Re: [PATCH v2 1/2] mm: Allow the page cache to allocate large pages

[Song Liu <songliubraving@fb.com>]

> On Jul 29, 2019, at 2:09 PM, William Kucharski <william.kucharski@oracle.com> wrote:
> 

I guess we need "From: Matthew Wilcox <willy@infradead.org>" here?

> Add an order field to __page_cache_alloc() to allow for the allocation
> of large memory page page cache entries.
> 
> Signed-off-by: Matthew Wilcox <willy@infradead.org>
> Signed-off-by: William Kucharski <william.kucharski@oracle.com>
> Reported-by: kbuild test robot <lkp@intel.com>
> ---
> fs/afs/dir.c            |  2 +-
> fs/btrfs/compression.c  |  2 +-
> fs/cachefiles/rdwr.c    |  4 ++--
> fs/ceph/addr.c          |  2 +-
> fs/ceph/file.c          |  2 +-
> include/linux/pagemap.h | 13 +++++++++----
> mm/filemap.c            | 25 +++++++++++++------------
> mm/readahead.c          |  2 +-
> net/ceph/pagelist.c     |  4 ++--
> net/ceph/pagevec.c      |  2 +-
> 10 files changed, 32 insertions(+), 26 deletions(-)
> 
> diff --git a/fs/afs/dir.c b/fs/afs/dir.c
> index e640d67274be..0a392214f71e 100644
> --- a/fs/afs/dir.c
> +++ b/fs/afs/dir.c
> @@ -274,7 +274,7 @@ static struct afs_read *afs_read_dir(struct afs_vnode *dvnode, struct key *key)
> 				afs_stat_v(dvnode, n_inval);
> 
> 			ret = -ENOMEM;
> -			req->pages[i] = __page_cache_alloc(gfp);
> +			req->pages[i] = __page_cache_alloc(gfp, 0);
> 			if (!req->pages[i])
> 				goto error;
> 			ret = add_to_page_cache_lru(req->pages[i],
> diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c
> index 60c47b417a4b..5280e7477b7e 100644
> --- a/fs/btrfs/compression.c
> +++ b/fs/btrfs/compression.c
> @@ -466,7 +466,7 @@ static noinline int add_ra_bio_pages(struct inode *inode,
> 		}
> 
> 		page = __page_cache_alloc(mapping_gfp_constraint(mapping,
> -								 ~__GFP_FS));
> +								 ~__GFP_FS), 0);
> 		if (!page)
> 			break;
> 
> diff --git a/fs/cachefiles/rdwr.c b/fs/cachefiles/rdwr.c
> index 44a3ce1e4ce4..11d30212745f 100644
> --- a/fs/cachefiles/rdwr.c
> +++ b/fs/cachefiles/rdwr.c
> @@ -259,7 +259,7 @@ static int cachefiles_read_backing_file_one(struct cachefiles_object *object,
> 			goto backing_page_already_present;
> 
> 		if (!newpage) {
> -			newpage = __page_cache_alloc(cachefiles_gfp);
> +			newpage = __page_cache_alloc(cachefiles_gfp, 0);
> 			if (!newpage)
> 				goto nomem_monitor;
> 		}
> @@ -495,7 +495,7 @@ static int cachefiles_read_backing_file(struct cachefiles_object *object,
> 				goto backing_page_already_present;
> 
> 			if (!newpage) {
> -				newpage = __page_cache_alloc(cachefiles_gfp);
> +				newpage = __page_cache_alloc(cachefiles_gfp, 0);
> 				if (!newpage)
> 					goto nomem;
> 			}
> diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
> index e078cc55b989..bcb41fbee533 100644
> --- a/fs/ceph/addr.c
> +++ b/fs/ceph/addr.c
> @@ -1707,7 +1707,7 @@ int ceph_uninline_data(struct file *filp, struct page *locked_page)
> 		if (len > PAGE_SIZE)
> 			len = PAGE_SIZE;
> 	} else {
> -		page = __page_cache_alloc(GFP_NOFS);
> +		page = __page_cache_alloc(GFP_NOFS, 0);
> 		if (!page) {
> 			err = -ENOMEM;
> 			goto out;
> diff --git a/fs/ceph/file.c b/fs/ceph/file.c
> index 685a03cc4b77..ae58d7c31aa4 100644
> --- a/fs/ceph/file.c
> +++ b/fs/ceph/file.c
> @@ -1305,7 +1305,7 @@ static ssize_t ceph_read_iter(struct kiocb *iocb, struct iov_iter *to)
> 		struct page *page = NULL;
> 		loff_t i_size;
> 		if (retry_op == READ_INLINE) {
> -			page = __page_cache_alloc(GFP_KERNEL);
> +			page = __page_cache_alloc(GFP_KERNEL, 0);
> 			if (!page)
> 				return -ENOMEM;
> 		}
> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
> index c7552459a15f..e9004e3cb6a3 100644
> --- a/include/linux/pagemap.h
> +++ b/include/linux/pagemap.h
> @@ -208,17 +208,17 @@ static inline int page_cache_add_speculative(struct page *page, int count)
> }
> 
> #ifdef CONFIG_NUMA
> -extern struct page *__page_cache_alloc(gfp_t gfp);
> +extern struct page *__page_cache_alloc(gfp_t gfp, unsigned int order);
> #else
> -static inline struct page *__page_cache_alloc(gfp_t gfp)
> +static inline struct page *__page_cache_alloc(gfp_t gfp, unsigned int order)
> {
> -	return alloc_pages(gfp, 0);
> +	return alloc_pages(gfp, order);
> }
> #endif
> 
> static inline struct page *page_cache_alloc(struct address_space *x)
> {
> -	return __page_cache_alloc(mapping_gfp_mask(x));
> +	return __page_cache_alloc(mapping_gfp_mask(x), 0);
> }
> 
> static inline gfp_t readahead_gfp_mask(struct address_space *x)
> @@ -240,6 +240,11 @@ pgoff_t page_cache_prev_miss(struct address_space *mapping,
> #define FGP_NOFS		0x00000010
> #define FGP_NOWAIT		0x00000020
> #define FGP_FOR_MMAP		0x00000040
> +/* If you add more flags, increment FGP_ORDER_SHIFT */
> +#define	FGP_ORDER_SHIFT		7
> +#define	FGP_PMD			((PMD_SHIFT - PAGE_SHIFT) << FGP_ORDER_SHIFT)
> +#define	FGP_PUD			((PUD_SHIFT - PAGE_SHIFT) << FGP_ORDER_SHIFT)
> +#define	fgp_get_order(fgp)	((fgp) >> FGP_ORDER_SHIFT)

