Can't debug FVP_BaseR_AEMv8R in Aarch64 mode using Arm DS

Hi again

The Cortex-R82 FVP, and a license for its use, is provided with Arm DS 2022.1, so you could use that.

Support for some of the more advanced features of the Arm Architecture offered within the AEMs (AEMvA, AEMv8R & AEMv8M) is provided only in the "Platinum" edition of Arm DS (though unfortunately there is no Evaluation version for that).

If you really do need to use the AEMv8R FVP, then a workaround is to set the cpu type (that the debugger will use for e.g. displaying system registers, etc) to Cortex-R82, as follows:

On the Arm DS Command Line, launch the FVP with:
FVP_BaseR_AEMv8R -C cluster0.has_aarch64=1 --iris-server

In the Arm DS IDE, select File > New > Other, Configuration Database > Model Configuration, click "Create New Database" and give it a name e.g. "ExtensionDB" [Next], select Iris, Browse for model running on localhost, and select it [Finish].

In the model.mdf dialog, for each of the 4 cpus, select its line, click Edit, and change its Type from "ARMAEMv8-R_" to "Cortex-R82".

Finally, click on Save, Import, then Debug.

I've only lightly tested this, but it seems to work OK.  If you need full Armv8-R architecture support (as opposed the Cortex-R82 subset), then I'm afraid you'll need to upgrade to Platinum.

Hope this helps

Stephen

Hi again

The Cortex-R82 FVP, and a license for its use, is provided with Arm DS 2022.1, so you could use that.

Support for some of the more advanced features of the Arm Architecture offered within the AEMs (AEMvA, AEMv8R & AEMv8M) is provided only in the "Platinum" edition of Arm DS (though unfortunately there is no Evaluation version for that).

If you really do need to use the AEMv8R FVP, then a workaround is to set the cpu type (that the debugger will use for e.g. displaying system registers, etc) to Cortex-R82, as follows:

On the Arm DS Command Line, launch the FVP with:
FVP_BaseR_AEMv8R -C cluster0.has_aarch64=1 --iris-server

In the Arm DS IDE, select File > New > Other, Configuration Database > Model Configuration, click "Create New Database" and give it a name e.g. "ExtensionDB" [Next], select Iris, Browse for model running on localhost, and select it [Finish].

In the model.mdf dialog, for each of the 4 cpus, select its line, click Edit, and change its Type from "ARMAEMv8-R_" to "Cortex-R82".

Finally, click on Save, Import, then Debug.

I've only lightly tested this, but it seems to work OK.  If you need full Armv8-R architecture support (as opposed the Cortex-R82 subset), then I'm afraid you'll need to upgrade to Platinum.

Hope this helps

Stephen

That was really helpful! Thanks!