On 6/13/23 11:56 AM, Stefan Wahren wrote:
[...]
>>>> The driver overrides the error codes and IRQ0 returned by platform_get_irq()
>>>> to -EINVAL, so if it returns -EPROBE_DEFER, the driver will fail the probe
>>>> permanently instead of the deferred probing. Switch to propagating the error
>>>> codes upstream. IRQ0 is no longer returned by platform_get_irq(), so we now
>>>> can safely ignore it...
>>>>
>>>> Fixes: 660fc733bd74 ("mmc: bcm2835: Add new driver for the sdhost controller.")
>>>
>>> I know this is very theoretical, but does the statement "IRQ0 is no longer returned by platform_get_irq()" also applies to the time of the fixes commit?
>>
>> Unfortunately, no. IRQ0 finally ceased to be returned in 5.19; there was a fat
>> warning in platform_get_irq() and friends before that (which is still there)...
>
> Okay, in this case the usage of the fixes tag is wrong.
Why? Returning -EPROBE_DEFER from platform_get_irq() predates this driver.
> Maybe we should refer to the commit which changed platform_get_irq()?
No, IRQ0 is a different issue than that I'm trying to solve here.
>>> I'm asking because the fix could be backported to Linux 4.14.
>>
>> I think the deferred probing can currently occur only with DT platforms
ACPI too (I was too lazy to look on the code yesterday).
>> (I may be wrong here). Is this your case?
>
> AFAIK Raspberry Pi was always a DT platform in the mainline kernel. At least in Linux 4.14.
Good to know. :-)
>>> Best regards
>>>
>>>> Signed-off-by: Sergey Shtylyov <s.shtylyov@xxxxxx>
>> [...]
MBR, Sergey
