Re: [PATCH 3/3] cm-fx6: add cm-fx6 machine configuration

[Igor Grinberg <grinberg@compulab.co.il>]

On 04/13/15 16:47, Otavio Salvador wrote:
> On Sat, Apr 11, 2015 at 6:39 AM, Valentin Raevsky
> <valentin@compulab.co.il> wrote:
>> Add cm-fx6 machine configuration: u-boot and kernel.
>> The cm-fx6 is a CompuLab CoM. The board features:
>> -- Freescale i.MX6 quad-core Cortex-A9 CPU, 1.2GHz
> 
> Quad / DualLite ?

Yes, thanks. We use Quad, Dual, and Solo (DL).

> 
>> -- Up to 4GB DDR3 and up to 32GB on-board SSD storage
>> -- Integrated GPU and 1080p VPU, dual display support
>> -- WiFi, BT 3.0, GbE, PCIe, SATA, USB x5, UART x5, CAN x2
>>
>> For more information visit:
>> http://www.compulab.co.il/products/computer-on-modules/cm-fx6/
> 
> Great :)
> 
>> Signed-off-by: Valentin Raevsky <valentin@compulab.co.il>
>> ---
>>  conf/machine/cm-fx6.conf |   22 ++++++++++++++++++++++
>>  1 file changed, 22 insertions(+)
>>  create mode 100644 conf/machine/cm-fx6.conf
>>
>> diff --git a/conf/machine/cm-fx6.conf b/conf/machine/cm-fx6.conf
>> new file mode 100644
>> index 0000000..1b357cd
>> --- /dev/null
>> +++ b/conf/machine/cm-fx6.conf
>> @@ -0,0 +1,22 @@
>> +#@TYPE: Machine
>> +#@NAME: CompuLab CM-FX6
>> +#@SOC: i.MX6 Q/DL
>> +#@DESCRIPTION: Machine configuration for CompuLab CM-FX6 machines
>> +#@MAINTAINER: Valentin Raevsky <valentin@compulab.co.il>
> 
> Great.
> 
>> +require conf/machine/include/imx-base.inc

Valentin told me that including this results in some strange build
breakage...
Is it because we tried with kernel kernel 3.10.17?

>> +require conf/machine/include/tune-cortexa9.inc
>> +
>> +SOC_FAMILY = "mx6:mx6dl:mx6q"
>> +
>> +PREFERRED_PROVIDER_u-boot = "u-boot-cm-fx6"
>> +PREFERRED_PROVIDER_virtual/kernel = "linux-cm-fx6"
> 
> This will need to be adjusted for v2, as part of the renaming.

No problem, I just want to understand, how should we add boards to the
same recipes (linux-compulab, u-boot-compulab).

> 
> ...
>> +KERNEL_IMAGETYPE = "zImage"
> 
> zImage is the BSP layer default so please drop it.

Ok.

> 
> Apart from this, it is very good.
> 
> Generally the most important remark is the way you've been sending the
> patchset. You need to add the proper subject-prefix when sending it.
> 
> You can take a look at the README[1] as it describes the proper way to do it.
> 
> 1. https://github.com/Freescale/meta-fsl-arm-extra

Thanks for the pointer!

> 
> Thanks for the hard work you've been doing at cooking this patches. I
> know the first iterations are tough but this ends being a pleasant
> experience as soon you get more used to it, I apologize in send so
> many "remarks" to the patchset but those are necessary so we can keep
> the good level of quality of the BSP layer in long term.

Well, thank you for reviewing those patches and being patient!
We have a mutual goal, so no problem with remarks or their amount.
As more remarks come in now, less we will have to deal with later ;-)

-- 
Regards,
Igor.