Re: [Qemu-devel] Re: [PATCH 5/5] RFC: distinguish warm reset from cold reset.

[Gleb Natapov]

On Tue, Aug 31, 2010 at 04:29:43PM +0300, Avi Kivity wrote:
>  On 08/31/2010 04:26 PM, Anthony Liguori wrote:
> >On 08/31/2010 08:21 AM, Gleb Natapov wrote:
> >>On Tue, Aug 31, 2010 at 08:20:51AM -0500, Anthony Liguori wrote:
> >>>On 08/31/2010 08:14 AM, Gleb Natapov wrote:
> >>>>System_reset should do cold reset like it does now.
> >>>Why?
> >>>
> >>Because I should not be forced to restart qemu to bring devices to
> >>initial state.
> >
> >IOW, you use system_reset for debugging purposes to reset the
> >device model.
> >
> >Point taken but functionally speaking, system_reset should map to
> >a RESET signal and from what I can tell in this thread, that's a
> >warm reset.
> >
> 
> Note, for most devices there's no difference.  x86 has INIT and
> RESET, with the keyboard controller RESET signal sometimes wired to
> INIT, and RAM doesn't have RESET.  Otherwise most devices don't see
> a difference.
> 
Actually soft reset is defined as INIT in PIIX4 spec:

Bits 1 and 2 in this register are used by PIIX4 to generate a hard reset
or a soft reset. During a hard reset, PIIX4 asserts CPURST, PCIRST#, and
RSTDRV, as well as reset its core and suspend well logic. During
a soft reset, PIIX4 asserts INIT.

--
                        Gleb.