On Mon, Oct 2, 2023, at 21:07, Frank Li wrote:
> On Mon, Oct 02, 2023 at 08:55:23PM +0200, Arnd Bergmann wrote:
>
>> A few other thoughts from my side, all of which could be ignored:
>>
>> - if the ioport access is not an important feature, we can instead
>> support 64-bit readl() as I commented in a previous email. We just
>> can't easily have both.
>
> We will get 64bit dma edma soon. So I can test and upstream it when I get
> it.
Ok, so if we already know this is going to be needed, then I would skip
the PIO support and just use read{bwlq}() with the optional swab() instead
of the ioread variants. Otherwise there is a risk that someone starts
relying on the port I/O feature and make it harder to remove.
Arnd
