Re: [PATCH] mips: econet: fulfill unmet kconfig dependencies

[Randy Dunlap]

On 8/12/26 11:50 AM, Caleb James DeLisle wrote:
> 
> On 12/08/2026 19:32, Jonas Gorski wrote:
>> On Wed, Aug 12, 2026 at 5:19 PM Randy Dunlap <rdunlap@xxxxxxxxxxxxx> wrote:
>>> Hui,
>>>
>>> On 8/12/26 6:12 AM, Jonas Gorski wrote:
>>>> Hi,
>>>>
>>>> On Wed, Aug 12, 2026 at 5:27 AM Randy Dunlap <rdunlap@xxxxxxxxxxxxx> wrote:
>>>>> ECONET selects SERIAL_8250 even though its TTY dependency is not met.
>>>>> ECONET selects SERIAL_OF_PLATFORM even though its TTY dependency is not
>>>>> met.
>>>>> ECONET selects EARLY_PRINTK_8250 even though its EARLY_PRINTK dependency
>>>>> is not met.
>>>>>
>>>>> Fix these kconfig warnings by adding selects for TTY and EARLY_PRINTK.
>>>>>
>>>>> WARNING: unmet direct dependencies detected for SERIAL_8250
>>>>>    Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && !S390
>>>>>    Selected by [y]:
>>>>>    - ECONET [=y]
>>>>>
>>>>> WARNING: unmet direct dependencies detected for SERIAL_OF_PLATFORM
>>>>>    Depends on [n]: TTY [=n] && HAS_IOMEM [=y] && SERIAL_8250 [=y] && OF [=y]
>>>>>    Selected by [y]:
>>>>>    - ECONET [=y]
>>>>>
>>>>> WARNING: unmet direct dependencies detected for EARLY_PRINTK_8250
>>>>>    Depends on [n]: EARLY_PRINTK [=n] && USE_GENERIC_EARLY_PRINTK_8250 [=y]
>>>>>    Selected by [y]:
>>>>>    - ECONET [=y]
>>>>>
>>>>> Signed-off-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
>>>>> ---
>>>>> Cc: Caleb James DeLisle <cjd@xxxxxxxx>
>>>>> Cc: Thomas Bogendoerfer <tsbogend@xxxxxxxxxxxxxxxx>
>>>>> Cc: linux-mips@xxxxxxxxxxxxxxx
>>>>>
>>>>>   arch/mips/Kconfig |    2 ++
>>>>>   1 file changed, 2 insertions(+)
>>>>>
>>>>> --- linux-next-20260810.orig/arch/mips/Kconfig
>>>>> +++ linux-next-20260810/arch/mips/Kconfig
>>>>> @@ -396,6 +396,7 @@ config ECONET
>>>>>          bool "EcoNet MIPS family"
>>>>>          select BOOT_RAW
>>>>>          select DEBUG_ZBOOT if DEBUG_KERNEL
>>>>> +       select EARLY_PRINTK
>>>>>          select EARLY_PRINTK_8250
>>>> I think the correct solution here is to drop EARLY_PRINTK_8250 instead.
>>>>
>>> Gladly.
>>>
>>>> ECONET already selects SYS_HAS_EARLY_PRINTK and
>>>> USE_GENERIC_EARLY_PRINTK_8250. EARLY_PRINTK defaults to y if
>>>> SYS_HAS_EARLY_PRINTK is y (unless EXPERT is enabled), and
>>>> EARLY_PRINTK_8250 defaults to y if both EARLY_PRINTK and
>>>> USE_GENERIC_EARLY_PRINTK_8250 are y.
>>>>
>>>> That way users can actually disable EARLY_PRINTK if they wish.
>>>>
>>>>>          select ECONET_EN751221_TIMER
>>>>>          select SERIAL_8250
>>>>> @@ -407,6 +408,7 @@ config ECONET
>>>>>          select SYS_SUPPORTS_32BIT_KERNEL
>>>>>          select SYS_SUPPORTS_MIPS16
>>>>>          select SYS_SUPPORTS_ZBOOT_UART16550
>>>>> +       select TTY
>>>> Here I don't know. ECONET seems to be the only platform (treewide)
>>>> that selects SERIAL_OF_PLATFORM and SERIAL_8250. Everyone else just
>>>> enables these via their defconfigs. No hard feelings one way or the
>>>> other, just noticed ECONET is the exception here.
>>> I don't prefer using "select" - it just went along with the rest of the
>>> Kconfig file. IOW, I would prefer using "depends on TTY".
>>> But you are saying that it could go into some (non-existing) defconfig file?
>> More like I grepped through the whole tree and ECONET was the only one
>> selecting these symbols, and all other occurrences where in
>> defconfigs.
>>
>> But as you correctly pointed out, there is no defconfig for econet, so
>> I guess as a quick fix selecting TTY is okay, and once a defconfig
>> exists we can drop it.
>>
>> Caleb, could you maybe add one?
> 
> 
> The use of select was not for any particular reason, when I implemented this I was still figuring out the platform and the best practices. I'll be happy to send a patch to normalize this in the next couple of days.

Great. I'll drop my patch and look forward to seeing yours.

Thanks.

-- 
~Randy