Although that this is functionally equivalent (since Zvfhmin depends on Zve32f), it makes clearer that Zvfh is a strict superset of Zvfhmin.
Again, it will not change the contents of both extensions.
But I believe that this change will ease some developers that depend on the expanded ISA string (e.g. after expanding G → IMAFD_Zicsr_Zifencei) or similar data structure (like which QEMU uses) because if we need to probe the vector FP16 conversion capability, we just need to probe for the Zvfhmin extension, not Zvfhmin or Zvfh.
Although that this is functionally equivalent (since
Zvfhmin
depends onZve32f
), it makes clearer thatZvfh
is a strict superset ofZvfhmin
.Again, it will not change the contents of both extensions. But I believe that this change will ease some developers that depend on the expanded ISA string (e.g. after expanding
G
→IMAFD_Zicsr_Zifencei
) or similar data structure (like which QEMU uses) because if we need to probe the vector FP16 conversion capability, we just need to probe for theZvfhmin
extension, notZvfhmin
orZvfh
.