Le 20/10/2021 à 16:08, Richard Henderson a écrit :
> On 10/19/21 2:47 AM, Frédéric Pétrot wrote:
>> +
>> +#define REQUIRE_64_OR_128BIT(ctx) do { \
>> + if (get_xl(ctx) == MXL_RV32) { \
>> + return false; \
>> + } \
>> +} while (0)
>
> So... you've left REQUIRE_64BIT accepting RV128, so that means that your
> current
> REQUIRE_64_OR_128BIT is redundant. Is that intentional?
>
> It does seem like all places that accept RV128 should accept RV64, but perhaps
> that's just your "limited" caveat in the cover letter.
My bad, indeed there is no instruction only required by RV64. "Limited" was
related to the minimal support of the priviledge spec.
> You don't use REQUIRE_32_OR_64BIT at all. Remove it?
It's a bug : some compressed insns are only RV32/RV64 (this is linked to
the other bug in the order in which the insns stand in the insn16.decode
file that you pointed out).
Frédéric
>
>
> r~
--
+---------------------------------------------------------------------------+
| Frédéric Pétrot, Pr. Grenoble INP-Ensimag/TIMA, Ensimag deputy director |
| Mob/Pho: +33 6 74 57 99 65/+33 4 76 57 48 70 Ad augusta per angusta |
| http://tima.univ-grenoble-alpes.fr frederic.pet...@univ-grenoble-alpes.fr |
+---------------------------------------------------------------------------+
