[PATCH 1/2] fwts_acpi_object_eval: add a function to check package elements

[Alex Hung]

On Mon, May 25, 2020 at 2:01 AM ivanhu <ivan.hu at canonical.com> wrote:

>
>
> On 5/22/20 12:47 PM, Alex Hung wrote:
> >
> >
> > On Thu, May 21, 2020 at 9:01 PM ivanhu <ivan.hu at canonical.com
> > <mailto:ivan.hu at canonical.com>> wrote:
> >
> >
> >
> >     On 5/16/20 2:00 AM, Alex Hung wrote:
> >     > This function checks if all elements are integers in a returned
> >     package.
> >     > This is one of common forms returned by ACPI methods.
> >     >
> >     > Signed-off-by: Alex Hung <alex.hung at canonical.com
> >     <mailto:alex.hung at canonical.com>>
> >     > ---
> >     >  src/lib/include/fwts_acpi_object_eval.h |  1 +
> >     >  src/lib/src/fwts_acpi_object_eval.c     | 28
> >     +++++++++++++++++++++++++
> >     >  2 files changed, 29 insertions(+)
> >     >
> >     > diff --git a/src/lib/include/fwts_acpi_object_eval.h
> >     b/src/lib/include/fwts_acpi_object_eval.h
> >     > index 09f3f754..92177857 100644
> >     > --- a/src/lib/include/fwts_acpi_object_eval.h
> >     > +++ b/src/lib/include/fwts_acpi_object_eval.h
> >     > @@ -128,6 +128,7 @@ void
> >     fwts_method_test_buffer_return(fwts_framework *fw, char *name,
> >     ACPI_BUFFER
> >     >  void fwts_method_test_all_reference_package_return(fwts_framework
> >     *fw, char *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     >  void fwts_method_test_integer_reserved_bits_return(fwts_framework
> >     *fw, char *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     >  void fwts_method_test_integer_max_return(fwts_framework *fw, char
> >     *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     > +void fwts_method_test_package_integer_return(fwts_framework *fw,
> >     char *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     >  void fwts_method_test_passed_failed_return(fwts_framework *fw,
> >     char *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     >  void fwts_method_test_polling_return( fwts_framework *fw, char
> >     *name, ACPI_BUFFER *buf, ACPI_OBJECT *obj, void *private);
> >     >
> >     > diff --git a/src/lib/src/fwts_acpi_object_eval.c
> >     b/src/lib/src/fwts_acpi_object_eval.c
> >     > index 1e3575c9..613a6fd3 100644
> >     > --- a/src/lib/src/fwts_acpi_object_eval.c
> >     > +++ b/src/lib/src/fwts_acpi_object_eval.c
> >     > @@ -881,6 +881,34 @@ void fwts_method_test_integer_max_return(
> >     >               fwts_passed(fw, "%s correctly returned an integer.",
> >     name);
> >     >  }
> >     >
> >     > +/*
> >     > + *  fwts_method_test_package_integer_return
> >     > + *   check if all integers in a returned package
> >     > + */
> >     > +void fwts_method_test_package_integer_return(
> >     > +     fwts_framework *fw,
> >     > +     char *name,
> >     > +     ACPI_BUFFER *buf,
> >     > +     ACPI_OBJECT *obj,
> >     > +     void *private)
> >     > +{
> >     > +     char method[4];
> >
> >     Should this be method[5], include the null-terminated?
> >
> >
> > It seems ACPI names aren't exactly strings. I checked acpica code and it
> > uses 4 chars like below:
> >
> /*******************************************************************************
> >  *
> >  * FUNCTION:    AcpiDmDumpName
> >  *
> >  * PARAMETERS:  Name            - 4 character ACPI name
> >  *
> >  * RETURN:      Final length of name
> >  *
> >  * DESCRIPTION: Dump an ACPI name, minus any trailing underscores.
> >  *
> >
>  ******************************************************************************/
> >
> > UINT32
> > AcpiDmDumpName (
> >     UINT32                  Name)
> > {
> >     UINT32                  i;
> >     UINT32                  Length;
> >     char                    NewName[4];
> >
> > // ... skipped
> >
> > In other cases, it can also use 5 chars like "AcpiName" variable in the
> > function "AcpiDbFindNameInNamespace" (although its parameter takes a
> > 4-char string)
> >
> > I think both would work when passing to AML parser.
> >
> >
> OK. but I think we should make sure if those non null-terminated string
> have any potential issue, since they will be passed to the result log
> functions.
>

V2 was sent for reviews.



>
> >
> >
> >     > +     uint32_t *element_size = (uint32_t *) private;
> >     > +
> >     > +     memcpy(method, &name[strlen(name) - 4], 4);
> >     > +
> >     > +     if (fwts_method_check_type(fw, name, buf, ACPI_TYPE_PACKAGE)
> >     != FWTS_OK)
> >     > +             return;
> >     > +
> >     > +     if (fwts_method_package_count_equal(fw, name, method, obj,
> >     *element_size) != FWTS_OK)
> >     > +             return;
> >     > +
> >     > +     if (fwts_method_package_elements_all_type(fw, name, method,
> >     obj, ACPI_TYPE_INTEGER) != FWTS_OK)
> >     > +             return;
> >     > +
> >     > +     fwts_method_passed_sane(fw, name, "package");
> >     > +}
> >     > +
> >     >  /*
> >     >   *  fwts_method_test_passed_failed_return
> >     >   *   check if 0 or 1 (false/true) integer is returned
> >     >
> >
> >     --
> >     fwts-devel mailing list
> >     fwts-devel at lists.ubuntu.com <mailto:fwts-devel at lists.ubuntu.com>
> >     Modify settings or unsubscribe at:
> >     https://lists.ubuntu.com/mailman/listinfo/fwts-devel
> >
> >
> >
> > --
> > Cheers,
> > Alex Hung
>


-- 
Cheers,
Alex Hung
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.ubuntu.com/archives/fwts-devel/attachments/20200525/1d1d5960/attachment.html>