On 20 December 2016 at 15:31, Andrew Jones <drjo...@redhat.com> wrote: > Should we output both the destination PC (ELR) and the source PC (where > the eret was)? Otherwise if there are many erets to the same entry point, > then the logs won't fully enlighten us.
We don't really conveniently have the source PC, because it isn't written out to env->pc at the point when this helper function is called (the calling code in translate-a64.c doesn't call gen_a64_set_pc(), because it knows the helper doesn't need to care about the PC value). thanks -- PMM