This looks like we want support order up to 25 (32 - 7). I guess we don't 
need that many. How about we specify the highest order to support here? 

Also, fgp_flags is signed int, so we need to make sure fgp_flags is not
negative. 

> 
> struct page *pagecache_get_page(struct address_space *mapping, pgoff_t offset,
> 		int fgp_flags, gfp_t cache_gfp_mask);
> diff --git a/mm/filemap.c b/mm/filemap.c
> index d0cf700bf201..a96092243fc4 100644
> --- a/mm/filemap.c
> +++ b/mm/filemap.c
> @@ -954,7 +954,7 @@ int add_to_page_cache_lru(struct page *page, struct address_space *mapping,
> EXPORT_SYMBOL_GPL(add_to_page_cache_lru);
> 
> #ifdef CONFIG_NUMA
> -struct page *__page_cache_alloc(gfp_t gfp)
> +struct page *__page_cache_alloc(gfp_t gfp, unsigned int order)
> {
> 	int n;
> 	struct page *page;
> @@ -964,12 +964,12 @@ struct page *__page_cache_alloc(gfp_t gfp)
> 		do {
> 			cpuset_mems_cookie = read_mems_allowed_begin();
> 			n = cpuset_mem_spread_node();
> -			page = __alloc_pages_node(n, gfp, 0);
> +			page = __alloc_pages_node(n, gfp, order);
> 		} while (!page && read_mems_allowed_retry(cpuset_mems_cookie));
> 
> 		return page;
> 	}
> -	return alloc_pages(gfp, 0);
> +	return alloc_pages(gfp, order);
> }
> EXPORT_SYMBOL(__page_cache_alloc);
> #endif
> @@ -1597,12 +1597,12 @@ EXPORT_SYMBOL(find_lock_entry);
>  * pagecache_get_page - find and get a page reference
>  * @mapping: the address_space to search
>  * @offset: the page index
> - * @fgp_flags: PCG flags
> + * @fgp_flags: FGP flags
>  * @gfp_mask: gfp mask to use for the page cache data page allocation
>  *
>  * Looks up the page cache slot at @mapping & @offset.
>  *
> - * PCG flags modify how the page is returned.
> + * FGP flags modify how the page is returned.
>  *
>  * @fgp_flags can be:
>  *
> @@ -1615,6 +1615,7 @@ EXPORT_SYMBOL(find_lock_entry);
>  * - FGP_FOR_MMAP: Similar to FGP_CREAT, only we want to allow the caller to do
>  *   its own locking dance if the page is already in cache, or unlock the page
>  *   before returning if we had to add the page to pagecache.
> + * - FGP_PMD: If FGP_CREAT is specified, attempt to allocate a PMD-sized page.
>  *
>  * If FGP_LOCK or FGP_CREAT are specified then the function may sleep even
>  * if the GFP flags specified for FGP_CREAT are atomic.
> @@ -1660,12 +1661,13 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t offset,
> no_page:
> 	if (!page && (fgp_flags & FGP_CREAT)) {
> 		int err;
> -		if ((fgp_flags & FGP_WRITE) && mapping_cap_account_dirty(mapping))
> +		if ((fgp_flags & FGP_WRITE) &&
> +			mapping_cap_account_dirty(mapping))
> 			gfp_mask |= __GFP_WRITE;
> 		if (fgp_flags & FGP_NOFS)
> 			gfp_mask &= ~__GFP_FS;
> 
> -		page = __page_cache_alloc(gfp_mask);
> +		page = __page_cache_alloc(gfp_mask, fgp_get_order(fgp_flags));
> 		if (!page)
> 			return NULL;
> 
> @@ -2802,15 +2804,14 @@ static struct page *wait_on_page_read(struct page *page)
