On 23 February 2018 at 11:40, Peter Maydell <peter.mayd...@linaro.org> wrote: > I realized while working through the rest of the series that this is > too early to do the sve_access_check() and fp_access_check(). Those > only apply to instructions which actually exist, so we mustn't > do the checks until after we've dealt with all the unallocated_encoding() > cases. I think you need to push them down inside disas_sve() somehow. > See also my comments on patch 8.
...I meant "patch 9". -- PMM