svn commit: r342085 - head/sys/dev/tpm

[Marcin Wojtas]

Hi John,

pt., 14 gru 2018 o 18:57 John Baldwin <jhb at freebsd.org> napisał(a):
>
> On 12/14/18 9:43 AM, Marcin Wojtas wrote:
> > Author: mw
> > Date: Fri Dec 14 17:43:35 2018
> > New Revision: 342085
> > URL: https://svnweb.freebsd.org/changeset/base/342085
> >
> > Log:
> >   Fix TPM driver compilation from r342084
> >
> >   Include recent ACPI_ID_PROBE API change.
> >
> > Modified:
> >   head/sys/dev/tpm/tpm_crb.c
> >   head/sys/dev/tpm/tpm_tis.c
> >
> > Modified: head/sys/dev/tpm/tpm_crb.c
> > ==============================================================================
> > --- head/sys/dev/tpm/tpm_crb.c        Fri Dec 14 16:14:36 2018        (r342084)
> > +++ head/sys/dev/tpm/tpm_crb.c        Fri Dec 14 17:43:35 2018        (r342085)
> > @@ -107,7 +107,7 @@ tpmcrb_acpi_probe(device_t dev)
> >       int rid = 0;
> >       uint32_t caps;
> >
> > -     if (ACPI_ID_PROBE(device_get_parent(dev), dev, tpmcrb_ids) == NULL)
> > +     if (ACPI_ID_PROBE(device_get_parent(dev), dev, tpmcrb_ids, NULL) == NULL)
> >               return (ENXIO);
>
> I think it doesn't return a pointer anymore either.  I think you want to use
> its return value as the return value from probe to honor the relative priority
> of CID vs HID matches, so something like:
>
>      int error;
>
>      error = ACPI_ID_PROBE(...);
>      if (error > 0)
>          return (error);
>
>      ...
>
>      return (error);
>
> Instead of 'return (BUS_GENERIC_PROBE)' or some such.
>

Thank you for pointing this, I addressed your remark in r342104.

Best regards,
Marcin