> static struct page *do_read_cache_page(struct address_space *mapping,
> 				pgoff_t index,
> 				int (*filler)(void *, struct page *),
> -				void *data,
> -				gfp_t gfp)
> +				void *data, unsigned int order, gfp_t gfp)
> {
> 	struct page *page;
> 	int err;
> repeat:
> 	page = find_get_page(mapping, index);
> 	if (!page) {
> -		page = __page_cache_alloc(gfp);
> +		page = __page_cache_alloc(gfp, order);
> 		if (!page)
> 			return ERR_PTR(-ENOMEM);
> 		err = add_to_page_cache_lru(page, mapping, index, gfp);
> @@ -2917,7 +2918,7 @@ struct page *read_cache_page(struct address_space *mapping,
> 				int (*filler)(void *, struct page *),
> 				void *data)
> {
> -	return do_read_cache_page(mapping, index, filler, data,
> +	return do_read_cache_page(mapping, index, filler, data, 0,
> 			mapping_gfp_mask(mapping));
> }
> EXPORT_SYMBOL(read_cache_page);
> @@ -2939,7 +2940,7 @@ struct page *read_cache_page_gfp(struct address_space *mapping,
> 				pgoff_t index,
> 				gfp_t gfp)
> {
> -	return do_read_cache_page(mapping, index, NULL, NULL, gfp);
> +	return do_read_cache_page(mapping, index, NULL, NULL, 0, gfp);
> }
> EXPORT_SYMBOL(read_cache_page_gfp);
> 
> diff --git a/mm/readahead.c b/mm/readahead.c
> index 2fe72cd29b47..954760a612ea 100644
> --- a/mm/readahead.c
> +++ b/mm/readahead.c
> @@ -193,7 +193,7 @@ unsigned int __do_page_cache_readahead(struct address_space *mapping,
> 			continue;
> 		}
> 
> -		page = __page_cache_alloc(gfp_mask);
> +		page = __page_cache_alloc(gfp_mask, 0);
> 		if (!page)
> 			break;
> 		page->index = page_offset;
> diff --git a/net/ceph/pagelist.c b/net/ceph/pagelist.c
> index 65e34f78b05d..0c3face908dc 100644
> --- a/net/ceph/pagelist.c
> +++ b/net/ceph/pagelist.c
> @@ -56,7 +56,7 @@ static int ceph_pagelist_addpage(struct ceph_pagelist *pl)
> 	struct page *page;
> 
> 	if (!pl->num_pages_free) {
> -		page = __page_cache_alloc(GFP_NOFS);
> +		page = __page_cache_alloc(GFP_NOFS, 0);
> 	} else {
> 		page = list_first_entry(&pl->free_list, struct page, lru);
> 		list_del(&page->lru);
> @@ -107,7 +107,7 @@ int ceph_pagelist_reserve(struct ceph_pagelist *pl, size_t space)
> 	space = (space + PAGE_SIZE - 1) >> PAGE_SHIFT;   /* conv to num pages */
> 
> 	while (space > pl->num_pages_free) {
> -		struct page *page = __page_cache_alloc(GFP_NOFS);
> +		struct page *page = __page_cache_alloc(GFP_NOFS, 0);
> 		if (!page)
> 			return -ENOMEM;
> 		list_add_tail(&page->lru, &pl->free_list);
> diff --git a/net/ceph/pagevec.c b/net/ceph/pagevec.c
> index 64305e7056a1..1d07e639216d 100644
> --- a/net/ceph/pagevec.c
> +++ b/net/ceph/pagevec.c
> @@ -45,7 +45,7 @@ struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags)
> 	if (!pages)
> 		return ERR_PTR(-ENOMEM);
> 	for (i = 0; i < num_pages; i++) {
> -		pages[i] = __page_cache_alloc(flags);
> +		pages[i] = __page_cache_alloc(flags, 0);
> 		if (pages[i] == NULL) {
> 			ceph_release_page_vector(pages, i);
> 			return ERR_PTR(-ENOMEM);
> -- 
> 2.21.0
>