Message-ID: <87pmbur99h.fsf@nvidia.com>
Date:   Wed, 04 Jan 2023 18:12:09 +1100
From:   Alistair Popple <apopple@...dia.com>
To:     Huang Ying <ying.huang@...el.com>
Cc:     Andrew Morton <akpm@...ux-foundation.org>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org, Zi Yan <ziy@...dia.com>,
        Yang Shi <shy828301@...il.com>,
        Baolin Wang <baolin.wang@...ux.alibaba.com>,
        Oscar Salvador <osalvador@...e.de>,
        Matthew Wilcox <willy@...radead.org>,
        Bharata B Rao <bharata@....com>,
        haoxin <xhao@...ux.alibaba.com>
Subject: Re: [PATCH 7/8] migrate_pages: share more code between _unmap and
 _move


Huang Ying <ying.huang@...el.com> writes:

> This is a code cleanup patch to reduce the duplicated code between the
> _unmap and _move stages of migrate_pages().  No functionality change
> is expected.
>
> Signed-off-by: "Huang, Ying" <ying.huang@...el.com>
> Cc: Zi Yan <ziy@...dia.com>
> Cc: Yang Shi <shy828301@...il.com>
> Cc: Baolin Wang <baolin.wang@...ux.alibaba.com>
> Cc: Oscar Salvador <osalvador@...e.de>
> Cc: Matthew Wilcox <willy@...radead.org>
> Cc: Bharata B Rao <bharata@....com>
> Cc: Alistair Popple <apopple@...dia.com>
> Cc: haoxin <xhao@...ux.alibaba.com>
> ---
>  mm/migrate.c | 208 ++++++++++++++++++++-------------------------------
>  1 file changed, 82 insertions(+), 126 deletions(-)
>
> diff --git a/mm/migrate.c b/mm/migrate.c
> index 70b987391296..70a40b8fee1f 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -1030,21 +1030,26 @@ static void __migrate_folio_extract(struct folio *dst,
>  static void migrate_folio_undo_src(struct folio *src,
>  				   int page_was_mapped,
>  				   struct anon_vma *anon_vma,
> +				   bool locked,
>  				   struct list_head *ret)
>  {
>  	if (page_was_mapped)
>  		remove_migration_ptes(src, src, false);
>  	if (anon_vma)
>  		put_anon_vma(anon_vma);
> -	folio_unlock(src);
> -	list_move_tail(&src->lru, ret);
> +	if (locked)
> +		folio_unlock(src);
> +	if (ret)
> +		list_move_tail(&src->lru, ret);
>  }
>  
>  static void migrate_folio_undo_dst(struct folio *dst,
> +				   bool locked,
>  				   free_page_t put_new_page,
>  				   unsigned long private)
>  {
> -	folio_unlock(dst);
> +	if (locked)
> +		folio_unlock(dst);
>  	if (put_new_page)
>  		put_new_page(&dst->page, private);
>  	else
> @@ -1068,14 +1073,44 @@ static void migrate_folio_done(struct folio *src,
>  		folio_put(src);
>  }
>  
> -static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
> -				 int force, bool force_lock, enum migrate_mode mode)
> +/* Obtain the lock on page, remove all ptes. */
> +static int migrate_folio_unmap(new_page_t get_new_page, free_page_t put_new_page,
> +			       unsigned long private, struct folio *src,
> +			       struct folio **dstp, int force, bool force_lock,
> +			       enum migrate_mode mode, enum migrate_reason reason,
> +			       struct list_head *ret)

Overall I think this should be refactored into some smaller, simpler
functions as the error handling and the giant switch statement in
migrate_pages_batch() is making my head hurt :-)

>  {
> -	int rc = -EAGAIN;
> +	struct folio *dst;
> +	int rc = MIGRATEPAGE_UNMAP;
> +	struct page *newpage = NULL;
>  	int page_was_mapped = 0;
>  	struct anon_vma *anon_vma = NULL;
>  	bool is_lru = !__PageMovable(&src->page);
> +	bool locked = false;
> +	bool dst_locked = false;
> +
> +	if (!thp_migration_supported() && folio_test_transhuge(src))
> +		return -ENOSYS;

This would be easier to follow if it was just moved to the caller and
the -ENOSYS switch case removed.

