driver
> non-racy would make clean up later on any harder?

Then rename the functions so they don't confusingly look like the
standard kernel pattern that doesn't lock.

Jason