On Wed, Jun 15, 2022 at 03:52:14PM -0700, Jerry Snitselaar wrote:
> On Wed, Jun 15, 2022 at 12:55 PM Jarkko Sakkinen <jarkko@xxxxxxxxxx> wrote:
> >
> > On Mon, Jun 13, 2022 at 07:47:40AM -0700, Jerry Snitselaar wrote:
> > > This came to my attention this morning, and I was able to do a quick
> > > test and reproduce in a vm locally, using both the crb and tis
> > > emulation. 5.18 boots up fine.
> > >
> > > The only thing I see that could impact both crb and tis in the tpm
> > > changes would be:
> > >
> > > af402ee3c045 tpm: Add field upgrade mode support for Infineon TPM2
> > > modules | 2022-05-23 | (Stefan Mahnke-Hartmann)
> > > e57b2523bd37 tpm: Fix buffer access in tpm2_get_tpm_pt() | 2022-05-23
> > > | (Stefan Mahnke-Hartmann)
> > >
> > > I'm going to poke around some more and see if I can figure out what is
> > > going on, but wanted to give you a heads up.
> > >
> > > Regards,
> > > Jerry
> > >
> >
> > OK, thanks for spotting this.
> >
> > BR, Jarkko
> >
>
> Whatever it was it wasn't the tpm code, and it looks like it has
> disappeared with v5.19-rc2.
>
> Regards,
> Jerry
>
It looks like the problem was fixed by:
228432551bd8 ("virtio-rng: make device ready before making request")
If I revert this on 5.19-rc2, I see the problem again. One of the
commits it lists as fixing is:
8b4ec69d7e09 ("virtio: harden vring IRQ")
If I go to a build prior to that it comes up fine. So looks like
it was introduced by 8b4ec69d7e09 and fixed with 228432551bd8.
Regards,
Jerry
