Joern Rennecke wrote:

>>>According to the sh2-dsp manual that is still at the Renesas web site,
>>>the xx / yy fields are still present in the pmuls instruction.
>>
>>Hmm, well, are they used for anything?  I think I took them out to
>>resolve a conflict with another insn (but I don't remember for sure).
>>Since there's no corresponding register parameter, and the code
>>does not use them -- is there any harm?  If the other patterns
>>are not used now, they probably will be someday.
> 
> 
> According to the manual, because of the ignored x,y,u operands, there are
> 64 valid pmuls opcodes for each e,f,g combination.  I can't find anything
> in the manual that says that any one of these opcodes is preferred.
> If there is a conflict, it appears that the sh4a-dsp is not actually
> backward compatible with the sh3-dsp, and we'll need another opcode table
> modification depending on bfd_get_mach.  OTOH I seem to remember that we
> were asked to zero unused fields in the assembler, but that might be just
> to get reproducible results.  You should probably ask Renesas for
> clarification.

OK, I'll restore the xxyy for now, just to get it committed,
and address it either as it comes up or as clarification obtains.


>>>What are these FIXMEs supposed to mean?
>>
>>I did this work 4 months ago.  Probably  I thought the
>>comment was wrong, and that the actual set of registers
>>was as shown.  Your second opinion would be appreciated.
> 
> 
> The first part of the comments is right, the FIXMEs are wrong.  The
> registers are numbered like in the movs instruction which is the same
> as in the the Dz parameter, with A1G / A0G having a number 8 higher
> than A1 / A0.  See also the macros in interp.c after DSR_R.

OK, removing the fixme comments.  Committing with changes
as attached below: