Re: uninitialized pmem struct pages

[David Hildenbrand]

On 05.01.21 08:44, Michal Hocko wrote:
> On Mon 04-01-21 17:30:48, David Hildenbrand wrote:
>>>> Let's assume this is indeed a reserved pfn in the altmap. What's the
>>>> actual address of the memmap?
>>>
>>> Not sure what exactly you are asking for but crash says
>>> crash> kmem -p 6060000
>>>       PAGE          PHYSICAL      MAPPING       INDEX CNT FLAGS
>>> fffff8c600181800     6060000                0        0  0 fffffc0000000
>>
>> ^ this looks like it was somewhat initialized. All flags zero, nid/zone
>> set to -1 (wild guess) and thus the crash? weird
> 
> Yes that made me scratch my head as well.

My best guess would be that this is indeed part of the altmap reserve
with a memmap on the altmap. Maybe these values are just leftovers on
the PMEM device (e.g., from older kernels)?

In the kernel, I cannot find anything that would initialize node/zone to
-1 only.


@Dan: Will get_dev_pagemap() return a valid dev_pagemap for pfns falling
into the altmap reserved region? I think we should also initialize the
memmap of the altmap reserved region somewhow.

-- 
Thanks,

David / dhildenb