On Thu, May 30, 2013 at 3:08 AM, Paolo Bonzini <pbonz...@redhat.com> wrote: > Il 29/05/2013 19:04, Edgar E. Iglesias ha scritto: >>> > + for (i = 0; i < R_MAX; ++i) { >>> > + RegisterInfo *r = &s->regs_info[i]; >>> > + >>> > + *r = (RegisterInfo) { >>> > + .data = &s->regs[i], >>> > + .data_size = sizeof(uint32_t), >>> > + .access = &xilinx_devcfg_regs_info[i], >>> > + .debug = XILINX_DEVCFG_ERR_DEBUG, >>> > + .prefix = prefix, >>> > + .opaque = s, >>> > + }; >>> > + memory_region_init_io(&r->mem, &devcfg_reg_ops, r, >>> > "devcfg-regs", 4); >> Hi Peter, Should we be putting r->access->name here instead of "devcfg-regs"? > > Yes, that's why I preferred to wrap the memory_region_init_io into an > API that takes a RegisterInfo. :)
ACK, You've convinced me :). Will be in v4 (pending outcome of discussion with Anthony RE decoding) Regards, Peter > > Paolo >