Article: AMD's Mobile Strategy
By: Paul A. Clayton (paaronclayton.delete@this.gmail.com), December 21, 2011 4:08 pm
Room: Moderated Discussions
Exophase (exophase@gmail.com) on 12/21/11 wrote:
---------------------------
[snip]
>How exactly would you have proposed handling this? To
>support 32 or 64-bit displacements generally you need
>another bit in the REX prefix that you don't have. The
>alternative is to make all prefixes 64-bits instead of
>32-bits (or 8-bits). I don't think I have to explain why
>that would be a much worse idea.
An alternative would be to provide state (modes) that
determine the size of immediates in various circumstances.
Placing such state in the address translation table could
reduce decode issues with respect to variable length. (I
am not suggesting that such is necessarily a good idea, but
it could probably be made to work.)
---------------------------
[snip]
>How exactly would you have proposed handling this? To
>support 32 or 64-bit displacements generally you need
>another bit in the REX prefix that you don't have. The
>alternative is to make all prefixes 64-bits instead of
>32-bits (or 8-bits). I don't think I have to explain why
>that would be a much worse idea.
An alternative would be to provide state (modes) that
determine the size of immediates in various circumstances.
Placing such state in the address translation table could
reduce decode issues with respect to variable length. (I
am not suggesting that such is necessarily a good idea, but
it could probably be made to work.)