On Tue, 2020-04-21 at 20:56 +0000, Benoit HOUYERE wrote: >> Issue occurs on several legacy models and corrected on latest TPM >> versions. Several corrections are possible. Omar's proposal is quite >> simple, short and efficient. Penalty time is only condition check but >> for all TPM_status access. >> >> Other possibility is to check status register validity (bit 5 is >> always at 0) at the first reading and modify wait_for_stat function >> (already inserted for I2C patch). >Benoit, thank you for the explanation. >It sounds like by "already inserted for I2C patch", you mean that this proposed solution is part of the i2c patch set. If that is the case, to simplify backporting, the fix should be the first patch in the i2c patch set. >Mimi Ok, I will check with Amir before.
