On Thu, 11 Apr 2019 at 13:07, Igor Mammedov <[email protected]> wrote:
>
> On Wed, 10 Apr 2019 23:11:00 +0000
> Alistair Francis <[email protected]> wrote:
> > + /* register system main memory (actual RAM) */
> > + memory_region_init_ram(main_mem, NULL, "riscv.spike.ram",
> > + machine->ram_size, &error_fatal);
> do you really care about migration? if not then _nomigrate flavor would
> be more suitable.
Every machine model should care about migration, at least to
the extent of making savevm/loadvm snapshots work. Using
memory_region_init_ram_nomigrate() is a bad idea in new
code: it exists only for the benefit of older code which
is handling the migration of the backing RAM by hand for
migration-compatibility reasons.
> > /* Initialize SOC */
> > object_initialize_child(OBJECT(machine), "soc", &s->soc,
> > sizeof(s->soc),
> > TYPE_RISCV_HART_ARRAY, &error_abort, NULL);
> > @@ -254,6 +343,12 @@ static void spike_v1_09_1_board_init(MachineState
> > *machine)
> > MemoryRegion *mask_rom = g_new(MemoryRegion, 1);
> > int i;
> >
> > + if (!qtest_enabled()) {
> > + info_report("The Spike v1.09.1 machine has been depreceated. "
> > + "Please use the deneric spike machine and specify the
> > ISA "
"generic"
> > + "versions using -cpu.");
> > + }
> > +
thanks
-- PMM
