Re: [PATCH v3 1/2] xen/device_tree: fix Eclair findings for MISRA C 2012 Rule 20.7

[Luca Fancellu <Luca.Fancellu@xxxxxxx>]

> On 6 Feb 2023, at 14:37, Michal Orzel <michal.orzel@xxxxxxx> wrote:
> 
> Hi Luca,
> 
> On 06/02/2023 15:31, Luca Fancellu wrote:
>> 
>> 
>> Hi Xenia,
>> 
>>> On 3 Feb 2023, at 19:09, Xenia Ragiadakou <burzalodowa@xxxxxxxxx> wrote:
>>> 
>>> Signed-off-by: Xenia Ragiadakou <burzalodowa@xxxxxxxxx>
>>> ---
>>> 
>>> Changes in v3:
>>> - the fixes are based solely to Eclair findings (the tool has been
>>>   adjusted to report only those violations that can result to a bug)
>>> - reflect this in the commit title
>>> 
>>> xen/include/xen/device_tree.h | 14 +++++++-------
>>> 1 file changed, 7 insertions(+), 7 deletions(-)
>>> 
>>> diff --git a/xen/include/xen/device_tree.h b/xen/include/xen/device_tree.h
>>> index a28937d12ae8..7839a199e311 100644
>>> --- a/xen/include/xen/device_tree.h
>>> +++ b/xen/include/xen/device_tree.h
>>> @@ -37,11 +37,11 @@ struct dt_device_match {
>>>    const void *data;
>>> };
>>> 
>>> -#define __DT_MATCH_PATH(p)              .path = p
>>> -#define __DT_MATCH_TYPE(typ)            .type = typ
>>> -#define __DT_MATCH_COMPATIBLE(compat)   .compatible = compat
>>> +#define __DT_MATCH_PATH(p)              .path = (p)
>>> +#define __DT_MATCH_TYPE(typ)            .type = (typ)
>>> +#define __DT_MATCH_COMPATIBLE(compat)   .compatible = (compat)
>>> #define __DT_MATCH_NOT_AVAILABLE()      .not_available = 1
>>> -#define __DT_MATCH_PROP(p)              .prop = p
>>> +#define __DT_MATCH_PROP(p)              .prop = (p)
>>> 
>>> #define DT_MATCH_PATH(p)                { __DT_MATCH_PATH(p) }
>>> #define DT_MATCH_TYPE(typ)              { __DT_MATCH_TYPE(typ) }
>>> @@ -226,13 +226,13 @@ dt_find_interrupt_controller(const struct 
>>> dt_device_match *matches);
>>> #define DT_ROOT_NODE_SIZE_CELLS_DEFAULT 1
>>> 
>>> #define dt_for_each_property_node(dn, pp)                   \
>>> -    for ( pp = dn->properties; pp != NULL; pp = pp->next )
>>> +    for ( pp = (dn)->properties; (pp) != NULL; pp = (pp)->next )
>>> 
>>> #define dt_for_each_device_node(dt, dn)                     \
>>> -    for ( dn = dt; dn != NULL; dn = dn->allnext )
>>> +    for ( dn = dt; (dn) != NULL; dn = (dn)->allnext )
>>> 
>>> #define dt_for_each_child_node(dt, dn)                      \
>>> -    for ( dn = dt->child; dn != NULL; dn = dn->sibling )
>>> +    for ( dn = (dt)->child; (dn) != NULL; dn = (dn)->sibling )
>>> 
>>> /* Helper to read a big number; size is in cells (not bytes) */
>>> static inline u64 dt_read_number(const __be32 *cell, int size)
>>> --
>>> 2.37.2
>>> 
>>> 
>> 
>> While the changes looks sensible to me, I’ve had a look in eclair but I was 
>> unable to find the findings,
>> here what findings I have in the latest report:
>> https://eclairit.com:8443/job/XEN/Target=ARM64,agent=docker1/lastBuild/eclair/packageName.832204620/fileName.1811675806/
> 
> Eclair does not report violations at the definition but rather at the use.
> Check domain_build.c for example to see the reports for 20.7 related to these 
> macros.

Wow yes that’s right, a bit annoying to link the issues! Who knows if there is 
a mode to make it point out the violation at the definition
like cppcheck.

I’ll review this patch later this week if I manage to find the time

> 
> ~Michal