[GIT UPDATE] pagecache tree

[GIT UPDATE] pagecache tree

[Matthew Wilcox]

I've just finished going through my email from last week, fixing up
various patches and adding new R-b lines.  You can find the git tree
here: git://git.infradead.org/users/willy/pagecache.git for-next
or https://git.infradead.org/users/willy/pagecache.git/shortlog/refs/heads/for-next

I'll send out some separate series as a reply to this mail, but here's
the shortlog:

Matthew Wilcox (Oracle) (105):
      scsicam: Fix use of page cache
      ext4: Use page_symlink() instead of __page_symlink()
      namei: Merge page_symlink() and __page_symlink()
      namei: Convert page_symlink() to use memalloc_nofs_save()
      f2fs: Convert f2fs_grab_cache_page() to use scoped memory APIs
      ext4: Allow GFP_FS allocations in ext4_da_convert_inline_data_to_extent()
      ext4: Use scoped memory API in mext_page_double_lock()
      ext4: Use scoped memory APIs in ext4_da_write_begin()
      ext4: Use scoped memory APIs in ext4_write_begin()
      fs: Remove AOP_FLAG_NOFS
      fs: Remove aop_flags parameter from netfs_write_begin()
      fs: Remove aop flags parameter from block_write_begin()
      fs: Remove aop flags parameter from cont_write_begin()
      fs: Remove aop flags parameter from grab_cache_page_write_begin()
      fs: Remove aop flags parameter from nobh_write_begin()
      fs: Remove flags parameter from aops->write_begin
      buffer: Call aops write_begin() and write_end() directly
      namei: Call aops write_begin() and write_end() directly
      ntfs3: Call ntfs_write_begin() and ntfs_write_end() directly
      ntfs3: Remove fsdata parameter from ntfs_extend_initialized_size()
      hfs: Call hfs_write_begin() and generic_write_end() directly
      hfsplus: Call hfsplus_write_begin() and generic_write_end() directly
      ext4: Call aops write_begin() and write_end() directly
      f2fs: Call aops write_begin() and write_end() directly
      i915: Call aops write_begin() and write_end() directly
      fs: Remove pagecache_write_begin() and pagecache_write_end()
      filemap: Update the folio_lock documentation
      filemap: Update the folio_mark_dirty documentation
      readahead: Use a folio in read_pages()
      fs: Convert is_dirty_writeback() to take a folio
      mm/readahead: Convert page_cache_async_readahead to take a folio
      buffer: Rewrite nobh_truncate_page() to use folios
      fs: Introduce aops->read_folio
      fs: Add read_folio documentation
      fs: Convert netfs_readpage to netfs_read_folio
      fs: Convert iomap_readpage to iomap_read_folio
      fs: Convert block_read_full_page() to block_read_full_folio()
      fs: Convert mpage_readpage to mpage_read_folio
      fs: Convert simple_readpage to simple_read_folio
      affs: Convert affs to read_folio
      afs: Convert afs_symlink_readpage to afs_symlink_read_folio
      befs: Convert befs to read_folio
      btrfs: Convert btrfs to read_folio
      cifs: Convert cifs to read_folio
      coda: Convert coda to read_folio
      cramfs: Convert cramfs to read_folio
      ecryptfs: Convert ecryptfs to read_folio
      efs: Convert efs symlinks to read_folio
      erofs: Convert erofs zdata to read_folio
      ext4: Convert ext4 to read_folio
      f2fs: Convert f2fs to read_folio
      freevxfs: Convert vxfs_immed to read_folio
      fuse: Convert fuse to read_folio
      hostfs: Convert hostfs to read_folio
      hpfs: Convert symlinks to read_folio
      isofs: Convert symlinks and zisofs to read_folio
      jffs2: Convert jffs2 to read_folio
      jfs: Convert metadata pages to read_folio
      nfs: Convert nfs to read_folio
      ntfs: Convert ntfs to read_folio
      ocfs2: Convert ocfs2 to read_folio
      orangefs: Convert orangefs to read_folio
      romfs: Convert romfs to read_folio
      squashfs: Convert squashfs to read_folio
      ubifs: Convert ubifs to read_folio
      udf: Convert adinicb and symlinks to read_folio
      vboxsf: Convert vboxsf to read_folio
      mm: Convert swap_readpage to call read_folio instead of readpage
      mm,fs: Remove aops->readpage
      jffs2: Pass the file pointer to jffs2_do_readpage_unlock()
      nfs: Pass the file pointer to nfs_symlink_filler()
      fs: Change the type of filler_t
      mm/filemap: Hoist filler_t decision to the top of do_read_cache_folio()
      fs: Add aops->release_folio
      iomap: Convert to release_folio
      9p: Convert to release_folio
      afs: Convert to release_folio
      btrfs: Convert to release_folio
      ceph: Convert to release_folio
      cifs: Convert to release_folio
      erofs: Convert to release_folio
      ext4: Convert to release_folio
      f2fs: Convert to release_folio
      gfs2: Convert to release_folio
      hfs: Convert to release_folio
      hfsplus: Convert to release_folio
      jfs: Convert to release_folio
      nfs: Convert to release_folio
      nilfs2: Remove comment about releasepage
      ocfs2: Convert to release_folio
      orangefs: Convert to release_folio
      reiserfs: Convert to release_folio
      ubifs: Convert to release_folio
      fs: Remove last vestiges of releasepage
      reiserfs: Convert release_buffer_page() to use a folio
      jbd2: Convert jbd2_journal_try_to_free_buffers to take a folio
      jbd2: Convert release_buffer_page() to use a folio
      fs: Change try_to_free_buffers() to take a folio
      fs: Convert drop_buffers() to use a folio
      fs: Add free_folio address space operation
      orangefs: Convert to free_folio
      nfs: Convert to free_folio
      secretmem: Convert to free_folio
      fs: Remove aops->freepage
      Appoint myself page cache maintainer

Miaohe Lin (1):
      filemap: Remove obsolete comment in lock_page

[PATCH] Appoint myself page cache maintainer

[Matthew Wilcox (Oracle)]

This feels like a sufficiently distinct area of responsibility to be
worth separating out from both MM and VFS.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 MAINTAINERS | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 9d47c5e7c6ae..5871ec2e1b3e 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -14833,6 +14833,18 @@ F:	Documentation/core-api/padata.rst
 F:	include/linux/padata.h
 F:	kernel/padata.c
 
+PAGE CACHE
+M:	Matthew Wilcox (Oracle) <willy@infradead.org>
+L:	linux-fsdevel@vger.kernel.org
+S:	Supported
+T:	git git://git.infradead.org/users/willy/pagecache.git
+F:	Documentation/filesystems/locking.rst
+F:	Documentation/filesystems/vfs.rst
+F:	include/linux/pagemap.h
+F:	mm/filemap.c
+F:	mm/page-writeback.c
+F:	mm/readahead.c
+
 PAGE POOL
 M:	Jesper Dangaard Brouer <hawk@kernel.org>
 M:	Ilias Apalodimas <ilias.apalodimas@linaro.org>
-- 
2.34.1

Re: [PATCH] Appoint myself page cache maintainer

[Dave Chinner]

On Sun, May 08, 2022 at 09:28:48PM +0100, Matthew Wilcox (Oracle) wrote:
> This feels like a sufficiently distinct area of responsibility to be
> worth separating out from both MM and VFS.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> ---
>  MAINTAINERS | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 9d47c5e7c6ae..5871ec2e1b3e 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -14833,6 +14833,18 @@ F:	Documentation/core-api/padata.rst
>  F:	include/linux/padata.h
>  F:	kernel/padata.c
>  
> +PAGE CACHE
> +M:	Matthew Wilcox (Oracle) <willy@infradead.org>
> +L:	linux-fsdevel@vger.kernel.org
> +S:	Supported
> +T:	git git://git.infradead.org/users/willy/pagecache.git
> +F:	Documentation/filesystems/locking.rst
> +F:	Documentation/filesystems/vfs.rst
> +F:	include/linux/pagemap.h
> +F:	mm/filemap.c
> +F:	mm/page-writeback.c
> +F:	mm/readahead.c

A big thumbs up from me.

Reviewed-by: Dave Chinner <dchinner@redhat.com>

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

Re: [PATCH] Appoint myself page cache maintainer

[Darrick J. Wong]

On Mon, May 09, 2022 at 09:16:44AM +1000, Dave Chinner wrote:
> On Sun, May 08, 2022 at 09:28:48PM +0100, Matthew Wilcox (Oracle) wrote:
> > This feels like a sufficiently distinct area of responsibility to be
> > worth separating out from both MM and VFS.
> > 
> > Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Yes plz,
Reviewed-by: Darrick J. Wong <djwong@kernel.org>

--D

> > ---
> >  MAINTAINERS | 12 ++++++++++++
> >  1 file changed, 12 insertions(+)
> > 
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index 9d47c5e7c6ae..5871ec2e1b3e 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -14833,6 +14833,18 @@ F:	Documentation/core-api/padata.rst
> >  F:	include/linux/padata.h
> >  F:	kernel/padata.c
> >  
> > +PAGE CACHE
> > +M:	Matthew Wilcox (Oracle) <willy@infradead.org>
> > +L:	linux-fsdevel@vger.kernel.org
> > +S:	Supported
> > +T:	git git://git.infradead.org/users/willy/pagecache.git
> > +F:	Documentation/filesystems/locking.rst
> > +F:	Documentation/filesystems/vfs.rst
> > +F:	include/linux/pagemap.h
> > +F:	mm/filemap.c
> > +F:	mm/page-writeback.c
> > +F:	mm/readahead.c
> 
> A big thumbs up from me.
> 
> Reviewed-by: Dave Chinner <dchinner@redhat.com>
> 
> Cheers,
> 
> Dave.
> -- 
> Dave Chinner
> david@fromorbit.com

Re: [PATCH] Appoint myself page cache maintainer

[Jeff Layton]

On Mon, 2022-05-09 at 09:16 +1000, Dave Chinner wrote:
> On Sun, May 08, 2022 at 09:28:48PM +0100, Matthew Wilcox (Oracle) wrote:
> > This feels like a sufficiently distinct area of responsibility to be
> > worth separating out from both MM and VFS.
> > 
> > Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> > ---
> >  MAINTAINERS | 12 ++++++++++++
> >  1 file changed, 12 insertions(+)
> > 
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index 9d47c5e7c6ae..5871ec2e1b3e 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -14833,6 +14833,18 @@ F:	Documentation/core-api/padata.rst
> >  F:	include/linux/padata.h
> >  F:	kernel/padata.c
> >  
> > +PAGE CACHE
> > +M:	Matthew Wilcox (Oracle) <willy@infradead.org>
> > +L:	linux-fsdevel@vger.kernel.org
> > +S:	Supported
> > +T:	git git://git.infradead.org/users/willy/pagecache.git
> > +F:	Documentation/filesystems/locking.rst
> > +F:	Documentation/filesystems/vfs.rst
> > +F:	include/linux/pagemap.h
> > +F:	mm/filemap.c
> > +F:	mm/page-writeback.c
> > +F:	mm/readahead.c
> 
> A big thumbs up from me.
> 
> Reviewed-by: Dave Chinner <dchinner@redhat.com>
> 

Ditto. Thanks for stepping up!

Reviewed-by: Jeff Layton <jlayton@kernel.org>

Re: [PATCH] Appoint myself page cache maintainer

[Christian Brauner]

On Sun, May 08, 2022 at 09:28:48PM +0100, Matthew Wilcox wrote:
> This feels like a sufficiently distinct area of responsibility to be
> worth separating out from both MM and VFS.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> ---

+1!
Acked-by: Christian Brauner (Microsoft) <brauner@kernel.org>

Re: [PATCH] Appoint myself page cache maintainer

[Vlastimil Babka]

On 5/8/22 22:28, Matthew Wilcox (Oracle) wrote:
> This feels like a sufficiently distinct area of responsibility to be
> worth separating out from both MM and VFS.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Acked-by: Vlastimil Babka <vbabka@suse.cz>

> ---
>  MAINTAINERS | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 9d47c5e7c6ae..5871ec2e1b3e 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -14833,6 +14833,18 @@ F:	Documentation/core-api/padata.rst
>  F:	include/linux/padata.h
>  F:	kernel/padata.c
>  
> +PAGE CACHE
> +M:	Matthew Wilcox (Oracle) <willy@infradead.org>
> +L:	linux-fsdevel@vger.kernel.org
> +S:	Supported
> +T:	git git://git.infradead.org/users/willy/pagecache.git
> +F:	Documentation/filesystems/locking.rst
> +F:	Documentation/filesystems/vfs.rst
> +F:	include/linux/pagemap.h
> +F:	mm/filemap.c
> +F:	mm/page-writeback.c
> +F:	mm/readahead.c
> +
>  PAGE POOL
>  M:	Jesper Dangaard Brouer <hawk@kernel.org>
>  M:	Ilias Apalodimas <ilias.apalodimas@linaro.org>

[PATCH] scsicam: Fix use of page cache

[Matthew Wilcox (Oracle)]

Convert scsicam to use a folio instead of a page.  There is no need to
check the error flag here; read_cache_folio() will return -EIO if the
folio cannot be read correctly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 drivers/scsi/scsicam.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/scsi/scsicam.c b/drivers/scsi/scsicam.c
index acdc0aceca5e..e2c7d8ef205f 100644
--- a/drivers/scsi/scsicam.c
+++ b/drivers/scsi/scsicam.c
@@ -34,15 +34,14 @@ unsigned char *scsi_bios_ptable(struct block_device *dev)
 {
 	struct address_space *mapping = bdev_whole(dev)->bd_inode->i_mapping;
 	unsigned char *res = NULL;
-	struct page *page;
+	struct folio *folio;
 
-	page = read_mapping_page(mapping, 0, NULL);
-	if (IS_ERR(page))
+	folio = read_mapping_folio(mapping, 0, NULL);
+	if (IS_ERR(folio))
 		return NULL;
 
-	if (!PageError(page))
-		res = kmemdup(page_address(page) + 0x1be, 66, GFP_KERNEL);
-	put_page(page);
+	res = kmemdup(folio_address(folio) + 0x1be, 66, GFP_KERNEL);
+	folio_put(folio);
 	return res;
 }
 EXPORT_SYMBOL(scsi_bios_ptable);
-- 
2.34.1

[PATCH 00/25] Remove AOP flags (and related cleanups)

[Matthew Wilcox (Oracle)]

This started out as just removing the AOP_FLAG_*, but turned into
removing pagecache_write_begin/end(), and then Smatch found an
unnecessary stack variable in ntfs3.

