Re: [PATCH 3/3] misra: deviate MISRA C Rule 5.5 for ARM SMMU fault handlers

[Nicola Vetrini <nicola.vetrini@xxxxxxxxxxx>]

On 2026-05-25 14:05, Dmytro Prokopchuk1 wrote:
> MISRA C Rule 5.5 states: "Identifiers shall be distinct from macro
> names".
>
> Update ECLAIR configuration to deviate intentional identifier and
> macro name clashes by specifying the macros that should be ignored.
> This includes 'arm_smmu_global_fault' and 'arm_smmu_context_fault',
> where the original ARM SMMU handler returns irqreturn_t, while the
> same-named macro redirects later uses to a Xen wrapper with the
> void interrupt-handler signature expected by Xen 'request_irq()'.
>
> Update deviations.rst and rules.rst accordingly.
>
> Signed-off-by: Dmytro Prokopchuk <dmytro_prokopchuk1@xxxxxxxx>
> ---
>  automation/eclair_analysis/ECLAIR/deviations.ecl | 6 ++++++
>  docs/misra/deviations.rst                        | 7 +++++++
>  docs/misra/rules.rst                             | 4 ++++
>  3 files changed, 17 insertions(+)
>
> diff --git a/automation/eclair_analysis/ECLAIR/deviations.ecl 
> b/automation/eclair_analysis/ECLAIR/deviations.ecl
> index 7aa64796ba..419a23ce8e 100644
> --- a/automation/eclair_analysis/ECLAIR/deviations.ecl
> +++ b/automation/eclair_analysis/ECLAIR/deviations.ecl
> @@ -144,6 +144,12 @@ ARM SMMU driver calls by reordering the handler 
> and flags arguments."
>  
> -config=MC3A2.R5.5,ignored_macros+="name(request_irq)&&loc(file(^xen/drivers/passthrough/arm/smmu\\.c$))"
>  -doc_end
>
> +-doc_begin="Clashes between ARM SMMU fault handler functions and macro 
> names in 'xen/drivers/passthrough/arm/smmu.c' are deliberate.
> +The original ARM SMMU handler returns irqreturn_t, while the macro 
> redirects later uses to a Xen wrapper with the
> +void interrupt-handler signature expected by Xen 'request_irq()'."
> +-config=MC3A2.R5.5,ignored_macros+="name(arm_smmu_global_fault||arm_smmu_context_fault)&&loc(file(^xen/drivers/passthrough/arm/smmu\\.c$))"
> +-doc_end
> +
>  -doc_begin="The type \"ret_t\" is deliberately defined multiple times,
>  depending on the guest."
>  
> -config=MC3A2.R5.6,reports+={deliberate,"any_area(any_loc(text(^.*ret_t.*$)))"}
> diff --git a/docs/misra/deviations.rst b/docs/misra/deviations.rst
> index 01f141b73f..15c88b82f4 100644
> --- a/docs/misra/deviations.rst
> +++ b/docs/misra/deviations.rst
> @@ -192,6 +192,13 @@ Deviations related to MISRA C:2012 Rules:
>         reordering the handler and flags arguments.
>       - ECLAIR has been configured to ignore this macro.
>
> +   * - R5.5
> +     - Clashes between the ARM SMMU fault handler functions and macro 
> names are
> +       intentional. The original ARM SMMU handler returns irqreturn_t, 
> while
> +       the macro redirects later uses to a Xen wrapper with the void
> +       interrupt-handler signature expected by Xen 'request_irq()'.
> +     - ECLAIR has been configured to ignore these macros.
> +
>     * - R5.6
>       - The type ret_t is deliberately defined multiple times depending 
> on the
>         type of guest to service.
> diff --git a/docs/misra/rules.rst b/docs/misra/rules.rst
> index 097c20432e..88d2c5fa7c 100644
> --- a/docs/misra/rules.rst
> +++ b/docs/misra/rules.rst
> @@ -230,6 +230,10 @@ maintainers if you want to suggest a change.
>         the macro is a local compatibility wrapper that adapts ARM SMMU 
> driver
>         call to Xen's request_irq() argument order.
>
> +       Clashes between ARM SMMU fault handler functions and macro 
> names are
> +       allowed because the macros redirect ARM SMMU driver uses to a 
> Xen interrupt
> +       wrapper with the handler signature expected by request_irq().
> +
>     * - `Rule 5.6 
> <https://gitlab.com/MISRA/MISRA-C/MISRA-C-2012/Example-Suite/-/blob/master/R_05_06.c>`_
>       - Required
>       - A typedef name shall be a unique identifier
Seeing the growing number of exceptions for R5.5 and 5.6 for one reason 
or another in deviations.ecl, I wonder whether a more structured 
approach to the configuration would be better (e.g., a JSON file that is 
then processed just before the build to produce an ECL file fragment 
evaluated in the the right place). The change itself is fine, but I 
think this would tidy up a bit the file from these targeted exceptions.
--
Nicola Vetrini, B.Sc.
Software Engineer
BUGSENG (https://bugseng.com)
LinkedIn: https://www.linkedin.com/in/nicola-vetrini-a42471253