Re: [PATCH v7 3/7] mm,hugetlb: Clear HPageFreed outside of the lock

From: Michal Hocko <mhocko@suse.com>
To: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Oscar Salvador <osalvador@suse.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	David Hildenbrand <david@redhat.com>,
	Muchun Song <songmuchun@bytedance.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH v7 3/7] mm,hugetlb: Clear HPageFreed outside of the lock
Date: Wed, 14 Apr 2021 08:04:21 +0200	[thread overview]
Message-ID: <YHaF5efHcJJ71UP9@dhcp22.suse.cz> (raw)
In-Reply-To: <b8f36340-df56-1180-2a14-7b20cc1a0cda@oracle.com>

On Tue 13-04-21 14:19:03, Mike Kravetz wrote:
> On 4/13/21 6:23 AM, Michal Hocko wrote:
> > On Tue 13-04-21 12:47:43, Oscar Salvador wrote:
[...]
> > Or do we need it for giga pages which are not allocated by the page
> > allocator? If yes then moving it to prep_compound_gigantic_page would be
> > better.
> 
> I am pretty sure dynamically allocated giga pages have page->Private
> cleared as well.  It is not obvious, but the alloc_contig_range code
> used to put together giga pages will end up calling isolate_freepages_range
> that will call split_map_pages and then post_alloc_hook for each MAX_ORDER
> block.

Thanks for saving me from crawling that code.

> As mentioned, this is not obvious and I would hate to rely on this
> behavior not changing.

Thinking about it some more, having some (page granularity) allocator
not clearing page private would be a serious problem for anybody relying
on its state. So I am not sure this can change.

> > So should we just drop it here?
> 
> The only place where page->private may not be initialized is when we do
> allocations at boot time from memblock.  In this case, we will add the
> pages to the free list via put_page/free_huge_page so the appropriate
> flags will be cleared before anyone notices.

Pages allocated by the bootmem should be pre initialized from the boot,
no?

> I'm wondering if we should just do a set_page_private(page, 0) here in
> prep_new_huge_page since we now use that field for flags.  Or, is that
> overkill?

I would rather not duplicate the work done by underlying allocators. I
do not think other users of the allocator want to do the same so why
should hugetlb be any different.
-- 
Michal Hocko
SUSE Labs