Matthew Wilcox (Oracle) (25):
  ext4: Use page_symlink() instead of __page_symlink()
  namei: Merge page_symlink() and __page_symlink()
  namei: Convert page_symlink() to use memalloc_nofs_save()
  f2fs: Convert f2fs_grab_cache_page() to use scoped memory APIs
  ext4: Allow GFP_FS allocations in
    ext4_da_convert_inline_data_to_extent()
  ext4: Use scoped memory API in mext_page_double_lock()
  ext4: Use scoped memory APIs in ext4_da_write_begin()
  ext4: Use scoped memory APIs in ext4_write_begin()
  fs: Remove AOP_FLAG_NOFS
  fs: Remove aop_flags parameter from netfs_write_begin()
  fs: Remove aop flags parameter from block_write_begin()
  fs: Remove aop flags parameter from cont_write_begin()
  fs: Remove aop flags parameter from grab_cache_page_write_begin()
  fs: Remove aop flags parameter from nobh_write_begin()
  fs: Remove flags parameter from aops->write_begin
  buffer: Call aops write_begin() and write_end() directly
  namei: Call aops write_begin() and write_end() directly
  ntfs3: Call ntfs_write_begin() and ntfs_write_end() directly
  ntfs3: Remove fsdata parameter from ntfs_extend_initialized_size()
  hfs: Call hfs_write_begin() and generic_write_end() directly
  hfsplus: Call hfsplus_write_begin() and generic_write_end() directly
  ext4: Call aops write_begin() and write_end() directly
  f2fs: Call aops write_begin() and write_end() directly
  i915: Call aops write_begin() and write_end() directly
  fs: Remove pagecache_write_begin() and pagecache_write_end()

 Documentation/filesystems/locking.rst       |  2 +-
 Documentation/filesystems/netfs_library.rst |  1 -
 Documentation/filesystems/porting.rst       |  2 +-
 Documentation/filesystems/vfs.rst           |  5 +--
 block/fops.c                                |  6 +--
 drivers/gpu/drm/i915/gem/i915_gem_shmem.c   | 23 ++++++------
 fs/9p/vfs_addr.c                            |  4 +-
 fs/adfs/inode.c                             |  4 +-
 fs/affs/file.c                              | 10 ++---
 fs/afs/internal.h                           |  2 +-
 fs/afs/write.c                              |  4 +-
 fs/bfs/file.c                               |  5 +--
 fs/buffer.c                                 | 27 +++++++-------
 fs/ceph/addr.c                              |  4 +-
 fs/cifs/file.c                              |  4 +-
 fs/ecryptfs/mmap.c                          |  4 +-
 fs/exfat/inode.c                            |  4 +-
 fs/ext2/inode.c                             | 11 ++----
 fs/ext4/ext4.h                              |  2 -
 fs/ext4/inline.c                            | 41 ++++++++++-----------
 fs/ext4/inode.c                             | 19 +++++-----
 fs/ext4/move_extent.c                       | 13 +++++--
 fs/ext4/namei.c                             |  7 +++-
 fs/ext4/verity.c                            |  9 +++--
 fs/f2fs/data.c                              |  5 +--
 fs/f2fs/f2fs.h                              |  9 ++++-
 fs/f2fs/super.c                             |  2 +-
 fs/f2fs/verity.c                            |  9 +++--
 fs/fat/inode.c                              |  4 +-
 fs/fuse/file.c                              |  7 ++--
 fs/hfs/extent.c                             |  6 +--
 fs/hfs/hfs_fs.h                             |  2 +
 fs/hfs/inode.c                              |  7 ++--
 fs/hfsplus/extents.c                        |  8 ++--
 fs/hfsplus/hfsplus_fs.h                     |  2 +
 fs/hfsplus/inode.c                          |  7 ++--
 fs/hostfs/hostfs_kern.c                     |  4 +-
 fs/hpfs/file.c                              |  4 +-
 fs/hugetlbfs/inode.c                        |  2 +-
 fs/jffs2/file.c                             |  6 +--
 fs/jfs/inode.c                              |  5 +--
 fs/libfs.c                                  |  4 +-
 fs/minix/inode.c                            |  5 +--
 fs/namei.c                                  | 28 ++++++--------
 fs/netfs/buffered_read.c                    | 10 ++---
 fs/nfs/file.c                               |  4 +-
 fs/nilfs2/inode.c                           |  5 +--
 fs/nilfs2/recovery.c                        |  2 +-
 fs/ntfs3/file.c                             |  7 +---
 fs/ntfs3/inode.c                            | 18 ++++-----
 fs/ntfs3/ntfs_fs.h                          |  5 +++
 fs/ocfs2/aops.c                             |  2 +-
 fs/omfs/file.c                              |  5 +--
 fs/orangefs/inode.c                         |  7 ++--
 fs/reiserfs/inode.c                         |  4 +-
 fs/sysv/itree.c                             |  4 +-
 fs/ubifs/file.c                             | 11 +++---
 fs/udf/file.c                               |  4 +-
 fs/udf/inode.c                              |  4 +-
 fs/ufs/inode.c                              |  5 +--
 include/linux/buffer_head.h                 |  6 +--
 include/linux/fs.h                          | 22 +----------
 include/linux/netfs.h                       |  2 +-
 include/linux/pagemap.h                     |  2 +-
 include/trace/events/ext4.h                 | 21 ++++-------
 include/trace/events/f2fs.h                 | 12 ++----
 mm/filemap.c                                | 24 +-----------
 mm/folio-compat.c                           |  4 +-
 mm/shmem.c                                  |  2 +-
 69 files changed, 233 insertions(+), 298 deletions(-)

-- 
2.34.1

[PATCH 01/25] ext4: Use page_symlink() instead of __page_symlink()

[Matthew Wilcox (Oracle)]

By using the memalloc_nofs_save() functionality, we can call
page_symlink(), safe in the knowledge that it won't recurse into the
filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/ext4/namei.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 767b4bfe39c3..1e7c5deed5e3 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -29,6 +29,7 @@
 #include <linux/pagemap.h>
 #include <linux/time.h>
 #include <linux/fcntl.h>
+#include <linux/sched/mm.h>
 #include <linux/stat.h>
 #include <linux/string.h>
 #include <linux/quotaops.h>
@@ -3308,6 +3309,8 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 	}
 
 	if ((disk_link.len > EXT4_N_BLOCKS * 4)) {
+		unsigned int flags;
+
 		if (!IS_ENCRYPTED(inode))
 			inode->i_op = &ext4_symlink_inode_operations;
 		inode_nohighmem(inode);
@@ -3329,7 +3332,9 @@ static int ext4_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 		handle = NULL;
 		if (err)
 			goto err_drop_inode;
-		err = __page_symlink(inode, disk_link.name, disk_link.len, 1);
+		flags = memalloc_nofs_save();
+		err = page_symlink(inode, disk_link.name, disk_link.len);
+		memalloc_nofs_restore(flags);
 		if (err)
 			goto err_drop_inode;
 		/*
-- 
2.34.1

[PATCH 02/25] namei: Merge page_symlink() and __page_symlink()

[Matthew Wilcox (Oracle)]

There are no callers of __page_symlink() left, so we can remove that
entry point.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Christian Brauner <brauner@kernel.org>
---
 Documentation/filesystems/porting.rst |  2 +-
 fs/namei.c                            | 13 ++-----------
 include/linux/fs.h                    |  2 --
 3 files changed, 3 insertions(+), 14 deletions(-)

diff --git a/Documentation/filesystems/porting.rst b/Documentation/filesystems/porting.rst
index 7c1583dbeb59..2e0e4f0e0c6f 100644
--- a/Documentation/filesystems/porting.rst
+++ b/Documentation/filesystems/porting.rst
@@ -624,7 +624,7 @@ any symlink that might use page_follow_link_light/page_put_link() must
 have inode_nohighmem(inode) called before anything might start playing with
 its pagecache.  No highmem pages should end up in the pagecache of such
 symlinks.  That includes any preseeding that might be done during symlink
-creation.  __page_symlink() will honour the mapping gfp flags, so once
+creation.  page_symlink() will honour the mapping gfp flags, so once
 you've done inode_nohighmem() it's safe to use, but if you allocate and
 insert the page manually, make sure to use the right gfp flags.
 
diff --git a/fs/namei.c b/fs/namei.c
index 509657fdf4f5..6153581073b1 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -5001,12 +5001,10 @@ int page_readlink(struct dentry *dentry, char __user *buffer, int buflen)
 }
 EXPORT_SYMBOL(page_readlink);
 
-/*
- * The nofs argument instructs pagecache_write_begin to pass AOP_FLAG_NOFS
- */
-int __page_symlink(struct inode *inode, const char *symname, int len, int nofs)
+int page_symlink(struct inode *inode, const char *symname, int len)
 {
 	struct address_space *mapping = inode->i_mapping;
+	bool nofs = !mapping_gfp_constraint(mapping, __GFP_FS);
 	struct page *page;
 	void *fsdata;
 	int err;
@@ -5034,13 +5032,6 @@ int __page_symlink(struct inode *inode, const char *symname, int len, int nofs)
 fail:
 	return err;
 }
-EXPORT_SYMBOL(__page_symlink);
-
-int page_symlink(struct inode *inode, const char *symname, int len)
-{
-	return __page_symlink(inode, symname, len,
-			!mapping_gfp_constraint(inode->i_mapping, __GFP_FS));
-}
 EXPORT_SYMBOL(page_symlink);
 
 const struct inode_operations page_symlink_inode_operations = {
diff --git a/include/linux/fs.h b/include/linux/fs.h
index bbde95387a23..e108aff23a28 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -3109,8 +3109,6 @@ extern int page_readlink(struct dentry *, char __user *, int);
 extern const char *page_get_link(struct dentry *, struct inode *,
 				 struct delayed_call *);
 extern void page_put_link(void *);
-extern int __page_symlink(struct inode *inode, const char *symname, int len,
-		int nofs);
 extern int page_symlink(struct inode *inode, const char *symname, int len);
 extern const struct inode_operations page_symlink_inode_operations;
 extern void kfree_link(void *);
-- 
2.34.1

[PATCH 03/25] namei: Convert page_symlink() to use memalloc_nofs_save()

[Matthew Wilcox (Oracle)]

Stop using AOP_FLAG_NOFS in favour of the scoped memory API.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/namei.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index 6153581073b1..0c84b4326dc9 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -22,6 +22,7 @@
 #include <linux/fs.h>
 #include <linux/namei.h>
 #include <linux/pagemap.h>
+#include <linux/sched/mm.h>
 #include <linux/fsnotify.h>
 #include <linux/personality.h>
 #include <linux/security.h>
@@ -5008,13 +5009,15 @@ int page_symlink(struct inode *inode, const char *symname, int len)
 	struct page *page;
 	void *fsdata;
 	int err;
-	unsigned int flags = 0;
-	if (nofs)
-		flags |= AOP_FLAG_NOFS;
+	unsigned int flags;
 
 retry:
+	if (nofs)
+		flags = memalloc_nofs_save();
 	err = pagecache_write_begin(NULL, mapping, 0, len-1,
-				flags, &page, &fsdata);
+				0, &page, &fsdata);
+	if (nofs)
+		memalloc_nofs_restore(flags);
 	if (err)
 		goto fail;
 
-- 
2.34.1

[PATCH 04/25] f2fs: Convert f2fs_grab_cache_page() to use scoped memory APIs

[Matthew Wilcox (Oracle)]

Prevent GFP_FS allocations by using memalloc_nofs_save() instead
of AOP_FLAG_NOFS.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/f2fs/f2fs.h | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 8c570de21ed5..74929ade4b5e 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -18,6 +18,7 @@
 #include <linux/kobject.h>
 #include <linux/sched.h>
 #include <linux/cred.h>
+#include <linux/sched/mm.h>
 #include <linux/vmalloc.h>
 #include <linux/bio.h>
 #include <linux/blkdev.h>
@@ -2654,6 +2655,7 @@ static inline struct page *f2fs_grab_cache_page(struct address_space *mapping,
 						pgoff_t index, bool for_write)
 {
 	struct page *page;
+	unsigned int flags;
 
 	if (IS_ENABLED(CONFIG_F2FS_FAULT_INJECTION)) {
 		if (!for_write)
@@ -2673,7 +2675,12 @@ static inline struct page *f2fs_grab_cache_page(struct address_space *mapping,
 
 	if (!for_write)
 		return grab_cache_page(mapping, index);
-	return grab_cache_page_write_begin(mapping, index, AOP_FLAG_NOFS);
+
+	flags = memalloc_nofs_save();
+	page = grab_cache_page_write_begin(mapping, index, 0);
+	memalloc_nofs_restore(flags);
+
+	return page;
 }
 
 static inline struct page *f2fs_pagecache_get_page(
-- 
2.34.1

[PATCH 05/25] ext4: Allow GFP_FS allocations in ext4_da_convert_inline_data_to_extent()

[Matthew Wilcox (Oracle)]

Since commit 8bc1379b82b8, the transaction is stopped before calling
ext4_da_convert_inline_data_to_extent(), which means we can do GFP_FS
allocations and recurse into the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/inline.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index 9c076262770d..93694ceb5a34 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -848,13 +848,12 @@ ext4_journalled_write_inline_data(struct inode *inode,
  */
 static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping,
 						 struct inode *inode,
-						 unsigned flags,
 						 void **fsdata)
 {
 	int ret = 0, inline_size;
 	struct page *page;
 
-	page = grab_cache_page_write_begin(mapping, 0, flags);
+	page = grab_cache_page_write_begin(mapping, 0, 0);
 	if (!page)
 		return -ENOMEM;
 
