Re: [PATCH v4] Kconfig: introduce HAS_IOPORT option and select it as necessary

[H. Peter Anvin]

On April 5, 2023 8:12:38 AM PDT, Niklas Schnelle <schnelle at linux.ibm.com> wrote:
>On Thu, 2023-03-23 at 17:33 +0100, Niklas Schnelle wrote:
>> We introduce a new HAS_IOPORT Kconfig option to indicate support for I/O
>> Port access. In a future patch HAS_IOPORT=n will disable compilation of
>> the I/O accessor functions inb()/outb() and friends on architectures
>> which can not meaningfully support legacy I/O spaces such as s390.
>> 
>> The following architectures do not select HAS_IOPORT:
>> 
>> * ARC
>> * C-SKY
>> * Hexagon
>> * Nios II
>> * OpenRISC
>> * s390
>> * User-Mode Linux
>> * Xtensa
>> 
>> All other architectures select HAS_IOPORT at least conditionally.
>> 
>> The "depends on" relations on HAS_IOPORT in drivers as well as ifdefs
>> for HAS_IOPORT specific sections will be added in subsequent patches on
>> a per subsystem basis.
>> 
>> Co-developed-by: Arnd Bergmann <arnd at kernel.org>
>> Signed-off-by: Arnd Bergmann <arnd at kernel.org>
>> Acked-by: Johannes Berg <johannes at sipsolutions.net> # for ARCH=um
>> Acked-by: Geert Uytterhoeven <geert at linux-m68k.org>
>> Signed-off-by: Niklas Schnelle <schnelle at linux.ibm.com>
>> ---
>> Note: This patch is the initial patch of a larger series[0]. This patch
>> introduces the HAS_IOPORT config option while the rest of the series adds
>> driver dependencies and the final patch removes inb() / outb() and friends on
>> platforms that don't support them. 
>> 
>> Thus each of the per-subsystem patches is independent from each other but
>> depends on this patch while the final patch depends on the whole series. Thus
>> splitting this initial patch off allows the per-subsytem HAS_IOPORT dependency
>> addition be merged separately via different trees without breaking the build.
>> 
>> [0] https://lore.kernel.org/lkml/20230314121216.413434-1-schnelle@linux.ibm.com/
>> 
>> Changes since v3:
>> - List archs without HAS_IOPORT in commit message (Arnd)
>> - Select HAS_IOPORT for LoongArch (Arnd)
>> - Use "select HAS_IOPORT if (E)ISA || .." instead of a "depends on" for (E)ISA
>>   for m68k and parisc
>> - Select HAS_IOPORT with config GSC on parisc (Arnd)
>> - Drop "depends on HAS_IOPORT" for um's config ISA (Johannes)
>> - Drop "depends on HAS_IOPORT" for config ISA on x86 and parisc where it is
>>   always selected (Arnd)
>> 
>
>Gentle ping. As far as I can tell this hasn't been picked to any tree
>sp far but also hasn't seen complains so I'm wondering if I should send
>a new version of the combined series of this patch plus the added
>HAS_IOPORT dependencies per subsystem or wait until this is picked up.
>
>Thanks,
>Niklas
>
>

You need this on a system supporting not just ISA but also PCI.

Typically on non-x86 architectures this is simply mapped into a memory window.