Re: [hatari-devel] 68030 cycle timings for 64bit div/mul

[ Thread Index | Date Index | More lists.tuxfamily.org/hatari-devel Archives ]


On 29 Jan 2013 at 17:15, laurent.sallafranque@xxxxxxx wrote:

> Motorola doc gives a max value in the docs for the mul and div instructions
> (for every declinaison of these instruction, ie .w, .l, 64 bits, and for the
> signed/unsigned instructions).
> 
> I've inplemented these max values into hatari, but of course, this is not
> accurate.
> As long as we don't manage to figure how motorola implemented these instr into
> the CPU, 68030 emu won't be accurate here.
> 
I've been working on the divide stuff on and off.  I have results for DIVS.W 
and DIVU.W which I can post together with the source code for the program that 
obtained them.

I started work on the 32-bit divides but that's paused ATM.  The problem there 
of course is that it's impossible to perform every possible divide and check 
the results, so the code tries to do quasi-random, quasi-boundary tests.  I 
have some preliminary results, but I need to fire up the TT again to continue.  
If anyone is interested in me continuing (or in getting my code & pursuing it 
themselves), let me know.

Regards,
Roger Burrows




Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/