@@ -942,7 +941,6 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 		ext4_journal_stop(handle);
 		ret = ext4_da_convert_inline_data_to_extent(mapping,
 							    inode,
-							    flags,
 							    fsdata);
 		if (ret == -ENOSPC &&
 		    ext4_should_retry_alloc(inode->i_sb, &retries))
-- 
2.34.1

[PATCH 06/25] ext4: Use scoped memory API in mext_page_double_lock()

[Matthew Wilcox (Oracle)]

Replace use of AOP_FLAG_NOFS with calls to memalloc_nofs_save()
and memalloc_nofs_restore().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/move_extent.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 95aa212f0863..56f21272fb00 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -8,6 +8,7 @@
 #include <linux/fs.h>
 #include <linux/quotaops.h>
 #include <linux/slab.h>
+#include <linux/sched/mm.h>
 #include "ext4_jbd2.h"
 #include "ext4.h"
 #include "ext4_extents.h"
@@ -127,7 +128,7 @@ mext_page_double_lock(struct inode *inode1, struct inode *inode2,
 		      pgoff_t index1, pgoff_t index2, struct page *page[2])
 {
 	struct address_space *mapping[2];
-	unsigned fl = AOP_FLAG_NOFS;
+	unsigned int flags;
 
 	BUG_ON(!inode1 || !inode2);
 	if (inode1 < inode2) {
@@ -139,11 +140,15 @@ mext_page_double_lock(struct inode *inode1, struct inode *inode2,
 		mapping[1] = inode1->i_mapping;
 	}
 
-	page[0] = grab_cache_page_write_begin(mapping[0], index1, fl);
-	if (!page[0])
+	flags = memalloc_nofs_save();
+	page[0] = grab_cache_page_write_begin(mapping[0], index1, 0);
+	if (!page[0]) {
+		memalloc_nofs_restore(flags);
 		return -ENOMEM;
+	}
 
-	page[1] = grab_cache_page_write_begin(mapping[1], index2, fl);
+	page[1] = grab_cache_page_write_begin(mapping[1], index2, 0);
+	memalloc_nofs_restore(flags);
 	if (!page[1]) {
 		unlock_page(page[0]);
 		put_page(page[0]);
-- 
2.34.1

[PATCH 07/25] ext4: Use scoped memory APIs in ext4_da_write_begin()

[Matthew Wilcox (Oracle)]

Instead of setting AOP_FLAG_NOFS, use memalloc_nofs_save() and
memalloc_nofs_restore() to prevent GFP_FS allocations recursing
into the filesystem with a journal already started.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/ext4.h   |  1 -
 fs/ext4/inline.c | 16 ++++++++--------
 fs/ext4/inode.c  |  3 +--
 3 files changed, 9 insertions(+), 11 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index a743b1e3b89e..90677e30e52d 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -3604,7 +3604,6 @@ ext4_journalled_write_inline_data(struct inode *inode,
 extern int ext4_da_write_inline_data_begin(struct address_space *mapping,
 					   struct inode *inode,
 					   loff_t pos, unsigned len,
-					   unsigned flags,
 					   struct page **pagep,
 					   void **fsdata);
 extern int ext4_try_add_inline_entry(handle_t *handle,
diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index 93694ceb5a34..d965ba08f68f 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -906,7 +906,6 @@ static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping,
 int ext4_da_write_inline_data_begin(struct address_space *mapping,
 				    struct inode *inode,
 				    loff_t pos, unsigned len,
-				    unsigned flags,
 				    struct page **pagep,
 				    void **fsdata)
 {
@@ -915,6 +914,7 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 	struct page *page;
 	struct ext4_iloc iloc;
 	int retries = 0;
+	unsigned int flags;
 
 	ret = ext4_get_inode_loc(inode, &iloc);
 	if (ret)
@@ -931,12 +931,6 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 	if (ret && ret != -ENOSPC)
 		goto out_journal;
 
-	/*
-	 * We cannot recurse into the filesystem as the transaction
-	 * is already started.
-	 */
-	flags |= AOP_FLAG_NOFS;
-
 	if (ret == -ENOSPC) {
 		ext4_journal_stop(handle);
 		ret = ext4_da_convert_inline_data_to_extent(mapping,
@@ -948,7 +942,13 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 		goto out;
 	}
 
-	page = grab_cache_page_write_begin(mapping, 0, flags);
+	/*
+	 * We cannot recurse into the filesystem as the transaction
+	 * is already started.
+	 */
+	flags = memalloc_nofs_save();
+	page = grab_cache_page_write_begin(mapping, 0, 0);
+	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
 		goto out_journal;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 646ece9b3455..21ebcb3c59ba 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -2954,8 +2954,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	trace_ext4_da_write_begin(inode, pos, len, flags);
 
 	if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) {
-		ret = ext4_da_write_inline_data_begin(mapping, inode,
-						      pos, len, flags,
+		ret = ext4_da_write_inline_data_begin(mapping, inode, pos, len,
 						      pagep, fsdata);
 		if (ret < 0)
 			return ret;
-- 
2.34.1

[PATCH 08/25] ext4: Use scoped memory APIs in ext4_write_begin()

[Matthew Wilcox (Oracle)]

Instead of setting AOP_FLAG_NOFS, use memalloc_nofs_save() and
memalloc_nofs_restore() to prevent GFP_FS allocations recursing
into the filesystem with a journal already started.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Acked-by: Theodore Ts'o <tytso@mit.edu>
---
 fs/ext4/ext4.h   |  1 -
 fs/ext4/inline.c | 21 ++++++++++-----------
 fs/ext4/inode.c  |  2 +-
 3 files changed, 11 insertions(+), 13 deletions(-)

diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 90677e30e52d..0c3308bac6c1 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -3591,7 +3591,6 @@ extern int ext4_readpage_inline(struct inode *inode, struct page *page);
 extern int ext4_try_to_write_inline_data(struct address_space *mapping,
 					 struct inode *inode,
 					 loff_t pos, unsigned len,
-					 unsigned flags,
 					 struct page **pagep);
 extern int ext4_write_inline_data_end(struct inode *inode,
 				      loff_t pos, unsigned len,
diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index d965ba08f68f..b2ef5ba568bc 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -527,13 +527,13 @@ int ext4_readpage_inline(struct inode *inode, struct page *page)
 }
 
 static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
-					      struct inode *inode,
-					      unsigned flags)
+					      struct inode *inode)
 {
 	int ret, needed_blocks, no_expand;
 	handle_t *handle = NULL;
 	int retries = 0, sem_held = 0;
 	struct page *page = NULL;
+	unsigned int flags;
 	unsigned from, to;
 	struct ext4_iloc iloc;
 
@@ -562,9 +562,9 @@ static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
 
 	/* We cannot recurse into the filesystem as the transaction is already
 	 * started */
-	flags |= AOP_FLAG_NOFS;
-
-	page = grab_cache_page_write_begin(mapping, 0, flags);
+	flags = memalloc_nofs_save();
+	page = grab_cache_page_write_begin(mapping, 0, 0);
+	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
 		goto out;
@@ -649,11 +649,11 @@ static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
 int ext4_try_to_write_inline_data(struct address_space *mapping,
 				  struct inode *inode,
 				  loff_t pos, unsigned len,
-				  unsigned flags,
 				  struct page **pagep)
 {
 	int ret;
 	handle_t *handle;
+	unsigned int flags;
 	struct page *page;
 	struct ext4_iloc iloc;
 
@@ -691,9 +691,9 @@ int ext4_try_to_write_inline_data(struct address_space *mapping,
 	if (ret)
 		goto out;
 
-	flags |= AOP_FLAG_NOFS;
-
-	page = grab_cache_page_write_begin(mapping, 0, flags);
+	flags = memalloc_nofs_save();
+	page = grab_cache_page_write_begin(mapping, 0, 0);
+	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
 		goto out;
@@ -727,8 +727,7 @@ int ext4_try_to_write_inline_data(struct address_space *mapping,
 	brelse(iloc.bh);
 	return ret;
 convert:
-	return ext4_convert_inline_data_to_extent(mapping,
-						  inode, flags);
+	return ext4_convert_inline_data_to_extent(mapping, inode);
 }
 
 int ext4_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len,
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 21ebcb3c59ba..01a55647c959 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1156,7 +1156,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping,
 
 	if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) {
 		ret = ext4_try_to_write_inline_data(mapping, inode, pos, len,
-						    flags, pagep);
+						    pagep);
 		if (ret < 0)
 			return ret;
 		if (ret == 1)
-- 
2.34.1

[PATCH 09/25] fs: Remove AOP_FLAG_NOFS

[Matthew Wilcox (Oracle)]

With all users of this flag gone, we can stop testing whether it's set.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/netfs/buffered_read.c | 6 +-----
 include/linux/fs.h       | 4 ----
 mm/folio-compat.c        | 2 --
 3 files changed, 1 insertion(+), 11 deletions(-)

diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index 281a88a5b8dc..65c17c5a5567 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -302,7 +302,6 @@ static bool netfs_skip_folio_read(struct folio *folio, loff_t pos, size_t len,
  * @mapping: The mapping to read from
  * @pos: File position at which the write will begin
  * @len: The length of the write (may extend beyond the end of the folio chosen)
- * @aop_flags: AOP_* flags
  * @_folio: Where to put the resultant folio
  * @_fsdata: Place for the netfs to store a cookie
  *
@@ -335,16 +334,13 @@ int netfs_write_begin(struct file *file, struct address_space *mapping,
 	struct netfs_io_request *rreq;
 	struct netfs_i_context *ctx = netfs_i_context(file_inode(file ));
 	struct folio *folio;
-	unsigned int fgp_flags;
+	unsigned int fgp_flags = FGP_LOCK | FGP_WRITE | FGP_CREAT | FGP_STABLE;
 	pgoff_t index = pos >> PAGE_SHIFT;
 	int ret;
 
 	DEFINE_READAHEAD(ractl, file, NULL, mapping, index);
 
 retry:
-	fgp_flags = FGP_LOCK | FGP_WRITE | FGP_CREAT | FGP_STABLE;
-	if (aop_flags & AOP_FLAG_NOFS)
-		fgp_flags |= FGP_NOFS;
 	folio = __filemap_get_folio(mapping, index, fgp_flags,
 				    mapping_gfp_mask(mapping));
 	if (!folio)
diff --git a/include/linux/fs.h b/include/linux/fs.h
index e108aff23a28..f81bc5cbcbb6 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -275,10 +275,6 @@ enum positive_aop_returns {
 	AOP_TRUNCATED_PAGE	= 0x80001,
 };
 
-#define AOP_FLAG_NOFS			0x0002 /* used by filesystem to direct
-						* helper code (eg buffer layer)
-						* to clear GFP_FS from alloc */
-
 /*
  * oh the beauties of C type declarations.
  */
diff --git a/mm/folio-compat.c b/mm/folio-compat.c
index 46fa179e32fb..3e42ddb81918 100644
--- a/mm/folio-compat.c
+++ b/mm/folio-compat.c
@@ -135,8 +135,6 @@ struct page *grab_cache_page_write_begin(struct address_space *mapping,
 {
 	unsigned fgp_flags = FGP_LOCK | FGP_WRITE | FGP_CREAT | FGP_STABLE;
 
-	if (flags & AOP_FLAG_NOFS)
-		fgp_flags |= FGP_NOFS;
 	return pagecache_get_page(mapping, index, fgp_flags,
 			mapping_gfp_mask(mapping));
 }
-- 
2.34.1

[PATCH 10/25] fs: Remove aop_flags parameter from netfs_write_begin()

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 Documentation/filesystems/netfs_library.rst | 1 -
 fs/9p/vfs_addr.c                            | 2 +-
 fs/afs/write.c                              | 2 +-
 fs/ceph/addr.c                              | 2 +-
 fs/netfs/buffered_read.c                    | 4 ++--
 include/linux/netfs.h                       | 2 +-
 6 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/Documentation/filesystems/netfs_library.rst b/Documentation/filesystems/netfs_library.rst
index 69f00179fdfe..d51c2a5ccf57 100644
--- a/Documentation/filesystems/netfs_library.rst
+++ b/Documentation/filesystems/netfs_library.rst
@@ -142,7 +142,6 @@ Three read helpers are provided::
 			      struct address_space *mapping,
 			      loff_t pos,
 			      unsigned int len,
-			      unsigned int flags,
 			      struct folio **_folio,
 			      void **_fsdata);
 
diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c
index 501128188343..d311e68e21fd 100644
--- a/fs/9p/vfs_addr.c
+++ b/fs/9p/vfs_addr.c
@@ -275,7 +275,7 @@ static int v9fs_write_begin(struct file *filp, struct address_space *mapping,
 	 * file.  We need to do this before we get a lock on the page in case
 	 * there's more than one writer competing for the same cache block.
 	 */
-	retval = netfs_write_begin(filp, mapping, pos, len, flags, &folio, fsdata);
+	retval = netfs_write_begin(filp, mapping, pos, len, &folio, fsdata);
 	if (retval < 0)
 		return retval;
 
diff --git a/fs/afs/write.c b/fs/afs/write.c
index 4763132ca57e..af496c98d394 100644
--- a/fs/afs/write.c
+++ b/fs/afs/write.c
@@ -60,7 +60,7 @@ int afs_write_begin(struct file *file, struct address_space *mapping,
 	 * file.  We need to do this before we get a lock on the page in case
 	 * there's more than one writer competing for the same cache block.
 	 */
-	ret = netfs_write_begin(file, mapping, pos, len, flags, &folio, fsdata);
+	ret = netfs_write_begin(file, mapping, pos, len, &folio, fsdata);
 	if (ret < 0)
 		return ret;
 
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index aa25bffd4823..415f0886bc25 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1318,7 +1318,7 @@ static int ceph_write_begin(struct file *file, struct address_space *mapping,
 	struct folio *folio = NULL;
 	int r;
 