> +	if (folio_ref_count(src) == 1) {
> +		/* Folio was freed from under us. So we are done. */
> +		folio_clear_active(src);
> +		folio_clear_unevictable(src);
> +		/* free_pages_prepare() will clear PG_isolated. */
> +		list_del(&src->lru);
> +		migrate_folio_done(src, reason);
> +		return MIGRATEPAGE_SUCCESS;
> +	}

This is the only case that returns MIGRATEPAGE_SUCCESS so would also be
clearer if moved to the caller eliminating another switch case.

> +
> +	newpage = get_new_page(&src->page, private);
> +	if (!newpage)
> +		return -ENOMEM;
> +	dst = page_folio(newpage);
> +	*dstp = dst;
> +
> +	dst->private = NULL;

This could be moved until after the folio_test_writeback(), which might
make the split I suggest below easier.

>  
> +	rc = -EAGAIN;

We can just initialise rc to -EAGAIN.

>  	if (!folio_trylock(src)) {
>  		if (!force || mode == MIGRATE_ASYNC)
>  			goto out;
> @@ -1103,6 +1138,7 @@ static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
>  
>  		folio_lock(src);
>  	}
> +	locked = true;

A seperate helper for locking the folio would be better IMHO.

>  
>  	if (folio_test_writeback(src)) {
>  		/*
> @@ -1117,10 +1153,10 @@ static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
>  			break;
>  		default:
>  			rc = -EBUSY;
> -			goto out_unlock;
> +			goto out;
>  		}
>  		if (!force)
> -			goto out_unlock;
> +			goto out;
>  		folio_wait_writeback(src);
>  	}

This is the only path that return -EBUSY, so could be integrated into
the helper suggested above for locking the folio.
  
> @@ -1150,7 +1186,8 @@ static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
>  	 * This is much like races on refcount of oldpage: just don't BUG().
>  	 */
>  	if (unlikely(!folio_trylock(dst)))
> -		goto out_unlock;
> +		goto out;
> +	dst_locked = true;

So how about splitting migrate_folio_unmap() into two functions:

/*
 * Prepare a folio for migration by locking the source, ensuring
 * writeback is complete and allocating and locking a new destination
 * page.
 */
migrate_folio_prepare(new_page_t get_new_page, free_page_t put_new_page,
                      unsigned long private, struct folio *src,
	              struct folio **dstp, int force, bool force_lock)

migrate_folio_unmap(struct folio *src, struct folio *dst,
                    enum migrate_mode mode, enum migrate_reason reason)

Obviously we still have the various failure scenarios to deal with, but
I think it would be more readable if these were limited to undoing the
migrate_folio_prepare() step in the caller. I think the list
manipulation would also be more obvious if left to the caller.

>  	if (unlikely(!is_lru)) {
>  		__migrate_folio_record(dst, page_was_mapped, anon_vma);
> @@ -1172,7 +1209,7 @@ static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
>  	if (!src->mapping) {
>  		if (folio_test_private(src)) {
>  			try_to_free_buffers(src);
> -			goto out_unlock_both;
> +			goto out;
>  		}
>  	} else if (folio_mapped(src)) {
>  		/* Establish migration ptes */
> @@ -1187,75 +1224,27 @@ static int __migrate_folio_unmap(struct folio *src, struct folio *dst,
>  		return MIGRATEPAGE_UNMAP;

I reaslise this is pre-existing but the mixing of setting rc or
returning codes directly is a bit hard to follow.

