> -----Original Message-----
> From: Richard Henderson <[email protected]>
> Sent: Saturday, March 13, 2021 7:40 PM
> To: Taylor Simpson <[email protected]>; [email protected]
> Cc: [email protected]; [email protected]; [email protected];
> [email protected]; Brian Cain <[email protected]>
> Subject: Re: [PATCH v8 26/35] Hexagon (target/hexagon) TCG generation
>
> >> Except for HEX_DEBUG, why would we not know whether or not a
> predicate
> >> has been
> >> written twice?  It seems like we shouldn't need hex_pred_written for the
> >> non-debug case.
> >
> > It's legal to write to a predicate more than once in the same packet.
>
> I repeat -- why would you not know that?  You should be able to examine the
> packet and see whether it does or does not write to a predicate register
> more
> than once.
>
> Simply always generating code that assumes it has is lazy.

Sorry misunderstood the question - yes, it should be possible at TCG generation 
time (rather than TCG execution time) to determine if the predicate has already 
been written.  I will code this up.

Thanks,
Taylor

Reply via email to