-	r = netfs_write_begin(file, inode->i_mapping, pos, len, 0, &folio, NULL);
+	r = netfs_write_begin(file, inode->i_mapping, pos, len, &folio, NULL);
 	if (r == 0)
 		folio_wait_fscache(folio);
 	if (r < 0) {
diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index 65c17c5a5567..1d44509455a5 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -328,8 +328,8 @@ static bool netfs_skip_folio_read(struct folio *folio, loff_t pos, size_t len,
  * This is usable whether or not caching is enabled.
  */
 int netfs_write_begin(struct file *file, struct address_space *mapping,
-		      loff_t pos, unsigned int len, unsigned int aop_flags,
-		      struct folio **_folio, void **_fsdata)
+		      loff_t pos, unsigned int len, struct folio **_folio,
+		      void **_fsdata)
 {
 	struct netfs_io_request *rreq;
 	struct netfs_i_context *ctx = netfs_i_context(file_inode(file ));
diff --git a/include/linux/netfs.h b/include/linux/netfs.h
index c7bf1eaf51d5..1c29f317d907 100644
--- a/include/linux/netfs.h
+++ b/include/linux/netfs.h
@@ -276,7 +276,7 @@ struct readahead_control;
 extern void netfs_readahead(struct readahead_control *);
 extern int netfs_readpage(struct file *, struct page *);
 extern int netfs_write_begin(struct file *, struct address_space *,
-			     loff_t, unsigned int, unsigned int, struct folio **,
+			     loff_t, unsigned int, struct folio **,
 			     void **);
 
 extern void netfs_subreq_terminated(struct netfs_io_subrequest *, ssize_t, bool);
-- 
2.34.1

[PATCH 11/25] fs: Remove aop flags parameter from block_write_begin()

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 block/fops.c                | 3 +--
 fs/bfs/file.c               | 3 +--
 fs/buffer.c                 | 6 +++---
 fs/ext2/inode.c             | 3 +--
 fs/minix/inode.c            | 3 +--
 fs/nilfs2/inode.c           | 3 +--
 fs/nilfs2/recovery.c        | 2 +-
 fs/ntfs3/inode.c            | 4 ++--
 fs/omfs/file.c              | 3 +--
 fs/sysv/itree.c             | 2 +-
 fs/udf/inode.c              | 2 +-
 fs/ufs/inode.c              | 3 +--
 include/linux/buffer_head.h | 2 +-
 13 files changed, 16 insertions(+), 23 deletions(-)

diff --git a/block/fops.c b/block/fops.c
index 9f2ecec406b0..b432756570c6 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -401,8 +401,7 @@ static int blkdev_write_begin(struct file *file, struct address_space *mapping,
 		loff_t pos, unsigned len, unsigned flags, struct page **pagep,
 		void **fsdata)
 {
-	return block_write_begin(mapping, pos, len, flags, pagep,
-				 blkdev_get_block);
+	return block_write_begin(mapping, pos, len, pagep, blkdev_get_block);
 }
 
 static int blkdev_write_end(struct file *file, struct address_space *mapping,
diff --git a/fs/bfs/file.c b/fs/bfs/file.c
index 03139344568f..9408f45225cb 100644
--- a/fs/bfs/file.c
+++ b/fs/bfs/file.c
@@ -174,8 +174,7 @@ static int bfs_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep,
-				bfs_get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, bfs_get_block);
 	if (unlikely(ret))
 		bfs_write_failed(mapping, pos + len);
 
diff --git a/fs/buffer.c b/fs/buffer.c
index 2b5561ae5d0b..4ec6eb03c0eb 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2104,13 +2104,13 @@ static int __block_commit_write(struct inode *inode, struct page *page,
  * The filesystem needs to handle block truncation upon failure.
  */
 int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
-		unsigned flags, struct page **pagep, get_block_t *get_block)
+		struct page **pagep, get_block_t *get_block)
 {
 	pgoff_t index = pos >> PAGE_SHIFT;
 	struct page *page;
 	int status;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index, 0);
 	if (!page)
 		return -ENOMEM;
 
@@ -2460,7 +2460,7 @@ int cont_write_begin(struct file *file, struct address_space *mapping,
 		(*bytes)++;
 	}
 
-	return block_write_begin(mapping, pos, len, flags, pagep, get_block);
+	return block_write_begin(mapping, pos, len, pagep, get_block);
 }
 EXPORT_SYMBOL(cont_write_begin);
 
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 52377a0ee735..97192932ea56 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -892,8 +892,7 @@ ext2_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep,
-				ext2_get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, ext2_get_block);
 	if (ret < 0)
 		ext2_write_failed(mapping, pos + len);
 	return ret;
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index f1a6610e4ee6..5e8d7ba661cf 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -428,8 +428,7 @@ static int minix_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep,
-				minix_get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, minix_get_block);
 	if (unlikely(ret))
 		minix_write_failed(mapping, pos + len);
 
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 6045cea21f52..be09a0d10f04 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -258,8 +258,7 @@ static int nilfs_write_begin(struct file *file, struct address_space *mapping,
 	if (unlikely(err))
 		return err;
 
-	err = block_write_begin(mapping, pos, len, flags, pagep,
-				nilfs_get_block);
+	err = block_write_begin(mapping, pos, len, pagep, nilfs_get_block);
 	if (unlikely(err)) {
 		nilfs_write_failed(mapping, pos + len);
 		nilfs_transaction_abort(inode->i_sb);
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index 9e2ed76c0f25..0955b657938f 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -511,7 +511,7 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs,
 
 		pos = rb->blkoff << inode->i_blkbits;
 		err = block_write_begin(inode->i_mapping, pos, blocksize,
-					0, &page, nilfs_get_block);
+					&page, nilfs_get_block);
 		if (unlikely(err)) {
 			loff_t isize = inode->i_size;
 
diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c
index 9eab11e3b034..3914138fd8ba 100644
--- a/fs/ntfs3/inode.c
+++ b/fs/ntfs3/inode.c
@@ -894,7 +894,7 @@ static int ntfs_write_begin(struct file *file, struct address_space *mapping,
 			goto out;
 	}
 
-	err = block_write_begin(mapping, pos, len, flags, pagep,
+	err = block_write_begin(mapping, pos, len, pagep,
 				ntfs_get_block_write_begin);
 
 out:
@@ -975,7 +975,7 @@ int reset_log_file(struct inode *inode)
 
 		len = pos + PAGE_SIZE > log_size ? (log_size - pos) : PAGE_SIZE;
 
-		err = block_write_begin(mapping, pos, len, 0, &page,
+		err = block_write_begin(mapping, pos, len, &page,
 					ntfs_get_block_write_begin);
 		if (err)
 			goto out;
diff --git a/fs/omfs/file.c b/fs/omfs/file.c
index 3f297b541713..349b96d89c44 100644
--- a/fs/omfs/file.c
+++ b/fs/omfs/file.c
@@ -321,8 +321,7 @@ static int omfs_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep,
-				omfs_get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, omfs_get_block);
 	if (unlikely(ret))
 		omfs_write_failed(mapping, pos + len);
 
diff --git a/fs/sysv/itree.c b/fs/sysv/itree.c
index 409ab5e17803..96b7fd4facf3 100644
--- a/fs/sysv/itree.c
+++ b/fs/sysv/itree.c
@@ -482,7 +482,7 @@ static int sysv_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep, get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, get_block);
 	if (unlikely(ret))
 		sysv_write_failed(mapping, pos + len);
 
diff --git a/fs/udf/inode.c b/fs/udf/inode.c
index ca4fa710e562..88a95886ce8a 100644
--- a/fs/udf/inode.c
+++ b/fs/udf/inode.c
@@ -209,7 +209,7 @@ static int udf_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep, udf_get_block);
+	ret = block_write_begin(mapping, pos, len, pagep, udf_get_block);
 	if (unlikely(ret))
 		udf_write_failed(mapping, pos + len);
 	return ret;
diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c
index d0dda01620f0..bd0e0c66f93d 100644
--- a/fs/ufs/inode.c
+++ b/fs/ufs/inode.c
@@ -500,8 +500,7 @@ static int ufs_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = block_write_begin(mapping, pos, len, flags, pagep,
-				ufs_getfrag_block);
+	ret = block_write_begin(mapping, pos, len, pagep, ufs_getfrag_block);
 	if (unlikely(ret))
 		ufs_write_failed(mapping, pos + len);
 
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index bcb4fe9b8575..63e49dfa7738 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -226,7 +226,7 @@ int __block_write_full_page(struct inode *inode, struct page *page,
 int block_read_full_page(struct page*, get_block_t*);
 bool block_is_partially_uptodate(struct folio *, size_t from, size_t count);
 int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
-		unsigned flags, struct page **pagep, get_block_t *get_block);
+		struct page **pagep, get_block_t *get_block);
 int __block_write_begin(struct page *page, loff_t pos, unsigned len,
 		get_block_t *get_block);
 int block_write_end(struct file *, struct address_space *,
-- 
2.34.1

[PATCH 12/25] fs: Remove aop flags parameter from cont_write_begin()

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/adfs/inode.c             | 2 +-
 fs/affs/file.c              | 2 +-
 fs/buffer.c                 | 2 +-
 fs/exfat/inode.c            | 2 +-
 fs/fat/inode.c              | 2 +-
 fs/hfs/inode.c              | 2 +-
 fs/hfsplus/inode.c          | 2 +-
 fs/hpfs/file.c              | 2 +-
 include/linux/buffer_head.h | 2 +-
 9 files changed, 9 insertions(+), 9 deletions(-)

diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c
index 561bc748c04a..b6912496bb19 100644
--- a/fs/adfs/inode.c
+++ b/fs/adfs/inode.c
@@ -58,7 +58,7 @@ static int adfs_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 				adfs_get_block,
 				&ADFS_I(mapping->host)->mmu_private);
 	if (unlikely(ret))
diff --git a/fs/affs/file.c b/fs/affs/file.c
index b3f81d84ff4c..704911d6aeba 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -420,7 +420,7 @@ static int affs_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 				affs_get_block,
 				&AFFS_I(mapping->host)->mmu_private);
 	if (unlikely(ret))
diff --git a/fs/buffer.c b/fs/buffer.c
index 4ec6eb03c0eb..fb97646d1977 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2441,7 +2441,7 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping,
  * We may have to extend the file.
  */
 int cont_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata,
 			get_block_t *get_block, loff_t *bytes)
 {
diff --git a/fs/exfat/inode.c b/fs/exfat/inode.c
index fc0ea1684880..8ed3c4b700cd 100644
--- a/fs/exfat/inode.c
+++ b/fs/exfat/inode.c
@@ -395,7 +395,7 @@ static int exfat_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 			       exfat_get_block,
 			       &EXFAT_I(mapping->host)->i_size_ondisk);
 
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index bf6051bdf1d1..9b34ccef2501 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -232,7 +232,7 @@ static int fat_write_begin(struct file *file, struct address_space *mapping,
 	int err;
 
 	*pagep = NULL;
-	err = cont_write_begin(file, mapping, pos, len, flags,
+	err = cont_write_begin(file, mapping, pos, len,
 				pagep, fsdata, fat_get_block,
 				&MSDOS_I(mapping->host)->mmu_private);
 	if (err < 0)
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 55f45e9b4930..396735dd3407 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -56,7 +56,7 @@ static int hfs_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 				hfs_get_block,
 				&HFS_I(mapping->host)->phys_size);
 	if (unlikely(ret))
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 446a816aa8e1..435b6202532a 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -50,7 +50,7 @@ static int hfsplus_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 				hfsplus_get_block,
 				&HFSPLUS_I(mapping->host)->phys_size);
 	if (unlikely(ret))
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index 99493a23c5d0..8740b4ea0b52 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -200,7 +200,7 @@ static int hpfs_write_begin(struct file *file, struct address_space *mapping,
 	int ret;
 
 	*pagep = NULL;
-	ret = cont_write_begin(file, mapping, pos, len, flags, pagep, fsdata,
+	ret = cont_write_begin(file, mapping, pos, len, pagep, fsdata,
 				hpfs_get_block,
 				&hpfs_i(mapping->host)->mmu_private);
 	if (unlikely(ret))
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 63e49dfa7738..127b60fad77e 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -238,7 +238,7 @@ int generic_write_end(struct file *, struct address_space *,
 void page_zero_new_buffers(struct page *page, unsigned from, unsigned to);
 void clean_page_buffers(struct page *page);
 int cont_write_begin(struct file *, struct address_space *, loff_t,
-			unsigned, unsigned, struct page **, void **,
+			unsigned, struct page **, void **,
 			get_block_t *, loff_t *);
 int generic_cont_expand_simple(struct inode *inode, loff_t size);
 int block_commit_write(struct page *page, unsigned from, unsigned to);
-- 
2.34.1

[PATCH 13/25] fs: Remove aop flags parameter from grab_cache_page_write_begin()

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/affs/file.c          | 2 +-
 fs/buffer.c             | 4 ++--
 fs/cifs/file.c          | 2 +-
 fs/ecryptfs/mmap.c      | 2 +-
 fs/ext4/inline.c        | 8 ++++----
 fs/ext4/inode.c         | 4 ++--
 fs/ext4/move_extent.c   | 4 ++--
 fs/f2fs/f2fs.h          | 2 +-
 fs/fuse/file.c          | 4 ++--
 fs/hostfs/hostfs_kern.c | 2 +-
 fs/jffs2/file.c         | 2 +-
 fs/libfs.c              | 2 +-
 fs/nfs/file.c           | 2 +-
 fs/ntfs3/inode.c        | 2 +-
 fs/orangefs/inode.c     | 2 +-
 fs/reiserfs/inode.c     | 2 +-
 fs/ubifs/file.c         | 4 ++--
 fs/udf/file.c           | 2 +-
 include/linux/pagemap.h | 2 +-
 mm/folio-compat.c       | 2 +-
 20 files changed, 28 insertions(+), 28 deletions(-)

diff --git a/fs/affs/file.c b/fs/affs/file.c
index 704911d6aeba..06645d05c717 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -670,7 +670,7 @@ static int affs_write_begin_ofs(struct file *file, struct address_space *mapping
 	}
 
 	index = pos >> PAGE_SHIFT;
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/fs/buffer.c b/fs/buffer.c
index fb97646d1977..01630218c75f 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2110,7 +2110,7 @@ int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
 	struct page *page;
 	int status;
 
