Re: [tpmdd-devel] [PATCH RFC 2/4] tpm: validate TPM 2.0 commands

[James Bottomley <jejb@linux.vnet.ibm.com>]

On Wed, 2017-01-04 at 13:04 -0500, Stefan Berger wrote:
> Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com> wrote on 01/02/2017
> 08:22:08 AM:
> 
> > --- a/drivers/char/tpm/tpm2-cmd.c
> > +++ b/drivers/char/tpm/tpm2-cmd.c
> > @@ -943,7 +943,9 @@ EXPORT_SYMBOL_GPL(tpm2_probe);
> >   */
> >  int tpm2_auto_startup(struct tpm_chip *chip)
> >  {
> > +   u32 nr_commands;
> >     int rc;
> > +   int i;
> > 
> >     rc = tpm_get_timeouts(chip);
> >     if (rc)
> > @@ -967,8 +969,49 @@ int tpm2_auto_startup(struct tpm_chip *chip)
> >        }
> >     }
> > 
> > +   rc = tpm2_get_tpm_pt(chip, TPM_PT_TOTAL_COMMANDS, &nr_commands,
> NULL);
> > +   if (rc)
> > +      return rc;
> > +
> > +   chip->cc_attrs_tbl = devm_kzalloc(&chip->dev, 4 * nr_commands,
> > +                 GFP_KERNEL);
> 
> For some reason this devm_kzalloc bombs for the vtpm proxy driver. 
> The only reason I could come up with is that it's being called before
> tpm_add_char_device() has been called.

No, it should be sufficient that chip->dev be initialized (which it is
in tpm_chip_alloc()).  What's the error you're getting?

It does look like the intention was to have non-devm with
tpm_chip_alloc() and devm with tpmm_chip_alloc(), but devm_kzalloc
should just work regardless because it's tied to the device model.

James