Re: Byte ordering (was: Whether something is RISC or not)

On 03/10/2024 11:39, Anton Ertl wrote:

David Brown <david.brown@hesbynett.no> writes:

Since you mentioned POWER and PowerPC elsewhere, the bit numbering
challenges of the m68k world are nothing compared to the PowerPC world.
Originally, the PowerPC was 32-bit and numbered bits from 0 as the MSB
down to 31 as the LSB.  So your 32-bit address bus had lines from A0
down to A31.  Then it got extended to 64-bit (some devices had only
partial 64-bit extensions), and the chips got a wider address bus (you
never need all 64-bit lines physically) - the pins for the higher
address lines were numbered A-1, A-2, and so on.  For the internal
registers, some are 64-bit numbered bit 0 (MSB) down to bit 63.  Some
were original 32-bit and got extended to 64-bit, and so are numbered bit
-32 down to bit 31 for consistency.  Others are 32-bit but numbered from
bit 32 down to bit 63.

 Maybe they should have started with the MSB as bit -31 or -63, which
would have allowed them to always use bit 0 for the LSB while having
big-endian bit ordering.
 

That's a very "outside the box thinking" solution!

For bit ordering big-endian (as in the PowerPC manual) looked more
wrong to me than for byte ordering; I thought that that was just a
matter of getting used to the unfamiliar bit ordering, but maybe the
advantage of little-endian becomes more apparent in bit ordering, and
maybe that's why Motorola and Sun chose little-endian bit ordering
despite having big-endian byte ordering.
 For both bit and byte ordering, the advantage of little-endian shows
up when there are several widths involved.  So why is it more obvious
for bit-ordering?

I have certainly found big-endian bit numbering harder to get my head around than big-endian byte ordering.  One possible explanation is that with little-endian ordering, (1 << bit_no) gives you a 1 in the right bit number.  Another is that with little-endian bit ordering, the same bit number has the same value regardless of the size of the type.  And I work with electronics as well as software - virtually everything in hardware (except PowerPC microcontrollers!) uses little-endian bit numbering.  Smallest to largest, counting upwards from 0 or 1, is just more natural.

 BTW, at least in my 32-bit PowerPC manual the claim is that PowerPC is
a 64-bit architecture, and that the manual describes only the 32-bit
subset.  Maybe the original Power was 32-bit.
 

As I understand it - and my history here might not be completely accurate - PowerPC was specified for both 32-bit and 64-bit from early on, but first made in the 32-bit version.  There were quite a number of optional parts of the PowerPC architecture, including 64-bit width, floating point units, and support for little-endian data modes - IIRC these were referred to as books of various colours.  And yes, you could then have weird things like it being a 64-bit architecture where none of the 64-bit features were actually implemented.  One microcontroller I used had 64-bit GPRs, but almost no 64-bit instructions - I don't think you could even load or save all 64 bits at a time.  The only use of them was for transferring to and from the 64-bit double precision floating point registers (which could be loaded and saved in full).