-	page = grab_cache_page_write_begin(mapping, index, 0);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 
@@ -2591,7 +2591,7 @@ int nobh_write_begin(struct address_space *mapping,
 	from = pos & (PAGE_SIZE - 1);
 	to = from + len;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index d511a78383c3..91aeae7fced8 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -4695,7 +4695,7 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping,
 	cifs_dbg(FYI, "write_begin from %lld len %d\n", (long long)pos, len);
 
 start:
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page) {
 		rc = -ENOMEM;
 		goto out;
diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c
index 9ad61b582f07..84e399a921ad 100644
--- a/fs/ecryptfs/mmap.c
+++ b/fs/ecryptfs/mmap.c
@@ -272,7 +272,7 @@ static int ecryptfs_write_begin(struct file *file,
 	loff_t prev_page_end_size;
 	int rc = 0;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
index b2ef5ba568bc..6d253edebf9f 100644
--- a/fs/ext4/inline.c
+++ b/fs/ext4/inline.c
@@ -563,7 +563,7 @@ static int ext4_convert_inline_data_to_extent(struct address_space *mapping,
 	/* We cannot recurse into the filesystem as the transaction is already
 	 * started */
 	flags = memalloc_nofs_save();
-	page = grab_cache_page_write_begin(mapping, 0, 0);
+	page = grab_cache_page_write_begin(mapping, 0);
 	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
@@ -692,7 +692,7 @@ int ext4_try_to_write_inline_data(struct address_space *mapping,
 		goto out;
 
 	flags = memalloc_nofs_save();
-	page = grab_cache_page_write_begin(mapping, 0, 0);
+	page = grab_cache_page_write_begin(mapping, 0);
 	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
@@ -852,7 +852,7 @@ static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping,
 	int ret = 0, inline_size;
 	struct page *page;
 
-	page = grab_cache_page_write_begin(mapping, 0, 0);
+	page = grab_cache_page_write_begin(mapping, 0);
 	if (!page)
 		return -ENOMEM;
 
@@ -946,7 +946,7 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
 	 * is already started.
 	 */
 	flags = memalloc_nofs_save();
-	page = grab_cache_page_write_begin(mapping, 0, 0);
+	page = grab_cache_page_write_begin(mapping, 0);
 	memalloc_nofs_restore(flags);
 	if (!page) {
 		ret = -ENOMEM;
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 01a55647c959..512d8143c765 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1171,7 +1171,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping,
 	 * the page (if needed) without using GFP_NOFS.
 	 */
 retry_grab:
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	unlock_page(page);
@@ -2963,7 +2963,7 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	}
 
 retry:
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 56f21272fb00..4172a7d22471 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -141,13 +141,13 @@ mext_page_double_lock(struct inode *inode1, struct inode *inode2,
 	}
 
 	flags = memalloc_nofs_save();
-	page[0] = grab_cache_page_write_begin(mapping[0], index1, 0);
+	page[0] = grab_cache_page_write_begin(mapping[0], index1);
 	if (!page[0]) {
 		memalloc_nofs_restore(flags);
 		return -ENOMEM;
 	}
 
-	page[1] = grab_cache_page_write_begin(mapping[1], index2, 0);
+	page[1] = grab_cache_page_write_begin(mapping[1], index2);
 	memalloc_nofs_restore(flags);
 	if (!page[1]) {
 		unlock_page(page[0]);
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 74929ade4b5e..18df53ef3d7e 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -2677,7 +2677,7 @@ static inline struct page *f2fs_grab_cache_page(struct address_space *mapping,
 		return grab_cache_page(mapping, index);
 
 	flags = memalloc_nofs_save();
-	page = grab_cache_page_write_begin(mapping, index, 0);
+	page = grab_cache_page_write_begin(mapping, index);
 	memalloc_nofs_restore(flags);
 
 	return page;
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index f18d14d5fea1..e35e394264ad 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -1174,7 +1174,7 @@ static ssize_t fuse_fill_write_pages(struct fuse_io_args *ia,
 			break;
 
 		err = -ENOMEM;
-		page = grab_cache_page_write_begin(mapping, index, 0);
+		page = grab_cache_page_write_begin(mapping, index);
 		if (!page)
 			break;
 
@@ -2284,7 +2284,7 @@ static int fuse_write_begin(struct file *file, struct address_space *mapping,
 
 	WARN_ON(!fc->writeback_cache);
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		goto error;
 
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index 14f9ac973a2e..2bfd316e1bf1 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -468,7 +468,7 @@ static int hostfs_write_begin(struct file *file, struct address_space *mapping,
 {
 	pgoff_t index = pos >> PAGE_SHIFT;
 
-	*pagep = grab_cache_page_write_begin(mapping, index, flags);
+	*pagep = grab_cache_page_write_begin(mapping, index);
 	if (!*pagep)
 		return -ENOMEM;
 	return 0;
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index bd7d58d27bfc..142d3ba9f0a8 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -213,7 +213,7 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 	 * page in read_cache_page(), which causes a deadlock.
 	 */
 	mutex_lock(&c->alloc_sem);
-	pg = grab_cache_page_write_begin(mapping, index, flags);
+	pg = grab_cache_page_write_begin(mapping, index);
 	if (!pg) {
 		ret = -ENOMEM;
 		goto release_sem;
diff --git a/fs/libfs.c b/fs/libfs.c
index e64bdedef168..d4395e1c6696 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -557,7 +557,7 @@ int simple_write_begin(struct file *file, struct address_space *mapping,
 
 	index = pos >> PAGE_SHIFT;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 150b7fa8f0a7..d66088dd33e7 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -325,7 +325,7 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,
 		file, mapping->host->i_ino, len, (long long) pos);
 
 start:
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c
index 3914138fd8ba..16466c8648f3 100644
--- a/fs/ntfs3/inode.c
+++ b/fs/ntfs3/inode.c
@@ -872,7 +872,7 @@ static int ntfs_write_begin(struct file *file, struct address_space *mapping,
 	*pagep = NULL;
 	if (is_resident(ni)) {
 		struct page *page = grab_cache_page_write_begin(
-			mapping, pos >> PAGE_SHIFT, flags);
+			mapping, pos >> PAGE_SHIFT);
 
 		if (!page) {
 			err = -ENOMEM;
diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 79c1025d18ea..809690db8be2 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -338,7 +338,7 @@ static int orangefs_write_begin(struct file *file,
 
 	index = pos >> PAGE_SHIFT;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 36c59b25486c..aa31cf1dbba6 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -2764,7 +2764,7 @@ static int reiserfs_write_begin(struct file *file,
 
  	inode = mapping->host;
 	index = pos >> PAGE_SHIFT;
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 0383fbdc95ff..0911fc311434 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -244,7 +244,7 @@ static int write_begin_slow(struct address_space *mapping,
 	if (unlikely(err))
 		return err;
 
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (unlikely(!page)) {
 		ubifs_release_budget(c, &req);
 		return -ENOMEM;
@@ -437,7 +437,7 @@ static int ubifs_write_begin(struct file *file, struct address_space *mapping,
 		return -EROFS;
 
 	/* Try out the fast-path part first */
-	page = grab_cache_page_write_begin(mapping, index, flags);
+	page = grab_cache_page_write_begin(mapping, index);
 	if (unlikely(!page))
 		return -ENOMEM;
 
diff --git a/fs/udf/file.c b/fs/udf/file.c
index 0f6bf2504437..724bb3141fda 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -94,7 +94,7 @@ static int udf_adinicb_write_begin(struct file *file,
 
 	if (WARN_ON_ONCE(pos >= PAGE_SIZE))
 		return -EIO;
-	page = grab_cache_page_write_begin(mapping, 0, flags);
+	page = grab_cache_page_write_begin(mapping, 0);
 	if (!page)
 		return -ENOMEM;
 	*pagep = page;
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index 993994cd943a..65ae8f96554b 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -735,7 +735,7 @@ static inline unsigned find_get_pages_tag(struct address_space *mapping,
 }
 
 struct page *grab_cache_page_write_begin(struct address_space *mapping,
-			pgoff_t index, unsigned flags);
+			pgoff_t index);
 
 /*
  * Returns locked page at given index in given cache, creating it if needed.
diff --git a/mm/folio-compat.c b/mm/folio-compat.c
index 3e42ddb81918..20bc15b57d93 100644
--- a/mm/folio-compat.c
+++ b/mm/folio-compat.c
@@ -131,7 +131,7 @@ struct page *pagecache_get_page(struct address_space *mapping, pgoff_t index,
 EXPORT_SYMBOL(pagecache_get_page);
 
 struct page *grab_cache_page_write_begin(struct address_space *mapping,
-					pgoff_t index, unsigned flags)
+					pgoff_t index)
 {
 	unsigned fgp_flags = FGP_LOCK | FGP_WRITE | FGP_CREAT | FGP_STABLE;
 
-- 
2.34.1

[PATCH 14/25] fs: Remove aop flags parameter from nobh_write_begin()

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/buffer.c                 | 3 +--
 fs/ext2/inode.c             | 2 +-
 fs/jfs/inode.c              | 3 +--
 include/linux/buffer_head.h | 2 +-
 4 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index 01630218c75f..02b50e3e4fbb 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2568,8 +2568,7 @@ static void attach_nobh_buffers(struct page *page, struct buffer_head *head)
  * On exit the page is fully uptodate in the areas outside (from,to)
  * The filesystem needs to handle block truncation upon failure.
  */
-int nobh_write_begin(struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+int nobh_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata,
 			get_block_t *get_block)
 {
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index 97192932ea56..bfa69c52ce2c 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -917,7 +917,7 @@ ext2_nobh_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = nobh_write_begin(mapping, pos, len, flags, pagep, fsdata,
+	ret = nobh_write_begin(mapping, pos, len, pagep, fsdata,
 			       ext2_get_block);
 	if (ret < 0)
 		ext2_write_failed(mapping, pos + len);
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index d1943a7b4b04..e16f77b4e84c 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -319,8 +319,7 @@ static int jfs_write_begin(struct file *file, struct address_space *mapping,
 {
 	int ret;
 
-	ret = nobh_write_begin(mapping, pos, len, flags, pagep, fsdata,
-				jfs_get_block);
+	ret = nobh_write_begin(mapping, pos, len, pagep, fsdata, jfs_get_block);
 	if (unlikely(ret))
 		jfs_write_failed(mapping, pos + len);
 
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 127b60fad77e..6e5a64005fef 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -258,7 +258,7 @@ static inline vm_fault_t block_page_mkwrite_return(int err)
 }
 sector_t generic_block_bmap(struct address_space *, sector_t, get_block_t *);
 int block_truncate_page(struct address_space *, loff_t, get_block_t *);
-int nobh_write_begin(struct address_space *, loff_t, unsigned, unsigned,
+int nobh_write_begin(struct address_space *, loff_t, unsigned len,
 				struct page **, void **, get_block_t*);
 int nobh_write_end(struct file *, struct address_space *,
 				loff_t, unsigned, unsigned,
-- 
2.34.1

[PATCH 15/25] fs: Remove flags parameter from aops->write_begin

[Matthew Wilcox (Oracle)]

There are no more aop flags left, so remove the parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 Documentation/filesystems/locking.rst |  2 +-
 Documentation/filesystems/vfs.rst     |  5 +----
 block/fops.c                          |  3 +--
 fs/9p/vfs_addr.c                      |  2 +-
 fs/adfs/inode.c                       |  2 +-
 fs/affs/file.c                        |  6 +++---
 fs/afs/internal.h                     |  2 +-
 fs/afs/write.c                        |  2 +-
 fs/bfs/file.c                         |  2 +-
 fs/ceph/addr.c                        |  2 +-
 fs/cifs/file.c                        |  2 +-
 fs/ecryptfs/mmap.c                    |  2 +-
 fs/exfat/inode.c                      |  2 +-
 fs/ext2/inode.c                       |  6 ++----
 fs/ext4/inode.c                       | 10 +++++-----
 fs/f2fs/data.c                        |  5 ++---
 fs/f2fs/super.c                       |  2 +-
 fs/fat/inode.c                        |  2 +-
 fs/fuse/file.c                        |  3 +--
 fs/hfs/inode.c                        |  2 +-
 fs/hfsplus/inode.c                    |  2 +-
 fs/hostfs/hostfs_kern.c               |  2 +-
 fs/hpfs/file.c                        |  2 +-
 fs/hugetlbfs/inode.c                  |  2 +-
 fs/jffs2/file.c                       |  4 ++--
 fs/jfs/inode.c                        |  2 +-
 fs/libfs.c                            |  2 +-
 fs/minix/inode.c                      |  2 +-
 fs/nfs/file.c                         |  2 +-
 fs/nilfs2/inode.c                     |  2 +-
 fs/ntfs3/inode.c                      |  2 +-
 fs/ocfs2/aops.c                       |  2 +-
 fs/omfs/file.c                        |  2 +-
 fs/orangefs/inode.c                   |  5 ++---
 fs/reiserfs/inode.c                   |  2 +-
 fs/sysv/itree.c                       |  2 +-
 fs/ubifs/file.c                       |  7 +++----
 fs/udf/file.c                         |  2 +-
 fs/udf/inode.c                        |  2 +-
 fs/ufs/inode.c                        |  2 +-
 include/linux/fs.h                    |  4 ++--
 include/trace/events/ext4.h           | 21 ++++++++-------------
 include/trace/events/f2fs.h           | 12 ++++--------
 mm/filemap.c                          |  6 ++----
 mm/shmem.c                            |  2 +-
 45 files changed, 69 insertions(+), 90 deletions(-)

diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
index c26d854275a0..fd9d9caf09ab 100644
--- a/Documentation/filesystems/locking.rst
+++ b/Documentation/filesystems/locking.rst
@@ -242,7 +242,7 @@ prototypes::
 	bool (*dirty_folio)(struct address_space *, struct folio *folio);
 	void (*readahead)(struct readahead_control *);
 	int (*write_begin)(struct file *, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
+				loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata);
 	int (*write_end)(struct file *, struct address_space *mapping,
 				loff_t pos, unsigned len, unsigned copied,
diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
index 794bd1a66bfb..30f303180a7d 100644
--- a/Documentation/filesystems/vfs.rst
+++ b/Documentation/filesystems/vfs.rst
@@ -727,7 +727,7 @@ cache in your filesystem.  The following members are defined:
 		bool (*dirty_folio)(struct address_space *, struct folio *);
 		void (*readahead)(struct readahead_control *);
 		int (*write_begin)(struct file *, struct address_space *mapping,
-				   loff_t pos, unsigned len, unsigned flags,
+				   loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata);
 		int (*write_end)(struct file *, struct address_space *mapping,
 				 loff_t pos, unsigned len, unsigned copied,
@@ -832,9 +832,6 @@ cache in your filesystem.  The following members are defined:
 	passed to write_begin is greater than the number of bytes copied
 	into the page).
 
