Re: [PULL 22/94] target/sparc: Move RDASR, STBAR, MEMBAR to decodetree

[Richard Henderson]

On 11/3/23 12:07, Peter Maydell wrote:
> On Thu, 26 Oct 2023 at 01:25, Richard Henderson
> <richard.henderson@linaro.org> wrote:
> > Tested-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
> > Acked-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk>
> > Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
>
>
> > +static TCGv do_rdgsr(DisasContext *dc, TCGv dst)
> > +{
> > +    gen_trap_ifnofpu(dc);
> > +    return cpu_gsr;
> > +}
> > +
> > +TRANS(RDGSR, 64, do_rd_special, true, a->rd, do_rdgsr)
>
> Hi; Coverity complains (CID 1523706) that we don't check
> the return value from this call to gen_trap_ifnofpu(),
> whereas we do check it on all but one other call.
> Is this OK?

Yes.

While gen_trap_ifnofpu may or may not trap, we must return a legal value for the calling 
wrapper function.  If a trap is generated, the read of cpu_gsr will be removed as dead code.

For do_wrgsr, we *could* avoid generating the dead store to cpu_gsr, but avoiding that is 
not critical.


r~