Jason King wrote:
> As a potential follow on to my work with implementing an unencumbered
> sparc libdisasm, I was wondering if there would be interest in adding
> ASI name support to the disassembly.
>
> I.e. instead of:
>
> lda [%r1 + %r2] 0x80, %r3
>
> you'd see
>
> lda [%r1 + %r2] ASI_P, %r3
>
> or perhaps
>
> lda [%r1 + %r2] ASI_PRIMARY, %r3
>
> Similar for sta, etc.
>
> Myself, I find symbolic names easier to remember/deal with than
> arbitrary numbers, but I'd rather get the opinions of those who would
> be using this far more regularly than I would (I'm hoping they're
> reading these lists).
>   

I suspect that the use of alternate ASIs is infrequent enough to not 
really be worth it.

When I've used alternate ASIs they were to access chip specifics that I 
couldn't get any other way.  It was only in a tiny, tiny bit of code, 
and I don't think you could have used a meaningful name.  (IIRC, the ASI 
used was one reserved for implementations.)

    -- Garrett
> Anyone care to comment?
> _______________________________________________
> mdb-discuss mailing list
> mdb-discuss at opensolaris.org
>   


Reply via email to