-	flags is a field for AOP_FLAG_xxx flags, described in
-	include/linux/fs.h.
-
 	A void * may be returned in fsdata, which then gets passed into
 	write_end.
 
diff --git a/block/fops.c b/block/fops.c
index b432756570c6..712affe56e29 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -398,8 +398,7 @@ static void blkdev_readahead(struct readahead_control *rac)
 }
 
 static int blkdev_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned len, unsigned flags, struct page **pagep,
-		void **fsdata)
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	return block_write_begin(mapping, pos, len, pagep, blkdev_get_block);
 }
diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c
index d311e68e21fd..a2d57112f53e 100644
--- a/fs/9p/vfs_addr.c
+++ b/fs/9p/vfs_addr.c
@@ -260,7 +260,7 @@ v9fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 }
 
 static int v9fs_write_begin(struct file *filp, struct address_space *mapping,
-			    loff_t pos, unsigned int len, unsigned int flags,
+			    loff_t pos, unsigned int len,
 			    struct page **subpagep, void **fsdata)
 {
 	int retval;
diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c
index b6912496bb19..f7959b1a2d52 100644
--- a/fs/adfs/inode.c
+++ b/fs/adfs/inode.c
@@ -52,7 +52,7 @@ static void adfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int adfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/affs/file.c b/fs/affs/file.c
index 06645d05c717..b952f65c3f06 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -414,7 +414,7 @@ affs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 }
 
 static int affs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
@@ -650,7 +650,7 @@ affs_readpage_ofs(struct file *file, struct page *page)
 }
 
 static int affs_write_begin_ofs(struct file *file, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
+				loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata)
 {
 	struct inode *inode = mapping->host;
@@ -887,7 +887,7 @@ affs_truncate(struct inode *inode)
 		loff_t isize = inode->i_size;
 		int res;
 
-		res = mapping->a_ops->write_begin(NULL, mapping, isize, 0, 0, &page, &fsdata);
+		res = mapping->a_ops->write_begin(NULL, mapping, isize, 0, &page, &fsdata);
 		if (!res)
 			res = mapping->a_ops->write_end(NULL, mapping, isize, 0, 0, page, fsdata);
 		else
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 7b7ef945dc78..7a72e9c60423 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -1535,7 +1535,7 @@ bool afs_dirty_folio(struct address_space *, struct folio *);
 #define afs_dirty_folio filemap_dirty_folio
 #endif
 extern int afs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata);
 extern int afs_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
diff --git a/fs/afs/write.c b/fs/afs/write.c
index af496c98d394..5224e346fbad 100644
--- a/fs/afs/write.c
+++ b/fs/afs/write.c
@@ -42,7 +42,7 @@ static void afs_folio_start_fscache(bool caching, struct folio *folio)
  * prepare to perform part of a write to a page
  */
 int afs_write_begin(struct file *file, struct address_space *mapping,
-		    loff_t pos, unsigned len, unsigned flags,
+		    loff_t pos, unsigned len,
 		    struct page **_page, void **fsdata)
 {
 	struct afs_vnode *vnode = AFS_FS_I(file_inode(file));
diff --git a/fs/bfs/file.c b/fs/bfs/file.c
index 9408f45225cb..dc97c9b8f23b 100644
--- a/fs/bfs/file.c
+++ b/fs/bfs/file.c
@@ -169,7 +169,7 @@ static void bfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int bfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 415f0886bc25..e65541a51b68 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1311,7 +1311,7 @@ static int ceph_netfs_check_write_begin(struct file *file, loff_t pos, unsigned
  * clean, or already dirty within the same snap context.
  */
 static int ceph_write_begin(struct file *file, struct address_space *mapping,
-			    loff_t pos, unsigned len, unsigned aop_flags,
+			    loff_t pos, unsigned len,
 			    struct page **pagep, void **fsdata)
 {
 	struct inode *inode = file_inode(file);
diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index 91aeae7fced8..da362b5a0c96 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -4681,7 +4681,7 @@ bool is_size_safe_to_change(struct cifsInodeInfo *cifsInode, __u64 end_of_file)
 }
 
 static int cifs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int oncethru = 0;
diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c
index 84e399a921ad..47904d40ef88 100644
--- a/fs/ecryptfs/mmap.c
+++ b/fs/ecryptfs/mmap.c
@@ -264,7 +264,7 @@ static int fill_zeros_to_end_of_page(struct page *page, unsigned int to)
  */
 static int ecryptfs_write_begin(struct file *file,
 			struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	pgoff_t index = pos >> PAGE_SHIFT;
diff --git a/fs/exfat/inode.c b/fs/exfat/inode.c
index 8ed3c4b700cd..b9f63113db2d 100644
--- a/fs/exfat/inode.c
+++ b/fs/exfat/inode.c
@@ -389,7 +389,7 @@ static void exfat_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int exfat_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned int len, unsigned int flags,
+		loff_t pos, unsigned int len,
 		struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index bfa69c52ce2c..d8ca8050945a 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -887,8 +887,7 @@ static void ext2_readahead(struct readahead_control *rac)
 
 static int
 ext2_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned len, unsigned flags,
-		struct page **pagep, void **fsdata)
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	int ret;
 
@@ -912,8 +911,7 @@ static int ext2_write_end(struct file *file, struct address_space *mapping,
 
 static int
 ext2_nobh_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned len, unsigned flags,
-		struct page **pagep, void **fsdata)
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	int ret;
 
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 512d8143c765..d3a7e8581291 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -1130,7 +1130,7 @@ static int ext4_block_write_begin(struct page *page, loff_t pos, unsigned len,
 #endif
 
 static int ext4_write_begin(struct file *file, struct address_space *mapping,
-			    loff_t pos, unsigned len, unsigned flags,
+			    loff_t pos, unsigned len,
 			    struct page **pagep, void **fsdata)
 {
 	struct inode *inode = mapping->host;
@@ -1144,7 +1144,7 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping,
 	if (unlikely(ext4_forced_shutdown(EXT4_SB(inode->i_sb))))
 		return -EIO;
 
-	trace_ext4_write_begin(inode, pos, len, flags);
+	trace_ext4_write_begin(inode, pos, len);
 	/*
 	 * Reserve one block more for addition to orphan list in case
 	 * we allocate blocks but write fails for some reason
@@ -2931,7 +2931,7 @@ static int ext4_nonda_switch(struct super_block *sb)
 }
 
 static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
-			       loff_t pos, unsigned len, unsigned flags,
+			       loff_t pos, unsigned len,
 			       struct page **pagep, void **fsdata)
 {
 	int ret, retries = 0;
@@ -2948,10 +2948,10 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping,
 	    ext4_verity_in_progress(inode)) {
 		*fsdata = (void *)FALL_BACK_TO_NONDELALLOC;
 		return ext4_write_begin(file, mapping, pos,
-					len, flags, pagep, fsdata);
+					len, pagep, fsdata);
 	}
 	*fsdata = (void *)0;
-	trace_ext4_da_write_begin(inode, pos, len, flags);
+	trace_ext4_da_write_begin(inode, pos, len);
 
 	if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) {
 		ret = ext4_da_write_inline_data_begin(mapping, inode, pos, len,
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index 9a1a526f2092..b3cf49136b9f 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -3314,8 +3314,7 @@ static int prepare_write_begin(struct f2fs_sb_info *sbi,
 }
 
 static int f2fs_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned len, unsigned flags,
-		struct page **pagep, void **fsdata)
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	struct inode *inode = mapping->host;
 	struct f2fs_sb_info *sbi = F2FS_I_SB(inode);
@@ -3325,7 +3324,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping,
 	block_t blkaddr = NULL_ADDR;
 	int err = 0;
 
-	trace_f2fs_write_begin(inode, pos, len, flags);
+	trace_f2fs_write_begin(inode, pos, len);
 
 	if (!f2fs_is_checkpoint_ready(sbi)) {
 		err = -ENOSPC;
diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c
index 4368f90571bd..ed3e8b7a8260 100644
--- a/fs/f2fs/super.c
+++ b/fs/f2fs/super.c
@@ -2483,7 +2483,7 @@ static ssize_t f2fs_quota_write(struct super_block *sb, int type,
 		tocopy = min_t(unsigned long, sb->s_blocksize - offset,
 								towrite);
 retry:
-		err = a_ops->write_begin(NULL, mapping, off, tocopy, 0,
+		err = a_ops->write_begin(NULL, mapping, off, tocopy,
 							&page, &fsdata);
 		if (unlikely(err)) {
 			if (err == -ENOMEM) {
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index 9b34ccef2501..1f15b0fd1bb0 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -226,7 +226,7 @@ static void fat_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int fat_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int err;
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index e35e394264ad..bca8c2135ec5 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -2273,8 +2273,7 @@ static int fuse_writepages(struct address_space *mapping,
  * but how to implement it without killing performance need more thinking.
  */
 static int fuse_write_begin(struct file *file, struct address_space *mapping,
-		loff_t pos, unsigned len, unsigned flags,
-		struct page **pagep, void **fsdata)
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	pgoff_t index = pos >> PAGE_SHIFT;
 	struct fuse_conn *fc = get_fuse_conn(file_inode(file));
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 396735dd3407..93d9aa832139 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -50,7 +50,7 @@ static void hfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int hfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 435b6202532a..73010aa4623f 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -44,7 +44,7 @@ static void hfsplus_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int hfsplus_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index 2bfd316e1bf1..e658d8edde35 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -463,7 +463,7 @@ static int hostfs_readpage(struct file *file, struct page *page)
 }
 
 static int hostfs_write_begin(struct file *file, struct address_space *mapping,
-			      loff_t pos, unsigned len, unsigned flags,
+			      loff_t pos, unsigned len,
 			      struct page **pagep, void **fsdata)
 {
 	pgoff_t index = pos >> PAGE_SHIFT;
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index 8740b4ea0b52..8b590b3826c3 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -194,7 +194,7 @@ static void hpfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int hpfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index dd3a088db11d..2de9ca5d260d 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -383,7 +383,7 @@ static ssize_t hugetlbfs_read_iter(struct kiocb *iocb, struct iov_iter *to)
 
 static int hugetlbfs_write_begin(struct file *file,
 			struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	return -EINVAL;
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index 142d3ba9f0a8..2b35811772de 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -25,7 +25,7 @@ static int jffs2_write_end(struct file *filp, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
 			struct page *pg, void *fsdata);
 static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata);
 static int jffs2_readpage (struct file *filp, struct page *pg);
 
@@ -130,7 +130,7 @@ static int jffs2_readpage (struct file *filp, struct page *pg)
 }
 
 static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	struct page *pg;
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index e16f77b4e84c..aa9f112107b2 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -314,7 +314,7 @@ static void jfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int jfs_write_begin(struct file *file, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
+				loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/libfs.c b/fs/libfs.c
index d4395e1c6696..a1c10d3163e0 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -549,7 +549,7 @@ static int simple_readpage(struct file *file, struct page *page)
 }
 
 int simple_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	struct page *page;
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 5e8d7ba661cf..3add78bccedc 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -423,7 +423,7 @@ static void minix_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int minix_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index d66088dd33e7..314d2d7ba84a 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -313,7 +313,7 @@ static bool nfs_want_read_modify_write(struct file *file, struct page *page,
  * increment the page use counts until he is done with the page.
  */
 static int nfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index be09a0d10f04..02297ec8dc55 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -248,7 +248,7 @@ void nilfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int nilfs_write_begin(struct file *file, struct address_space *mapping,
-			     loff_t pos, unsigned len, unsigned flags,
+			     loff_t pos, unsigned len,
 			     struct page **pagep, void **fsdata)
 
