Le 06/23/2017 à 11:51 AM, Peter Maydell a écrit :
On 15 June 2017 at 16:15, Edgar E. Iglesias <edgar.igles...@xilinx.com> wrote:
On Thu, Jun 15, 2017 at 04:04:56PM +0100, Peter Maydell wrote:
The difference here is that the clock objects themselves have
internal state. That's not necessarily a bad idea, but it does
mean that something's got to migrate that state or otherwise
regenerate it. (Anthony once proposed that we should change qemu_irq
objects to have internal state, because that's effectively what real
hardware is and it would save the need for each device to track its
input line state and be notified if the line didn't actually change
state. It would just have been an enormous upheaval and migration
compat break...)
Thanks. I didn't realize that the internal clock state would be used in a
way that is not deriveable from other device state & input clocks.
I'm not sure this is a use-case we need to support, thoughts?
Maybe there's some value in keeping interrupt and clock handling alike,
e.g removing the internal state from clocks.
It occurs to me that it might be possible to split the
difference, ie the clock object still has internal state,
but the device that owns it is responsible for that state's
migration, so you would have
VMSTATE_CLOCK(myclock, MyDeviceState),
Yes I think that makes sense.
in the vmstate for every clock you owned. (Not sure whether
that should be "every input clock" or "every output clock"
or even both?)
As an aside, I still find it very odd that you get a clock
object for both an input clock and an output clock. I feel
like we should have one end owns the clock object and the
other just has a reference to it of some kind.
The point is beeing able to bind/unbind clocks inside the device
to model a clock selector for example. So you just bind/unbind
the input and output clocks.
Secondly you can print the clock input rate in the mtree command
so you can eg: debug it.
Thanks,
Fred
thanks
-- PMM
