On Mon Mar 25 19, Doug Fraser wrote:
I captured the following out of order. You added line is the first emitted at the time of the fault.
/var/log/messages:Mar 25 13:28:27 C05BUB0000A000004124 kern.info kernel: [ 1461.843825] tpm tpm0: tpm_tis_send_data: TPM_STS_DATA_EXPECT == 0: locality: 0 status: ff access: 81
Yes, that was what was being seen before with the registers. The
access register says it has a valid status and it is saying the
locality is not active. I've been dragged into another issue, so I
haven't been able to deal with backporting the patch to 4.14. It
should be straightforward, the only difference that I think needs to
be dealt with is the fact the release_locality has a void return
instead of int. I will hopefully get to it in the next couple of days.
/var/log/messages:Mar 25 13:28:27 C05BUB0000A000004124 kern.err kernel: [ 1461.858321] tpm tpm0: tpm_try_transmit: tpm_send: error -5
/var/log/messages:Mar 25 13:28:27 C05BUB0000A000004124 kern.err kernel: [ 1461.865369] tpm tpm0: A TPM error (357) occurred flushing context
Note also that the 4.18 TPM pull into my 4.14 kernel continues to perform well.