 {
diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c
index 16466c8648f3..1364174cc6c9 100644
--- a/fs/ntfs3/inode.c
+++ b/fs/ntfs3/inode.c
@@ -862,7 +862,7 @@ static int ntfs_get_block_write_begin(struct inode *inode, sector_t vbn,
 }
 
 static int ntfs_write_begin(struct file *file, struct address_space *mapping,
-			    loff_t pos, u32 len, u32 flags, struct page **pagep,
+			    loff_t pos, u32 len, struct page **pagep,
 			    void **fsdata)
 {
 	int err;
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 4b9af65cb61b..7cffe9dcad17 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -1881,7 +1881,7 @@ int ocfs2_write_begin_nolock(struct address_space *mapping,
 }
 
 static int ocfs2_write_begin(struct file *file, struct address_space *mapping,
-			     loff_t pos, unsigned len, unsigned flags,
+			     loff_t pos, unsigned len,
 			     struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/omfs/file.c b/fs/omfs/file.c
index 349b96d89c44..980b0a72c172 100644
--- a/fs/omfs/file.c
+++ b/fs/omfs/file.c
@@ -316,7 +316,7 @@ static void omfs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int omfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 809690db8be2..bc7ccd15d7a3 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -326,9 +326,8 @@ static int orangefs_readpage(struct file *file, struct page *page)
 }
 
 static int orangefs_write_begin(struct file *file,
-    struct address_space *mapping,
-    loff_t pos, unsigned len, unsigned flags, struct page **pagep,
-    void **fsdata)
+		struct address_space *mapping, loff_t pos, unsigned len,
+		struct page **pagep, void **fsdata)
 {
 	struct orangefs_write_range *wr;
 	struct folio *folio;
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index aa31cf1dbba6..46ba4892030a 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -2753,7 +2753,7 @@ static void reiserfs_truncate_failed_write(struct inode *inode)
 
 static int reiserfs_write_begin(struct file *file,
 				struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
+				loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata)
 {
 	struct inode *inode;
diff --git a/fs/sysv/itree.c b/fs/sysv/itree.c
index 96b7fd4facf3..96ad24fe0ffb 100644
--- a/fs/sysv/itree.c
+++ b/fs/sysv/itree.c
@@ -477,7 +477,7 @@ static void sysv_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int sysv_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 0911fc311434..81c085c4decf 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -215,8 +215,7 @@ static void release_existing_page_budget(struct ubifs_info *c)
 }
 
 static int write_begin_slow(struct address_space *mapping,
-			    loff_t pos, unsigned len, struct page **pagep,
-			    unsigned flags)
+			    loff_t pos, unsigned len, struct page **pagep)
 {
 	struct inode *inode = mapping->host;
 	struct ubifs_info *c = inode->i_sb->s_fs_info;
@@ -419,7 +418,7 @@ static int allocate_budget(struct ubifs_info *c, struct page *page,
  * without forcing write-back. The slow path does not make this assumption.
  */
 static int ubifs_write_begin(struct file *file, struct address_space *mapping,
-			     loff_t pos, unsigned len, unsigned flags,
+			     loff_t pos, unsigned len,
 			     struct page **pagep, void **fsdata)
 {
 	struct inode *inode = mapping->host;
@@ -493,7 +492,7 @@ static int ubifs_write_begin(struct file *file, struct address_space *mapping,
 		unlock_page(page);
 		put_page(page);
 
-		return write_begin_slow(mapping, pos, len, pagep, flags);
+		return write_begin_slow(mapping, pos, len, pagep);
 	}
 
 	/*
diff --git a/fs/udf/file.c b/fs/udf/file.c
index 724bb3141fda..3f4d5c44c784 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -87,7 +87,7 @@ static int udf_adinicb_writepage(struct page *page,
 
 static int udf_adinicb_write_begin(struct file *file,
 			struct address_space *mapping, loff_t pos,
-			unsigned len, unsigned flags, struct page **pagep,
+			unsigned len, struct page **pagep,
 			void **fsdata)
 {
 	struct page *page;
diff --git a/fs/udf/inode.c b/fs/udf/inode.c
index 88a95886ce8a..866f9a53248e 100644
--- a/fs/udf/inode.c
+++ b/fs/udf/inode.c
@@ -204,7 +204,7 @@ static void udf_readahead(struct readahead_control *rac)
 }
 
 static int udf_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c
index bd0e0c66f93d..6c973b71cab2 100644
--- a/fs/ufs/inode.c
+++ b/fs/ufs/inode.c
@@ -495,7 +495,7 @@ static void ufs_write_failed(struct address_space *mapping, loff_t to)
 }
 
 static int ufs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	int ret;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index f81bc5cbcbb6..a0e73432526f 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -346,7 +346,7 @@ struct address_space_operations {
 	void (*readahead)(struct readahead_control *);
 
 	int (*write_begin)(struct file *, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
+				loff_t pos, unsigned len,
 				struct page **pagep, void **fsdata);
 	int (*write_end)(struct file *, struct address_space *mapping,
 				loff_t pos, unsigned len, unsigned copied,
@@ -3179,7 +3179,7 @@ extern int noop_fsync(struct file *, loff_t, loff_t, int);
 extern ssize_t noop_direct_IO(struct kiocb *iocb, struct iov_iter *iter);
 extern int simple_empty(struct dentry *);
 extern int simple_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata);
 extern const struct address_space_operations ram_aops;
 extern int always_delete_dentry(const struct dentry *);
diff --git a/include/trace/events/ext4.h b/include/trace/events/ext4.h
index d06ffffad434..229e8fae66a3 100644
--- a/include/trace/events/ext4.h
+++ b/include/trace/events/ext4.h
@@ -335,17 +335,15 @@ TRACE_EVENT(ext4_begin_ordered_truncate,
 
 DECLARE_EVENT_CLASS(ext4__write_begin,
 
-	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len,
-		 unsigned int flags),
+	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len),
 
-	TP_ARGS(inode, pos, len, flags),
+	TP_ARGS(inode, pos, len),
 
 	TP_STRUCT__entry(
 		__field(	dev_t,	dev			)
 		__field(	ino_t,	ino			)
 		__field(	loff_t,	pos			)
 		__field(	unsigned int, len		)
-		__field(	unsigned int, flags		)
 	),
 
 	TP_fast_assign(
@@ -353,29 +351,26 @@ DECLARE_EVENT_CLASS(ext4__write_begin,
 		__entry->ino	= inode->i_ino;
 		__entry->pos	= pos;
 		__entry->len	= len;
-		__entry->flags	= flags;
 	),
 
-	TP_printk("dev %d,%d ino %lu pos %lld len %u flags %u",
+	TP_printk("dev %d,%d ino %lu pos %lld len %u",
 		  MAJOR(__entry->dev), MINOR(__entry->dev),
 		  (unsigned long) __entry->ino,
-		  __entry->pos, __entry->len, __entry->flags)
+		  __entry->pos, __entry->len)
 );
 
 DEFINE_EVENT(ext4__write_begin, ext4_write_begin,
 
-	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len,
-		 unsigned int flags),
+	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len),
 
-	TP_ARGS(inode, pos, len, flags)
+	TP_ARGS(inode, pos, len)
 );
 
 DEFINE_EVENT(ext4__write_begin, ext4_da_write_begin,
 
-	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len,
-		 unsigned int flags),
+	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len),
 
-	TP_ARGS(inode, pos, len, flags)
+	TP_ARGS(inode, pos, len)
 );
 
 DECLARE_EVENT_CLASS(ext4__write_end,
diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h
index 1779e133cea0..bea654a85e6b 100644
--- a/include/trace/events/f2fs.h
+++ b/include/trace/events/f2fs.h
@@ -1159,17 +1159,15 @@ DEFINE_EVENT_CONDITION(f2fs__bio, f2fs_submit_write_bio,
 
 TRACE_EVENT(f2fs_write_begin,
 
-	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len,
-				unsigned int flags),
+	TP_PROTO(struct inode *inode, loff_t pos, unsigned int len),
 
-	TP_ARGS(inode, pos, len, flags),
+	TP_ARGS(inode, pos, len),
 
 	TP_STRUCT__entry(
 		__field(dev_t,	dev)
 		__field(ino_t,	ino)
 		__field(loff_t,	pos)
 		__field(unsigned int, len)
-		__field(unsigned int, flags)
 	),
 
 	TP_fast_assign(
@@ -1177,14 +1175,12 @@ TRACE_EVENT(f2fs_write_begin,
 		__entry->ino	= inode->i_ino;
 		__entry->pos	= pos;
 		__entry->len	= len;
-		__entry->flags	= flags;
 	),
 
-	TP_printk("dev = (%d,%d), ino = %lu, pos = %llu, len = %u, flags = %u",
+	TP_printk("dev = (%d,%d), ino = %lu, pos = %llu, len = %u",
 		show_dev_ino(__entry),
 		(unsigned long long)__entry->pos,
-		__entry->len,
-		__entry->flags)
+		__entry->len)
 );
 
 TRACE_EVENT(f2fs_write_end,
diff --git a/mm/filemap.c b/mm/filemap.c
index 9a1eef6c5d35..0751843b052f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3628,8 +3628,7 @@ int pagecache_write_begin(struct file *file, struct address_space *mapping,
 {
 	const struct address_space_operations *aops = mapping->a_ops;
 
-	return aops->write_begin(file, mapping, pos, len, flags,
-							pagep, fsdata);
+	return aops->write_begin(file, mapping, pos, len, pagep, fsdata);
 }
 EXPORT_SYMBOL(pagecache_write_begin);
 
@@ -3754,7 +3753,6 @@ ssize_t generic_perform_write(struct kiocb *iocb, struct iov_iter *i)
 	const struct address_space_operations *a_ops = mapping->a_ops;
 	long status = 0;
 	ssize_t written = 0;
-	unsigned int flags = 0;
 
 	do {
 		struct page *page;
@@ -3784,7 +3782,7 @@ ssize_t generic_perform_write(struct kiocb *iocb, struct iov_iter *i)
 			break;
 		}
 
-		status = a_ops->write_begin(file, mapping, pos, bytes, flags,
+		status = a_ops->write_begin(file, mapping, pos, bytes,
 						&page, &fsdata);
 		if (unlikely(status < 0))
 			break;
diff --git a/mm/shmem.c b/mm/shmem.c
index 4b2fea33158e..0f557a512171 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2426,7 +2426,7 @@ static int shmem_initxattrs(struct inode *, const struct xattr *, void *);
 
 static int
 shmem_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len, unsigned flags,
+			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata)
 {
 	struct inode *inode = mapping->host;
-- 
2.34.1

[PATCH 16/25] buffer: Call aops write_begin() and write_end() directly

[Matthew Wilcox (Oracle)]

pagecache_write_begin() and pagecache_write_end() are now trivial
wrappers, so call the aops directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/buffer.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index 02b50e3e4fbb..d538495a0553 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2344,6 +2344,7 @@ EXPORT_SYMBOL(block_read_full_page);
 int generic_cont_expand_simple(struct inode *inode, loff_t size)
 {
 	struct address_space *mapping = inode->i_mapping;
+	const struct address_space_operations *aops = mapping->a_ops;
 	struct page *page;
 	void *fsdata;
 	int err;
@@ -2352,11 +2353,11 @@ int generic_cont_expand_simple(struct inode *inode, loff_t size)
 	if (err)
 		goto out;
 
-	err = pagecache_write_begin(NULL, mapping, size, 0, 0, &page, &fsdata);
+	err = aops->write_begin(NULL, mapping, size, 0, &page, &fsdata);
 	if (err)
 		goto out;
 
-	err = pagecache_write_end(NULL, mapping, size, 0, 0, page, fsdata);
+	err = aops->write_end(NULL, mapping, size, 0, 0, page, fsdata);
 	BUG_ON(err > 0);
 
 out:
@@ -2368,6 +2369,7 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping,
 			    loff_t pos, loff_t *bytes)
 {
 	struct inode *inode = mapping->host;
+	const struct address_space_operations *aops = mapping->a_ops;
 	unsigned int blocksize = i_blocksize(inode);
 	struct page *page;
 	void *fsdata;
@@ -2387,12 +2389,12 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping,
 		}
 		len = PAGE_SIZE - zerofrom;
 
-		err = pagecache_write_begin(file, mapping, curpos, len, 0,
+		err = aops->write_begin(file, mapping, curpos, len,
 					    &page, &fsdata);
 		if (err)
 			goto out;
 		zero_user(page, zerofrom, len);
-		err = pagecache_write_end(file, mapping, curpos, len, len,
+		err = aops->write_end(file, mapping, curpos, len, len,
 						page, fsdata);
 		if (err < 0)
 			goto out;
@@ -2420,12 +2422,12 @@ static int cont_expand_zero(struct file *file, struct address_space *mapping,
 		}
 		len = offset - zerofrom;
 
-		err = pagecache_write_begin(file, mapping, curpos, len, 0,
+		err = aops->write_begin(file, mapping, curpos, len,
 					    &page, &fsdata);
 		if (err)
 			goto out;
 		zero_user(page, zerofrom, len);
-		err = pagecache_write_end(file, mapping, curpos, len, len,
+		err = aops->write_end(file, mapping, curpos, len, len,
 						page, fsdata);
 		if (err < 0)
 			goto out;
-- 
2.34.1

[PATCH 17/25] namei: Call aops write_begin() and write_end() directly

[Matthew Wilcox (Oracle)]

pagecache_write_begin() and pagecache_write_end() are now trivial
wrappers, so call the aops directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/namei.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/namei.c b/fs/namei.c
index 0c84b4326dc9..896ade8b7400 100644
--- a/fs/namei.c
+++ b/fs/namei.c
@@ -5005,6 +5005,7 @@ EXPORT_SYMBOL(page_readlink);
 int page_symlink(struct inode *inode, const char *symname, int len)
 {
 	struct address_space *mapping = inode->i_mapping;
+	const struct address_space_operations *aops = mapping->a_ops;
 	bool nofs = !mapping_gfp_constraint(mapping, __GFP_FS);
 	struct page *page;
 	void *fsdata;
@@ -5014,8 +5015,7 @@ int page_symlink(struct inode *inode, const char *symname, int len)
 retry:
 	if (nofs)
 		flags = memalloc_nofs_save();
-	err = pagecache_write_begin(NULL, mapping, 0, len-1,
-				0, &page, &fsdata);
+	err = aops->write_begin(NULL, mapping, 0, len-1, &page, &fsdata);
 	if (nofs)
 		memalloc_nofs_restore(flags);
 	if (err)
@@ -5023,7 +5023,7 @@ int page_symlink(struct inode *inode, const char *symname, int len)
 
 	memcpy(page_address(page), symname, len-1);
 
-	err = pagecache_write_end(NULL, mapping, 0, len-1, len-1,
+	err = aops->write_end(NULL, mapping, 0, len-1, len-1,
 							page, fsdata);
 	if (err < 0)
 		goto fail;
-- 
2.34.1

[PATCH 18/25] ntfs3: Call ntfs_write_begin() and ntfs_write_end() directly

[Matthew Wilcox (Oracle)]

There is only one kind of write_begin/write_end aops, so we don't need to
look up which aop it is, just make ntfs_write_begin() and ntfs_write_end()
available to this file and call them directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Namjae Jeon <linkinjeon@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/ntfs3/file.c    |  5 ++---
 fs/ntfs3/inode.c   | 12 +++++-------
 fs/ntfs3/ntfs_fs.h |  5 +++++
 3 files changed, 12 insertions(+), 10 deletions(-)

diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c
index 787b53b984ee..c2e7e561958a 100644
--- a/fs/ntfs3/file.c
+++ b/fs/ntfs3/file.c
@@ -157,15 +157,14 @@ static int ntfs_extend_initialized_size(struct file *file,
 		if (pos + len > new_valid)
 			len = new_valid - pos;
 
