Re: TPM operation times out (very rarely)

["Jarkko Sakkinen" <jarkko@kernel.org>]

On Fri Jan 31, 2025 at 7:28 PM EET, Michal Suchánek wrote:
> On Fri, Jan 31, 2025 at 07:12:06PM +0200, Jarkko Sakkinen wrote:
> > On Fri Jan 31, 2025 at 3:02 PM EET, Michal Suchánek wrote:
> > > It looks like the timeout_b is used exclusively as the ready timeout *),
> > > with various sources of the value depending on chip type.
> > >
> > > Then increasing it should not cause any problem other than the kernel
> > > waiting longer when the TPM chip is really stuck.
> > >
> > > * There is one instance of use of timeout_b for TPM_STS_VALID in
> > > st33zp24_pm_resume.
> > 
> > Possible for you to give a shot for patch and try it out for a while?
> > I'm fine with 2x, or even 4x in this case.
>
> I will see what I can do. It will definitely take a while.
>
> How would you like to multiply it?
>
> At the sime the timeout_b is assigned, or at the time it's used?
>
> Any specific patch that you have in mind?

I'll think about this a bit and send a patch with RFC tag. Might take
to late next week.

BR, Jarkko