[U-Boot] [PATCH 1/2] serial: add skipping init option

Simon Glass sjg at chromium.org
Sat Mar 30 21:19:21 UTC 2019


Hi Lukasz,

On Wed, 27 Mar 2019 at 03:34, Lukasz Majewski <lukma at denx.de> wrote:
>
> Hi Simon, Jun,
>
> > Hi Jun,
> >
> > On Fri, 22 Mar 2019 at 16:02, Jun Nie <jun.nie at linaro.org> wrote:
> > >
> > > Simon Glass <sjg at chromium.org> 于2019年3月22日周五 下午3:56写道:
> > > >
> > > > Hi Jun,
> > > >
> > > > On Fri, 22 Mar 2019 at 15:20, Jun Nie <jun.nie at linaro.org>
> > > > wrote:
> > > > >
> > > > > add skipping init option to avoid corrupt data in console
> > > > > if serial is already initilized when u-boot start its excution.
> > > > >
> > > > > Signed-off-by: Jun Nie <jun.nie at linaro.org>
> > > > > ---
> > > > >  drivers/serial/Kconfig | 6 ++++++
> > > > >  1 file changed, 6 insertions(+)
> > > >
> > > > Could we use a device-tree property for this? Some UARTs have a
> > > > 'skip-init' property.
> > > >
> > > > Regards,
> > > > Simon
> > >
> > > Hi Simon,
> > >
> > > It is a good suggestion. But device tree is board specific, while
> > > the config is boot flow specific. For ATF -> OPTEE -> U-BOOT case,
> > > initialization should be skipped. Without earlier initialization
> > > from firmware, U-boot should do it by itself. So I need find a
> > > solution that is defconfig specific, instead of device tree
> > > specific.
> >
> > So you want something that is flow-specific, I think. Your thinking is
> > that the defconfig specifies this. Is that because certain boards
> > require ATF/Optee?
> >
> > Should we consider having a way to tell U-Boot that it is running
> > from Optee?
> >
> > Could the UART driver detect that the UART is already inited?
>
> Isn't this issue similar to:
> http://patchwork.ozlabs.org/patch/820824/
>
> and
>
> http://patchwork.ozlabs.org/patch/865963/
>

Yes. But in the case of this patch I'm hoping we don't need it.

Regards,
Simon


More information about the U-Boot mailing list