Re: [RFC 6/6] xen/arm: Replace early_printk call to printk call

[Julien Grall <julien.grall@linaro.org>]

On 20/02/14 09:04, Ian Campbell wrote:
> On Wed, 2014-02-19 at 17:56 +0000, Julien Grall wrote:
>> Hi Ian,
>>
>> On 02/19/2014 11:20 AM, Ian Campbell wrote:
>>> On Sun, 2014-01-05 at 21:26 +0000, Julien Grall wrote:
>>>> Now that the console supports earlyprintk, we can get a rid of
>>>> early_printk call.
>>>>
>>>> Signed-off-by: Julien Grall <julien.grall@linaro.org>
>>>
>>> Acked-by: Ian Campbell <ian.campbell@citrix.com>
>>
>> Thanks, due to some conflicts I have rebased this patch on top my patch
>> "xen/serial: Don't leak memory mapping if the serial initialization has
>> failed".
>>
>> I will resend this series later.
>
> Thanks.
>
>>> Now all we need is a way to make it a runtime option :-)
>>
>> I let you write a device tree parser in assembly ;).
>
> ;-)
>
> I was actually thinking more along the lines of a .word at a defined
> offset which you could hex edit to a specific value to activate a
> particular flavour of early printk handling. That would be sufficient
> e.g. for osstest to activate the appropriate stuff for the specific
> platform.

I don't see useful use case to have a such early printk implementation 
in Xen. When the board is fully supported, failed at early stage (e.g 
before console is initialized) is very unlikely. At least if you don't 
play with memory.

I see osstest like a more generic test suite. If it fails at early 
stage, the developper should give a try himself (most of the time early 
printk is not useful to find the really bug), that would mean it need to 
recompile Xen.

Cheers,

-- 
Julien Grall