On Fri, Jul 4, 2014 at 2:59 AM, Andreas Färber <afaer...@suse.de> wrote: > Am 03.07.2014 18:46, schrieb Peter Maydell: >> On 24 June 2014 07:06, Alistair Francis <alistair.fran...@xilinx.com> wrote: >>> SysBusDevice::init is deprecated. Convert to Object::init and > > Note that there is no Object::init, only TypeInfo::instance_init. >
So I've been using these C++ish ::s in commentry liberally without worrying about their literal correctness. TypeInfo requires a bit of QOM internal understanding to get the point, whearas a much wider audience has awarness of the Object type and the fact that is has an init. If anything I would just soften to plain english - "object init". >>> Device::realize as prescribed by QOM conventions. >>> >>> Signed-off-by: Alistair Francis <alistair.fran...@xilinx.com> >>> --- >>> >>> hw/char/cadence_uart.c | 29 ++++++++++++++++------------- >>> 1 files changed, 16 insertions(+), 13 deletions(-) >>> >>> diff --git a/hw/char/cadence_uart.c b/hw/char/cadence_uart.c >>> index bf0c853..5a22a72 100644 >>> --- a/hw/char/cadence_uart.c >>> +++ b/hw/char/cadence_uart.c >>> @@ -468,27 +468,30 @@ static void cadence_uart_reset(DeviceState *dev) >>> uart_update_status(s); >>> } >>> >>> -static int cadence_uart_init(SysBusDevice *dev) >>> +static void candence_uart_realize(DeviceState *dev, Error **errp) >> >> Typo in your new function name :-) >> Will fix v2 (i'll respin this one) >> Otherwise >> Reviewed-by: Peter Maydell <peter.mayd...@linaro.org> > Thanks, Regards, Peter > LGTM, but I wonder whether the work Alex is doing on SysBusDevice > requires us to introduce SysBusDevice::realize, called from > Device::realize in SysBusDevice code? > > Regards, > Andreas > > -- > SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany > GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg >