On 9/4/2023 7:20 PM, Nhi Pham wrote:
> On 9/2/2023 3:43 AM, Oliver Smith-Denny wrote:
>> On 8/31/2023 1:20 AM, Nhi Pham via groups.io wrote:
>>
>> If I am understanding this correctly, this is only an issue when
>> HOBs are created in StMM, i.e. not from HOBs that are passed in. Is this
>> correct?
> Yes, the issue only occurs when HOB are created in StandaloneMM by the
> HOB library instance
> StandaloneMmPkg/Library/StandaloneMmHobLib/StandaloneMmHobLib.inf
>>
>> If so, is HOB creation in StMM and supported use case? The only instance
> I think it is intended to work as the CreateHob() function is implemented.

Well, that may just be a copy/paste sort of thing :).

>> a quick search turns up is the ARM StMM Core entry, where some
>> information from TF-A is converted to HOB format. Do we have any other
>> use cases (and curious more on this use case). My thought process would
>> be that StMM would not create any HOBs. Depending on FW configuration,
>> it may receive HOBs from PEI.
>
> I have a use case when enabling the UEFI Variable driver running in
> StandaloneMM. Instead of using the PCDs, the in-memory NVRAM region is
> allocated **dynamically** at boot time in the StMM secure memory. Then,
> they will be passed into the gVariableFlashInfoHobGuid for being
> consumed by other variable MM drivers.
>

I do believe that per the PI spec, we should have HOB producer and HOB
consumer phases, where in this case PEI (if it was the launching entity
for StMM) is the HOB producer and StMM is the HOB producer. This is the
same pattern the PI spec details for PEI and DXE, where DXE is not
intended to create new HOBs, but just to consume information from the
previous phase.

As I mentioned, there are other interfaces for passing information
within a phase, such as protocols, dynamic PCDs, variables, etc. that
are built for this application. I think it is useful to adhere to the
model for HOBs (which are hand off blocks, one phase handing information
to another phase) and that we will create more issues if we rely on
HOB consumer phases producing HOBs.

My proposal would be to remove the HOB creation code from StMM
completely. I believe in your use case that you are describing a dynamic
PCD or a protocol could work to pass the information.

If we are saying that prior to your patch that HOB creation in StMM was
completely broken, anyway, it seems that folks were not relying on this
code?

Thanks,
Oliver