On Thu, Sep 22, 2022 at 5:55 PM Thomas Huth <th...@redhat.com> wrote: > > On 22/09/2022 17.38, David Hildenbrand wrote: > > From: "Jason A. Donenfeld" <ja...@zx2c4.com> > > > > In order to fully support MSA_EXT_5, we have to support the SHA-512 > > special instructions. So implement those. > > > > The implementation began as something TweetNacl-like, and then was > > adjusted to be useful here. It's not very beautiful, but it is quite > > short and compact, which is what we're going for. > ... > > @@ -52,6 +278,9 @@ uint32_t HELPER(msa)(CPUS390XState *env, uint32_t r1, > > uint32_t r2, uint32_t r3, > > cpu_stb_data_ra(env, param_addr, subfunc[i], ra); > > } > > break; > > + case 3: /* CPACF_*_SHA_512 */ > > + return cpacf_sha512(env, ra, env->regs[1], &env->regs[r2], > > + &env->regs[r2 + 1], type); > > I have to say that I liked Jason's v8 better here. Code 3 is also used for > other instructions with completely different meaning, e.g. PCKMO uses 3 for > TDEA-192 ... so having the "type" check here made more sense. > (meta comment: maybe we should split up the msa function and stop using just > one function for all crypto/rng related instructions? ... but that's of > course something for a different patch series)
Maybe just commit my v8, and then David's changes can layer on top as follow ups? Checking len alignment, for example, is a separate patch from the rest. Jason
