On 29/08/23 08:35, Jan Beulich wrote:
> On 28.08.2023 15:19, Simone Ballarin wrote:
>> --- a/automation/eclair_analysis/ECLAIR/deviations.ecl
>> +++ b/automation/eclair_analysis/ECLAIR/deviations.ecl
>> @@ -80,6 +80,7 @@ inline functions."
>>
>> -doc_begin="This header file is autogenerated or empty, therefore it poses no
>> risk if included more than once."
>
> While unrelated to, the change at hand, I still have a question on this:
> How come it is deemed universally safe to multi-include generated headers.
> I would have said that whether that's safe depends on the nature of the
> generated code in the header. Only truly empty ones are uniformly safe to
> include any number of times.
Yes, I agree with you. The mere fact that a file is auto-generated does
not imply anything, moreover, this deviation is not even reported in
rule.rst. In the next series, I'll drop it.
>
> Jan
>
>> +-config=MC3R1.D4.10,reports+={safe, "first_area(text(^/\\* empty \\*/$, begin-1))"}
>> -file_tag+={empty_header, "^xen/arch/arm/efi/runtime\\.h$"}
>> -file_tag+={autogen_headers, "^xen/include/xen/compile\\.h$||^xen/include/generated/autoconf.h$||^xen/include/xen/hypercall-defs.h$"}
>> -config=MC3R1.D4.10,reports+={safe, "all_area(all_loc(file(empty_header||autogen_headers)))"}
>
--
Simone Ballarin, M.Sc.
Field Application Engineer, BUGSENG (https://bugseng.com)