>  	}
>  
> -
> -	if (page_was_mapped)
> -		remove_migration_ptes(src, src, false);
> -
> -out_unlock_both:
> -	folio_unlock(dst);
> -out_unlock:
> -	/* Drop an anon_vma reference if we took one */
> -	if (anon_vma)
> -		put_anon_vma(anon_vma);
> -	folio_unlock(src);
>  out:
> -
> -	return rc;
> -}
> -
> -/* Obtain the lock on page, remove all ptes. */
> -static int migrate_folio_unmap(new_page_t get_new_page, free_page_t put_new_page,
> -			       unsigned long private, struct folio *src,
> -			       struct folio **dstp, int force, bool force_lock,
> -			       enum migrate_mode mode, enum migrate_reason reason,
> -			       struct list_head *ret)
> -{
> -	struct folio *dst;
> -	int rc = MIGRATEPAGE_UNMAP;
> -	struct page *newpage = NULL;
> -
> -	if (!thp_migration_supported() && folio_test_transhuge(src))
> -		return -ENOSYS;
> -
> -	if (folio_ref_count(src) == 1) {
> -		/* Folio was freed from under us. So we are done. */
> -		folio_clear_active(src);
> -		folio_clear_unevictable(src);
> -		/* free_pages_prepare() will clear PG_isolated. */
> -		list_del(&src->lru);
> -		migrate_folio_done(src, reason);
> -		return MIGRATEPAGE_SUCCESS;
> -	}
> -
> -	newpage = get_new_page(&src->page, private);
> -	if (!newpage)
> -		return -ENOMEM;
> -	dst = page_folio(newpage);
> -	*dstp = dst;
> -
> -	dst->private = NULL;
> -	rc = __migrate_folio_unmap(src, dst, force, force_lock, mode);
> -	if (rc == MIGRATEPAGE_UNMAP)
> -		return rc;
> -
>  	/*
>  	 * A page that has not been migrated will have kept its
>  	 * references and be restored.
>  	 */
>  	/* restore the folio to right list. */
> -	if (rc != -EAGAIN && rc != -EDEADLOCK)
> -		list_move_tail(&src->lru, ret);
> +	if (rc == -EAGAIN || rc == -EDEADLOCK)
> +		ret = NULL;
>  
> -	if (put_new_page)
> -		put_new_page(&dst->page, private);
> -	else
> -		folio_put(dst);
> +	migrate_folio_undo_src(src, page_was_mapped, anon_vma, locked, ret);
> +	if (dst)
> +		migrate_folio_undo_dst(dst, dst_locked, put_new_page, private);
>  
>  	return rc;
>  }
>  
> -static int __migrate_folio_move(struct folio *src, struct folio *dst,
> -				enum migrate_mode mode)
> +/* Migrate the folio to the newly allocated folio in dst. */
> +static int migrate_folio_move(free_page_t put_new_page, unsigned long private,
> +			      struct folio *src, struct folio *dst,
> +			      enum migrate_mode mode, enum migrate_reason reason,
> +			      struct list_head *ret)
>  {
>  	int rc;
>  	int page_was_mapped = 0;
> @@ -1264,9 +1253,10 @@ static int __migrate_folio_move(struct folio *src, struct folio *dst,
>  	__migrate_folio_extract(dst, &page_was_mapped, &anon_vma);
>  
>  	rc = move_to_new_folio(dst, src, mode);
> +	if (rc)
> +		goto out;
>  
> -	if (rc != -EAGAIN)
> -		list_del(&dst->lru);
> +	list_del(&dst->lru);
>  	/*
>  	 * When successful, push dst to LRU immediately: so that if it
>  	 * turns out to be an mlocked page, remove_migration_ptes() will
> @@ -1276,74 +1266,40 @@ static int __migrate_folio_move(struct folio *src, struct folio *dst,
>  	 * unsuccessful, and other cases when a page has been temporarily
>  	 * isolated from the unevictable LRU: but this case is the easiest.
