Re: [PATCH for-4.5 1/4] xen: arm: Add earlyprintk for McDivitt.

[Ian Campbell <Ian.Campbell@citrix.com>]

On Tue, 2014-11-18 at 16:59 +0000, Julien Grall wrote:
> Hi Ian,
> 
> On 11/18/2014 04:44 PM, Ian Campbell wrote:
> > Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
> > ---
> >  xen/arch/arm/Rules.mk |    6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/xen/arch/arm/Rules.mk b/xen/arch/arm/Rules.mk
> > index 572d854..ef887a5 100644
> > --- a/xen/arch/arm/Rules.mk
> > +++ b/xen/arch/arm/Rules.mk
> > @@ -95,6 +95,12 @@ EARLY_PRINTK_BAUD := 115200
> >  EARLY_UART_BASE_ADDRESS := 0x1c020000
> >  EARLY_UART_REG_SHIFT := 2
> >  endif
> > +ifeq ($(CONFIG_EARLY_PRINTK), xgene-mcdivitt)
> > +EARLY_PRINTK_INC := 8250
> > +EARLY_PRINTK_BAUD := 9600
> 
> EARLY_PRINTK_BAUD is not necessary as we don't use the initialization
> function (EARLY_PRINTK_INIT_UART is not set).

Oh yes, oops. Also the baud is not even what is actually used, so it's
not even serving a documentary purpose.

> With the EARLY_PRINTK_BAUD dropped, this could be merged with the
> xgene-storm  early printk

It's at a different base address. Long term I either want to make this
(somewhat) runtime configurable or at least to rationalise the options
into the form <soc/soc-family>-uart<N>, or perhaps even <8250|pl011|
etc>@<address>[,<rate><settings>], if it's not to skanky to arrange to
parse that somewhere in the build system. Not for 4.5 though.

> (I didn't really understand why the baud rate
> is different).

Different hardware might potentially have different baud rates
configured in firmware which we would want to seemlessly follow, but
it's moot since the right thing to do in most cases is leave the
bootloader provided cfg alone.

> But I don't think it's 4.5 material.

You mean the patch generally or the merging?

Ian.