Re: [PATCH] mm,memory_hotplug: Explicitly pass the head to isolate_huge_page

[Mike Kravetz]

On 2/12/19 6:40 AM, Michal Hocko wrote:
> On Tue 12-02-19 14:45:49, Oscar Salvador wrote:
>> On Tue, Feb 12, 2019 at 09:33:29AM +0100, Michal Hocko wrote:
>>>>  
>>>>  		if (PageHuge(page)) {
>>>>  			struct page *head = compound_head(page);
>>>> -			pfn = page_to_pfn(head) + (1<<compound_order(head)) - 1;
>>>>  			if (compound_order(head) > PFN_SECTION_SHIFT) {
>>>>  				ret = -EBUSY;
>>>>  				break;
>>>>  			}
>>>
>>> Why are we doing this, btw? 
>>
>> I assume you are referring to:
>>
>>>>                     if (compound_order(head) > PFN_SECTION_SHIFT) {
>>>>                             ret = -EBUSY;
>>>>                             break;
>>>>                     }
> 
> yes.
> 
>> I thought it was in case we stumble upon a gigantic page, and commit
>> (c8721bbbdd36 mm: memory-hotplug: enable memory hotplug to handle hugepage)
>> confirms it.
>>
>> But I am not really sure if the above condition would still hold on powerpc,
>> I wanted to check it but it is a bit more tricky than it is in x86_64 because
>> of the different hugetlb sizes.
>> Could it be that the above condition is not true, but still the order of that
>> hugetlb page goes beyond MAX_ORDER? It is something I have to check.

Well, commit 94310cbcaa3c ("mm/madvise: enable (soft|hard) offline of
HugeTLB pages at PGD level") should have allowed migration of gigantic
pages.  I believe it was added for 16GB pages on powerpc.  However, due
to subsequent changes I suspsect this no longer works.

> This check doesn't make much sense in principle. Why should we bail out
> based on a section size? We are offlining a pfn range. All that we care
> about is whether the hugetlb is migrateable.

Yes.  Do note that the do_migrate_range is only called from __offline_pages
with a start_pfn that was returned by scan_movable_pages.  scan_movable_pages
has the hugepage_migration_supported check for PageHuge pages.  So, it would
seem to be redundant to do another check in do_migrate_range.

-- 
Mike Kravetz