>  	 */
> -	if (rc == MIGRATEPAGE_SUCCESS) {
> -		folio_add_lru(dst);
> -		if (page_was_mapped)
> -			lru_add_drain();
> -	}
> -
> -	if (rc == -EAGAIN) {
> -		__migrate_folio_record(dst, page_was_mapped, anon_vma);
> -		return rc;
> -	}
> -
> +	folio_add_lru(dst);
>  	if (page_was_mapped)
> -		remove_migration_ptes(src,
> -			rc == MIGRATEPAGE_SUCCESS ? dst : src, false);
> +		lru_add_drain();
>  
> +	if (page_was_mapped)
> +		remove_migration_ptes(src, dst, false);
>  	folio_unlock(dst);
> -	/* Drop an anon_vma reference if we took one */
> -	if (anon_vma)
> -		put_anon_vma(anon_vma);
> -	folio_unlock(src);
> +	set_page_owner_migrate_reason(&dst->page, reason);
>  	/*
>  	 * If migration is successful, decrease refcount of dst,
>  	 * which will not free the page because new page owner increased
>  	 * refcounter.
>  	 */
> -	if (rc == MIGRATEPAGE_SUCCESS)
> -		folio_put(dst);
> -
> -	return rc;
> -}
> -
> -/* Migrate the folio to the newly allocated folio in dst. */
> -static int migrate_folio_move(free_page_t put_new_page, unsigned long private,
> -			      struct folio *src, struct folio *dst,
> -			      enum migrate_mode mode, enum migrate_reason reason,
> -			      struct list_head *ret)
> -{
> -	int rc;
> -
> -	rc = __migrate_folio_move(src, dst, mode);
> -	if (rc == MIGRATEPAGE_SUCCESS)
> -		set_page_owner_migrate_reason(&dst->page, reason);
> -
> -	if (rc != -EAGAIN) {
> -		/*
> -		 * A folio that has been migrated has all references
> -		 * removed and will be freed. A folio that has not been
> -		 * migrated will have kept its references and be restored.
> -		 */
> -		list_del(&src->lru);
> -	}
> +	folio_put(dst);
>  
>  	/*
> -	 * If migration is successful, releases reference grabbed during
> -	 * isolation. Otherwise, restore the folio to right list unless
> -	 * we want to retry.
> +	 * A page that has been migrated has all references removed
> +	 * and will be freed.
>  	 */
> -	if (rc == MIGRATEPAGE_SUCCESS) {
> -		migrate_folio_done(src, reason);
> -	} else if (rc != -EAGAIN) {
> -		list_add_tail(&src->lru, ret);
> +	list_del(&src->lru);
> +	migrate_folio_undo_src(src, 0, anon_vma, true, NULL);
> +	migrate_folio_done(src, reason);
>  
> -		if (put_new_page)
> -			put_new_page(&dst->page, private);
> -		else
> -			folio_put(dst);
> +	return rc;
> +out:
> +	if (rc == -EAGAIN) {
> +		__migrate_folio_record(dst, page_was_mapped, anon_vma);
> +		return rc;
>  	}
>  
> +	migrate_folio_undo_src(src, page_was_mapped, anon_vma, true, ret);
> +	list_del(&dst->lru);
> +	migrate_folio_undo_dst(dst, true, put_new_page, private);
> +
>  	return rc;
>  }
>  
> @@ -1849,9 +1805,9 @@ static int migrate_pages_batch(struct list_head *from, new_page_t get_new_page,
>  
>  		__migrate_folio_extract(dst, &page_was_mapped, &anon_vma);
>  		migrate_folio_undo_src(folio, page_was_mapped, anon_vma,
> -				       ret_folios);
> +				       true, ret_folios);
>  		list_del(&dst->lru);
> -		migrate_folio_undo_dst(dst, put_new_page, private);
> +		migrate_folio_undo_dst(dst, true, put_new_page, private);
>  		dst = dst2;
>  		dst2 = list_next_entry(dst, lru);
>  	}

Powered by blists - more mailing lists