Re: [PATCH] x86/x86_64_defconfig: Enable the serial console

From: Enric Balletbo i Serra
Date: Sun Oct 11 2020 - 07:43:56 EST


Hi Borislav and Randy,

Thank you for your comments.

On 8/10/20 20:31, Randy Dunlap wrote:
> On 10/8/20 9:40 AM, Borislav Petkov wrote:
>> On Thu, Oct 08, 2020 at 06:22:06PM +0200, Enric Balletbo i Serra wrote:
>>> As part of KernelCI, we added a bunch of different x86 based Chromebooks
>>> to do test booting and runtime testing. It will be useful have serial
>>> support for these platforms in the default defconfig, as this, is the
>>> defconfig used by default for the different maintainer's tree.
>>>
>>> SERIAL_8250_DW is the actual support for the console, but to have
>>> support we need to enable X86_AMD_PLATFORM_DEVICE for specific AMD
>>> boards and MFD_INTEL for specific Intel boards.
>>>
>>> While here, also enable USB_RTL8152 config which enables a common
>>> USB-Ethernet adapter used very commonly in the KernelCI labs.
>>
>> To me defconfig sounds like the config which contains items which are
>> needed on the majority of x86 hardware out there.
>
> to run in normal (or common) user mode(s).
>
> Nothing to do with automated testing like syzbot or CI IMO.
>
>> Are those which you enable that common and if not, why can't your build
>> supply a custom .config instead?
>
> I suppose that we are lacking a definition of a defconfig, but again, IMO,
> these Kconfig symbols don't meet the need for normal users.
>
>

We're also probably lacking a definition of what normal users mean, because I
don't think normal users build their own kernel. I think that at least
X86_AMD_PLATFORM_DEVICE and MFD_INTEL_LPSS_PCI could be common enough to match
within the category of needed to run in normal (or common) user mode(s). I can
send a patch with only these two options.

But, yes, the main purpose after this patch is the serial console for CI. I saw
that there are already some configs with a specific purpose (tiny.config and
xen.config). So, I am wondering if would be acceptable support another specific
config for CI (i.e kernelci.config). Will it be acceptable?

Thanks,
Enric