Re: [PATCH v8 05/10] mm: thp: enable thp migration in generic path

[Naoya Horiguchi]

On Sat, Jul 01, 2017 at 09:40:03AM -0400, Zi Yan wrote:
> From: Zi Yan <zi.yan@xxxxxxxxxxxxxx>
> 
> This patch adds thp migration's core code, including conversions
> between a PMD entry and a swap entry, setting PMD migration entry,
> removing PMD migration entry, and waiting on PMD migration entries.
> 
> This patch makes it possible to support thp migration.
> If you fail to allocate a destination page as a thp, you just split
> the source thp as we do now, and then enter the normal page migration.
> If you succeed to allocate destination thp, you enter thp migration.
> Subsequent patches actually enable thp migration for each caller of
> page migration by allowing its get_new_page() callback to
> allocate thps.
> 
> ChangeLog v1 -> v2:
> - support pte-mapped thp, doubly-mapped thp
> 
> Signed-off-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
> 
> ChangeLog v2 -> v3:
> - use page_vma_mapped_walk()
> - use pmdp_huge_clear_flush() instead of pmdp_huge_get_and_clear() in
>   set_pmd_migration_entry()
> 
> ChangeLog v3 -> v4:
> - factor out the code of removing pte pgtable page in zap_huge_pmd()
> 
> ChangeLog v4 -> v5:
> - remove unnecessary PTE-mapped THP code in remove_migration_pmd()
>   and set_pmd_migration_entry()
> - restructure the code in zap_huge_pmd() to avoid factoring out
>   the pte pgtable page code
> - in zap_huge_pmd(), check that PMD swap entries are migration entries
> - change author information
> 
> ChangeLog v5 -> v7
> - use macro to disable the code when thp migration is not enabled
> 
> ChangeLog v7 -> v8
> - use IS_ENABLED instead of macro to make code look clean in
>   zap_huge_pmd() and page_vma_mapped_walk()
> - remove BUILD_BUG() in pmd_to_swp_entry() and swp_entry_to_pmd() to
>   avoid compilation error
> - rename variable 'migration' to 'flush_needed' and invert the logic in
>   zap_huge_pmd() to make code more descriptive
> - use pmdp_invalidate() in set_pmd_migration_entry() to avoid race
>   with MADV_DONTNEED
> - remove unnecessary tlb flush in remove_migration_pmd()
> - add the missing migration flag check in page_vma_mapped_walk()
> 
> Signed-off-by: Zi Yan <zi.yan@xxxxxxxxxxxxxx>
> Cc: Kirill A. Shutemov <kirill.shutemov@xxxxxxxxxxxxxxx>
> ---
>  arch/x86/include/asm/pgtable_64.h |  2 +
>  include/linux/swapops.h           | 67 ++++++++++++++++++++++++++++++-
>  mm/huge_memory.c                  | 84 ++++++++++++++++++++++++++++++++++++---
>  mm/migrate.c                      | 32 ++++++++++++++-
>  mm/page_vma_mapped.c              | 18 +++++++--
>  mm/pgtable-generic.c              |  3 +-
>  mm/rmap.c                         | 13 ++++++
>  7 files changed, 207 insertions(+), 12 deletions(-)
> 
...

> diff --git a/mm/rmap.c b/mm/rmap.c
> index 91948fbbb0bb..b28f633cd569 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -1302,6 +1302,7 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
>  	bool ret = true;
>  	enum ttu_flags flags = (enum ttu_flags)arg;
>  
> +
>  	/* munlock has nothing to gain from examining un-locked vmas */
>  	if ((flags & TTU_MUNLOCK) && !(vma->vm_flags & VM_LOCKED))
>  		return true;
> @@ -1312,6 +1313,18 @@ static bool try_to_unmap_one(struct page *page, struct vm_area_struct *vma,
>  	}
>  
>  	while (page_vma_mapped_walk(&pvmw)) {
> +#ifdef CONFIG_ARCH_ENABLE_THP_MIGRATION
> +		/* PMD-mapped THP migration entry */
> +		if (flags & TTU_MIGRATION) {

My testing based on mmotm-2017-07-06-16-18 showed that migrating shmem thp
caused kernel crash. I don't think this is critical because that case is
just not-prepared yet. So in order to avoid the crash, please add
PageAnon(page) check here. This makes shmem thp migration just fail.

+			if (!PageAnon(page))
+				continue;

> +			if (!pvmw.pte && page) {

Just from curiosity, do we really need this page check?
try_to_unmap() always passes down the parameter 'page' to try_to_unmap_one()
via rmap_walk_* family, so I think we can assume page is always non-NULL.

Thanks,
Naoya Horiguchi

> +				VM_BUG_ON_PAGE(PageHuge(page) || !PageTransCompound(page),
> +						page);
> +				set_pmd_migration_entry(&pvmw, page);
> +				continue;
> +			}
> +		}
> +#endif
> +
>  		/*
>  		 * If the page is mlock()d, we cannot swap it out.
>  		 * If it's recently referenced (perhaps page_referenced
> -- 
> 2.11.0
> 
>