Define the FFI approach - #3139
Conversation
|
The created documentation from the pull request is available at: docu-html |
| The specific Assumptions of Use relevant only for the users of a specific module are documented in the module's safety manual. | ||
| That means that the platform safety manual always has to be read together with all its modules safety manuals. | ||
|
|
||
| A platform module may contain both ASIL-relevant and QM components. The |
There was a problem hiding this comment.
I think we should also define what that means for the verification of the component. What is the difference between asil relevant components and qm components in respect to our safety process
There was a problem hiding this comment.
Good overview here, what is required for QM and for ASIL
https://eclipse-score.github.io/score/pr-3139/platform_management_plan/software_verification.html#id1
There was a problem hiding this comment.
added a reference to ASIL and QM Verification methods
| The specific Assumptions of Use relevant only for the users of a specific module are documented in the module's safety manual. | ||
| That means that the platform safety manual always has to be read together with all its modules safety manuals. | ||
|
|
||
| A platform module may contain both ASIL-relevant and QM components. The |
There was a problem hiding this comment.
Good overview here, what is required for QM and for ASIL
https://eclipse-score.github.io/score/pr-3139/platform_management_plan/software_verification.html#id1
| integrated into a target hardware which includes safety mechanisms which cover hardware related errors. | ||
|
|
||
| The platform allows ASIL and QM components to coexist within the same module. | ||
| Where components executing in an ASIL context depend on functionality that has |
There was a problem hiding this comment.
As Assumed here, https://eclipse-score.github.io/score/pr-3139/platform_management_plan/software_verification.html#id1
there is already a requirement, if the concept requires components used as part of safety context, is must be ASIL, can directly derived from that stakeholder requirement
There was a problem hiding this comment.
this is not exactly a new requirement but an attempt to clarify how to apply that requirement in the context of a module that may contain components with different ASIL levels all while maintaining freedom from interference
There was a problem hiding this comment.
Yes, then it is more a safety concept to apply FFI on modules, which are QM, but also to be used from safety relevant component. The Safety Concept means, that the part for interacting with QM part shall be developed according to the required ASIL and run in the same context, etc.
There was a problem hiding this comment.
yes that was my intention, do you think the current update convey that or do you have a specific recommendation
There was a problem hiding this comment.
RC would be to explicetely mention, that it is a safety concept, and then explain it in a short as to split component in QM part and additional Safety Part, which allocates the safety requirements as you mentioned
aschemmel-tech
left a comment
There was a problem hiding this comment.
Is this the only change within score repository? Please also link to changes in other repositories if planned any?
| safety lifecycle and shall be developed, verified, and released according to | ||
| the corresponding ASIL process. Components that are not intended to execute | ||
| within an ASIL context may remain QM, provided that freedom from interference | ||
| with the safety context is justified in the respective module safety manual. |
There was a problem hiding this comment.
I would have said FFI has nothing to do with the module safety manual. But you can refer to the platform AoU in https://eclipse-score.github.io/score/main/requirements/platform_assumptions/index.html#aou_req__platform__process_isolation
There was a problem hiding this comment.
you're right perhaps the feature DFA is more appropriate justification for FFI.
https://eclipse-score.github.io/process_description//main/process_areas/safety_analysis/safety_analysis_workproducts.html#wp__feature_dfa
There was a problem hiding this comment.
Update speaks of a "is justified in respective feature DFA". I think it is not clear which feature DFA this is. I would propose to write instead "is justified in the feature DFAs of the ASIL rated features using the QM component"
aschemmel-tech
left a comment
There was a problem hiding this comment.
Inline comments and need to get rid of the "lifecycle" changes which are not relevant here.
| as a whole. Components that are intended to execute within an ASIL context are in scope of the | ||
| applicable ASIL safety lifecycle and shall be developed, verified, and released according to | ||
| the corresponding ASIL process, see :need:`stkh_req__dependability__no_mixed_asil`. | ||
| Components that are not intended to executewithin an ASIL context may remain QM, |
There was a problem hiding this comment.
typo "executewithin"
| services, the frontend forms part of the ASIL execution context and therefore | ||
| must fulfil additional safety requirements derived from the need to maintain | ||
| freedom from interference. In contrast, the ``datarouter`` daemon executes as a | ||
| separate process and may remain QM, provided that the required freedom from |
There was a problem hiding this comment.
I would remove the half-sentence "provided that the required freedom from provided that the required freedom from" as it was already stated above. But if you want to keep, please align with the proposed change above.
| safety requirements to be allocated to components that otherwise only provide | ||
| QM functionality. Such requirements exist to preserve the integrity of the | ||
| safety context and are applied in addition to the component's functional | ||
| requirements. |
There was a problem hiding this comment.
I would add here the clarification: "All the component's functional requirements will be promoted to ASIL level to ensure the application of the safety lifecycle." - because this is also what is done in the lifecycle example (at least this is what I understood during your presentation of the PR).
The PR defines our approach to handling modules that may contain components that are executed from within safe context in order to maintain FFI
#1122