Re: [PATCH v2] memblock: Anonotate memblock_is_reserved() with __init_memblock.

[Yueyi Li <liyueyi@live.com>]

On 2018/12/4 11:04, Wei Yang wrote:
> On Mon, Dec 03, 2018 at 04:00:08AM +0000, Yueyi Li wrote:
>> Found warning:
>>
>> WARNING: EXPORT symbol "gsi_write_channel_scratch" [vmlinux] version generation failed, symbol will not be versioned.
>> WARNING: vmlinux.o(.text+0x1e0a0): Section mismatch in reference from the function valid_phys_addr_range() to the function .init.text:memblock_is_reserved()
>> The function valid_phys_addr_range() references
>> the function __init memblock_is_reserved().
>> This is often because valid_phys_addr_range lacks a __init
>> annotation or the annotation of memblock_is_reserved is wrong.
>>
>> Use __init_memblock instead of __init.
> Not familiar with this error, the change looks good to me while have
> some questions.
>
> 1. I don't see valid_phys_addr_range() reference memblock_is_reserved().
>     This is in which file or arch?

Yes,  I modified valid_phys_addr_range() for some other debugging.

> 2. In case a function reference memblock_is_reserved(), should it has
>     the annotation of __init_memblock too? Or just __init is ok? If my
>     understanding is correct, annotation __init is ok. Well, I don't see
>     valid_phys_addr_range() has an annotation.
> 3. The only valid_phys_addr_range() reference some memblock function is
>     the one in arch/arm64/mm/mmap.c. Do we suppose to add an annotation to
>     this?

Actually, __init_memblock is null in arch arm64, this warning is due to
CONFIG_DEBUG_SECTION_MISMATCH enabled,  the help text in lib/Kconfig.debug.



Thanks,
Yueyi