On Fri, Aug 28, 2020 at 7:18 PM Jarkko Sakkinen
<jarkko.sakkinen@xxxxxxxxxxxxxxx> wrote:
>
> On Thu, Aug 27, 2020 at 11:24:45AM -0400, Jason Andryuk wrote:
> > James Bottomley wrote:
> > >On Wed, 2020-04-15 at 15:45 -0700, Omar Sandoval wrote:
> > >> From: Omar Sandoval <osandov@xxxxxx>
> > >>
> > >> We've encountered a particular model of STMicroelectronics TPM that
> > >> transiently returns a bad value in the status register. This causes
> > >> the kernel to believe that the TPM is ready to receive a command when
> > >> it actually isn't, which in turn causes the send to time out in
> > >> get_burstcount(). In testing, reading the status register one extra
> > >> time convinces the TPM to return a valid value.
> > >
> > >Interesting, I've got a very early upgradeable nuvoton that seems to be
> > >behaving like this.
> > >
> > >> Signed-off-by: Omar Sandoval <osandov@xxxxxx>
> > >> ---
> > >> drivers/char/tpm/tpm_tis_core.c | 12 ++++++++++++
> > >> 1 file changed, 12 insertions(+)
> > >>
> > >> diff --git a/drivers/char/tpm/tpm_tis_core.c
> > >> b/drivers/char/tpm/tpm_tis_core.c
> > >> index 27c6ca031e23..277a21027fc7 100644
> > >> --- a/drivers/char/tpm/tpm_tis_core.c
> > >> +++ b/drivers/char/tpm/tpm_tis_core.c
> > >> @@ -238,6 +238,18 @@ static u8 tpm_tis_status(struct tpm_chip *chip)
> > >> rc = tpm_tis_read8(priv, TPM_STS(priv->locality), &status);
> > >> if (rc < 0)
> > >> return 0;
> > >> + /*
> > >> + * Some STMicroelectronics TPMs have a bug where the status
> > >> register is
> > >> + * sometimes bogus (all 1s) if read immediately after the
> > >> access
> > >> + * register is written to. Bits 0, 1, and 5 are always
> > >> supposed to read
> > >> + * as 0, so this is clearly invalid. Reading the register a
> > >> second time
> > >> + * returns a valid value.
> > >> + */
> > >> + if (unlikely(status == 0xff)) {
> > >> + rc = tpm_tis_read8(priv, TPM_STS(priv->locality),
> > >> &status);
> > >> + if (rc < 0)
> > >> + return 0;
> > >> + }
> > >
> > >You theorize that your case is fixed by the second read, but what if it
> > >isn't and the second read also returns 0xff? Shouldn't we have a line
> > >here saying
> > >
> > >if (unlikely(status == 0xff))
> > > status = 0;
> > >
> > >So if we get a second 0xff we just pretend the thing isn't ready?
> >
> > Thanks for the fix, Omar!
> >
> > I tried the patch and it helps with STM TPM2 issues where commands fail
> > with the kernel reporting:
> > tpm tpm0: Unable to read burstcount
> > tpm tpm0: tpm_try_transmit: send(): error -16
> >
> > My testing was with 5.4, and I'd like to see this CC-ed stable.
> >
> > When trying to diagnose the issue before finding this patch, I found it
> > was timing sensitive. I was seeing failures in the OpenXT installer.
> > The system is basically idle when issuing TPM commands which frequently
> > failed. The same hardware booted into a Fedora Live USB image didn't
> > have any TPM command failures. One notable difference between the two
> > is Fedora is CONFIG_PREEMPT=y and OpenXT is CONFIG_PREEMPT_NONE=y.
> > Switching OpenXT to PREEMPT=y helped some, but there were still some
> > issues with commands failing. The second interesting thing was running tpm
> > commands in OpenXT under trace-cmd let them succeed. I guess that was enough
> > to throw the timing off.
> >
> > Anyway, I'd like to see this patch applied, please.
> >
> > Thanks,
> > Jason
>
> There was v2 sent after this:
>
> https://patchwork.kernel.org/patch/11492125/
Thanks! That one didn't come up in a search for STM on lore.kernel.org.
> Unfortunately it lacks changelog. What was changed between v1 and v2?
> Why v3 has not been sent yet? I see a discussion with no final
> conclusion.
Looks like v2 added James's suggestion with a comment (sorry the
formating is off):
+ /*
+ * The status is somehow still bad. This hasn't been observed in
+ * practice, but clear it just in case so that it doesn't appear
+ * to be ready.
+ */
+ if (unlikely(status == 0xff))
+ status = 0;
But, yes, the decision on the alternate approach is unresolved.
Thanks again,
Jason
