On 8/18/25 23:42, Peter Maydell wrote:
Looking at this again, I don't think we should set ID_AA64MMFR3_EL1.MEC if RME is not enabled.I guess the way to do this that fits in with how we already handle "user-facing property disabled feature X and that means we should also disable feature Y" is to put a check in arm_cpu_realizefn() for "if RME disabled then squash ID_AA64MMFR3.MEC to 0" ?
Agreed. r~
