On Mon, 7 Nov 2022 at 16:42, ~axelheider <[email protected]> wrote: > > From: Axel Heider <[email protected]> > > - inline software reset > - make hardware reset invoke software reset > - simplify code flow
I think this patch is fixing a bug, right? We weren't previously clearing CR for the hardware reset. If so, that's worth noting in the commit message. > Signed-off-by: Axel Heider <[email protected]> > +static void imx_epit_reset(DeviceState *dev) > +{ > + IMXEPITState *s = IMX_EPIT(dev); > + > + /* initialize CR and perform a software reset */ > + s->cr = 0; > + imx_epit_write_cr(s, CR_SWR); > +} Generally we prefer not to do this for the hardware reset function, as it makes it harder to see what the reset is doing (eg to confirm that it isn't changing qemu IRQ state). You can have a common helper function to do the work of the reset though if that helps. thanks -- PMM
