Re: floating-point loads vs dreamcast

>> Any idea what's going on here?  As I read the hardware docs,
>> fmov @rN,drM certainly ought to do a 64-bit double-precision load.
> Only if you set SZ to 1.  PR controls the size of arithmetic
> operations, SZ controls the size of loads and stores.

Well, my notes say

| When the manual says "SZ 0", it means "SZ=0 and PR=0", and when it says
| "SZ 1", it means "(SZ=1 and PR=0) or (SZ=0 and PR=1)".  (This learned
| from Marcus Comstedt on port-dreamcast.)

Did I misunderstand you, or is this true generally but not true of
fmov, or what?  I can't just set SZ; the doc specifically calls out
SZ=1 PR=1 as an invalid combination, so I'd have to clear PR.  And it
seems..odd..to design an FPU which requires changing settings to both
do arithmetic on double-precision numbers and load and store
double-precision numbers.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		[email protected]
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B