[RFC] Sandbox reset

[Simon Glass]

Hi Heinrich,

On Mon, 23 Mar 2020 at 09:01, Heinrich Schuchardt <xypron.glpk at gmx.de> wrote:
>
> On 3/23/20 2:18 PM, Simon Glass wrote:
> > Hi Heinrich,
> >
> > On Mon, 23 Mar 2020 at 00:40, Heinrich Schuchardt <xypron.glpk at gmx.de> wrote:
> >>
> >> Hello Simon,
> >>
> >> currently the reset command on the sandbox exits the program. Is there
> >> any requirement for this behavior in our current testing landscape? For
> >> testing the UEFI sub-system I would prefer that the reset command
> >> restarts the sandbox and the poweroff command is used for leaving.
> >
> > I think it could be changed to do that. There are a few tests that
> > need 'bootm' to exit (e.g. vboot) I think. There might be other
> > places.
> >
> > Actually 'poweroff' should move to driver model. The sysreset uclass
> > has power-off capability.
>
> We still have to implement cold reset in sandbox_sysreset_request().
>
> I guess it will take:
>
> os_fd_restore();
> state_uninit();
> dm_unint();
> os_free();
>
> and a longjmp() into main().
>
> Anything I missed?

Sounds right. Not to jump into the weeds, but it might be good to put
that code next to the current main().

Regards,
Simon