-		err = pagecache_write_begin(file, mapping, pos, len, 0, &page,
-					    &fsdata);
+		err = ntfs_write_begin(file, mapping, pos, len, &page, &fsdata);
 		if (err)
 			goto out;
 
 		zero_user_segment(page, zerofrom, PAGE_SIZE);
 
 		/* This function in any case puts page. */
-		err = pagecache_write_end(file, mapping, pos, len, len, page,
+		err = ntfs_write_end(file, mapping, pos, len, len, page,
 					  fsdata);
 		if (err < 0)
 			goto out;
diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c
index 1364174cc6c9..bfd71f384e21 100644
--- a/fs/ntfs3/inode.c
+++ b/fs/ntfs3/inode.c
@@ -861,9 +861,8 @@ static int ntfs_get_block_write_begin(struct inode *inode, sector_t vbn,
 				  bh_result, create, GET_BLOCK_WRITE_BEGIN);
 }
 
-static int ntfs_write_begin(struct file *file, struct address_space *mapping,
-			    loff_t pos, u32 len, struct page **pagep,
-			    void **fsdata)
+int ntfs_write_begin(struct file *file, struct address_space *mapping,
+		     loff_t pos, u32 len, struct page **pagep, void **fsdata)
 {
 	int err;
 	struct inode *inode = mapping->host;
@@ -904,10 +903,9 @@ static int ntfs_write_begin(struct file *file, struct address_space *mapping,
 /*
  * ntfs_write_end - Address_space_operations::write_end.
  */
-static int ntfs_write_end(struct file *file, struct address_space *mapping,
-			  loff_t pos, u32 len, u32 copied, struct page *page,
-			  void *fsdata)
-
+int ntfs_write_end(struct file *file, struct address_space *mapping,
+		   loff_t pos, u32 len, u32 copied, struct page *page,
+		   void *fsdata)
 {
 	struct inode *inode = mapping->host;
 	struct ntfs_inode *ni = ntfs_i(inode);
diff --git a/fs/ntfs3/ntfs_fs.h b/fs/ntfs3/ntfs_fs.h
index fb825059d488..8de129a6419b 100644
--- a/fs/ntfs3/ntfs_fs.h
+++ b/fs/ntfs3/ntfs_fs.h
@@ -689,6 +689,11 @@ int ntfs_set_size(struct inode *inode, u64 new_size);
 int reset_log_file(struct inode *inode);
 int ntfs_get_block(struct inode *inode, sector_t vbn,
 		   struct buffer_head *bh_result, int create);
+int ntfs_write_begin(struct file *file, struct address_space *mapping,
+		     loff_t pos, u32 len, struct page **pagep, void **fsdata);
+int ntfs_write_end(struct file *file, struct address_space *mapping,
+		   loff_t pos, u32 len, u32 copied, struct page *page,
+		   void *fsdata);
 int ntfs3_write_inode(struct inode *inode, struct writeback_control *wbc);
 int ntfs_sync_inode(struct inode *inode);
 int ntfs_flush_inodes(struct super_block *sb, struct inode *i1,
-- 
2.34.1

[PATCH 19/25] ntfs3: Remove fsdata parameter from ntfs_extend_initialized_size()

[Matthew Wilcox (Oracle)]

After the last patch, Smatch reports:

        fs/ntfs3/file.c:168 ntfs_extend_initialized_size()
        error: uninitialized symbol 'fsdata'.

fsdata is indeed unused.  This is not new, but Smatch couldn't see it
before because calls through pagecache_write_begin()/pagecache_write_end()
could theoretically call any implemention of ->write_begin/write_end,
some of which do use fsdata.  Now that the calls are direct, Smatch can
see they're never used.

Fix this by simply passing NULL.  While ntfs3 does pass this parameter
on to generic functions, those generic functions also never dereference
the fsdata parameter, so it's unnecessary to pass the address of a real
pointer.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ntfs3/file.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/fs/ntfs3/file.c b/fs/ntfs3/file.c
index c2e7e561958a..e61f335c9c63 100644
--- a/fs/ntfs3/file.c
+++ b/fs/ntfs3/file.c
@@ -115,7 +115,6 @@ static int ntfs_extend_initialized_size(struct file *file,
 	for (;;) {
 		u32 zerofrom, len;
 		struct page *page;
-		void *fsdata;
 		u8 bits;
 		CLST vcn, lcn, clen;
 
@@ -157,15 +156,14 @@ static int ntfs_extend_initialized_size(struct file *file,
 		if (pos + len > new_valid)
 			len = new_valid - pos;
 
-		err = ntfs_write_begin(file, mapping, pos, len, &page, &fsdata);
+		err = ntfs_write_begin(file, mapping, pos, len, &page, NULL);
 		if (err)
 			goto out;
 
 		zero_user_segment(page, zerofrom, PAGE_SIZE);
 
 		/* This function in any case puts page. */
-		err = ntfs_write_end(file, mapping, pos, len, len, page,
-					  fsdata);
+		err = ntfs_write_end(file, mapping, pos, len, len, page, NULL);
 		if (err < 0)
 			goto out;
 		pos += len;
-- 
2.34.1

[PATCH 20/25] hfs: Call hfs_write_begin() and generic_write_end() directly

[Matthew Wilcox (Oracle)]

There is only one kind of write_begin/write_end aops, so we don't need
to look up which aop it is, just make hfs_write_begin() available to
this file and call it directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/hfs/extent.c | 6 +++---
 fs/hfs/hfs_fs.h | 2 ++
 fs/hfs/inode.c  | 5 ++---
 3 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/fs/hfs/extent.c b/fs/hfs/extent.c
index 263d5028d9d1..3f7e9bef9874 100644
--- a/fs/hfs/extent.c
+++ b/fs/hfs/extent.c
@@ -491,10 +491,10 @@ void hfs_file_truncate(struct inode *inode)
 
 		/* XXX: Can use generic_cont_expand? */
 		size = inode->i_size - 1;
-		res = pagecache_write_begin(NULL, mapping, size+1, 0, 0,
-					    &page, &fsdata);
+		res = hfs_write_begin(NULL, mapping, size + 1, 0, &page,
+				&fsdata);
 		if (!res) {
-			res = pagecache_write_end(NULL, mapping, size+1, 0, 0,
+			res = generic_write_end(NULL, mapping, size + 1, 0, 0,
 					page, fsdata);
 		}
 		if (res)
diff --git a/fs/hfs/hfs_fs.h b/fs/hfs/hfs_fs.h
index b8eb0322a3e5..68d0305880f7 100644
--- a/fs/hfs/hfs_fs.h
+++ b/fs/hfs/hfs_fs.h
@@ -201,6 +201,8 @@ extern int hfs_get_block(struct inode *, sector_t, struct buffer_head *, int);
 extern const struct address_space_operations hfs_aops;
 extern const struct address_space_operations hfs_btree_aops;
 
+int hfs_write_begin(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata);
 extern struct inode *hfs_new_inode(struct inode *, const struct qstr *, umode_t);
 extern void hfs_inode_write_fork(struct inode *, struct hfs_extent *, __be32 *, __be32 *);
 extern int hfs_write_inode(struct inode *, struct writeback_control *);
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 93d9aa832139..9a26b9510da0 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -49,9 +49,8 @@ static void hfs_write_failed(struct address_space *mapping, loff_t to)
 	}
 }
 
-static int hfs_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len,
-			struct page **pagep, void **fsdata)
+int hfs_write_begin(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	int ret;
 
-- 
2.34.1

[PATCH 21/25] hfsplus: Call hfsplus_write_begin() and generic_write_end() directly

[Matthew Wilcox (Oracle)]

There is only one kind of write_begin/write_end aops, so we don't need
to look up which aop it is, just make hfsplus_write_begin() available to
this file and call it directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/hfsplus/extents.c    | 8 ++++----
 fs/hfsplus/hfsplus_fs.h | 2 ++
 fs/hfsplus/inode.c      | 5 ++---
 3 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/hfsplus/extents.c b/fs/hfsplus/extents.c
index 7054a542689f..721f779b4ec3 100644
--- a/fs/hfsplus/extents.c
+++ b/fs/hfsplus/extents.c
@@ -557,12 +557,12 @@ void hfsplus_file_truncate(struct inode *inode)
 		void *fsdata;
 		loff_t size = inode->i_size;
 
-		res = pagecache_write_begin(NULL, mapping, size, 0, 0,
-					    &page, &fsdata);
+		res = hfsplus_write_begin(NULL, mapping, size, 0,
+					  &page, &fsdata);
 		if (res)
 			return;
-		res = pagecache_write_end(NULL, mapping, size,
-			0, 0, page, fsdata);
+		res = generic_write_end(NULL, mapping, size, 0, 0,
+					page, fsdata);
 		if (res < 0)
 			return;
 		mark_inode_dirty(inode);
diff --git a/fs/hfsplus/hfsplus_fs.h b/fs/hfsplus/hfsplus_fs.h
index 1798949f269b..396e73aa0961 100644
--- a/fs/hfsplus/hfsplus_fs.h
+++ b/fs/hfsplus/hfsplus_fs.h
@@ -468,6 +468,8 @@ extern const struct address_space_operations hfsplus_aops;
 extern const struct address_space_operations hfsplus_btree_aops;
 extern const struct dentry_operations hfsplus_dentry_operations;
 
+int hfsplus_write_begin(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata);
 struct inode *hfsplus_new_inode(struct super_block *sb, struct inode *dir,
 				umode_t mode);
 void hfsplus_delete_inode(struct inode *inode);
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 73010aa4623f..905ae3660315 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -43,9 +43,8 @@ static void hfsplus_write_failed(struct address_space *mapping, loff_t to)
 	}
 }
 
-static int hfsplus_write_begin(struct file *file, struct address_space *mapping,
-			loff_t pos, unsigned len,
-			struct page **pagep, void **fsdata)
+int hfsplus_write_begin(struct file *file, struct address_space *mapping,
+		loff_t pos, unsigned len, struct page **pagep, void **fsdata)
 {
 	int ret;
 
-- 
2.34.1

[PATCH 22/25] ext4: Call aops write_begin() and write_end() directly

[Matthew Wilcox (Oracle)]

pagecache_write_begin() and pagecache_write_end() are now trivial
wrappers, so call the aops directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/ext4/verity.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/ext4/verity.c b/fs/ext4/verity.c
index eacbd489e3bf..b051d19b5c8a 100644
--- a/fs/ext4/verity.c
+++ b/fs/ext4/verity.c
@@ -69,6 +69,9 @@ static int pagecache_read(struct inode *inode, void *buf, size_t count,
 static int pagecache_write(struct inode *inode, const void *buf, size_t count,
 			   loff_t pos)
 {
+	struct address_space *mapping = inode->i_mapping;
+	const struct address_space_operations *aops = mapping->a_ops;
+
 	if (pos + count > inode->i_sb->s_maxbytes)
 		return -EFBIG;
 
@@ -79,15 +82,13 @@ static int pagecache_write(struct inode *inode, const void *buf, size_t count,
 		void *fsdata;
 		int res;
 
-		res = pagecache_write_begin(NULL, inode->i_mapping, pos, n, 0,
-					    &page, &fsdata);
+		res = aops->write_begin(NULL, mapping, pos, n, &page, &fsdata);
 		if (res)
 			return res;
 
 		memcpy_to_page(page, offset_in_page(pos), buf, n);
 
-		res = pagecache_write_end(NULL, inode->i_mapping, pos, n, n,
-					  page, fsdata);
+		res = aops->write_end(NULL, mapping, pos, n, n, page, fsdata);
 		if (res < 0)
 			return res;
 		if (res != n)
-- 
2.34.1

[PATCH 23/25] f2fs: Call aops write_begin() and write_end() directly

[Matthew Wilcox (Oracle)]

pagecache_write_begin() and pagecache_write_end() are now trivial
wrappers, so call the aops directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/f2fs/verity.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/f2fs/verity.c b/fs/f2fs/verity.c
index 3d793202cc9f..65395ae188aa 100644
--- a/fs/f2fs/verity.c
+++ b/fs/f2fs/verity.c
@@ -74,6 +74,9 @@ static int pagecache_read(struct inode *inode, void *buf, size_t count,
 static int pagecache_write(struct inode *inode, const void *buf, size_t count,
 			   loff_t pos)
 {
+	struct address_space *mapping = inode->i_mapping;
+	const struct address_space_operations *aops = mapping->a_ops;
+
 	if (pos + count > inode->i_sb->s_maxbytes)
 		return -EFBIG;
 
@@ -85,8 +88,7 @@ static int pagecache_write(struct inode *inode, const void *buf, size_t count,
 		void *addr;
 		int res;
 
-		res = pagecache_write_begin(NULL, inode->i_mapping, pos, n, 0,
-					    &page, &fsdata);
+		res = aops->write_begin(NULL, mapping, pos, n, &page, &fsdata);
 		if (res)
 			return res;
 
@@ -94,8 +96,7 @@ static int pagecache_write(struct inode *inode, const void *buf, size_t count,
 		memcpy(addr + offset_in_page(pos), buf, n);
 		kunmap_atomic(addr);
 
-		res = pagecache_write_end(NULL, inode->i_mapping, pos, n, n,
-					  page, fsdata);
+		res = aops->write_end(NULL, mapping, pos, n, n, page, fsdata);
 		if (res < 0)
 			return res;
 		if (res != n)
-- 
2.34.1

[PATCH 24/25] i915: Call aops write_begin() and write_end() directly

[Matthew Wilcox (Oracle)]

pagecache_write_begin() and pagecache_write_end() are now trivial
wrappers, so call the aops directly.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 drivers/gpu/drm/i915/gem/i915_gem_shmem.c | 23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
index 3a1c782ed791..e92cc9d7257c 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_shmem.c
@@ -408,6 +408,7 @@ shmem_pwrite(struct drm_i915_gem_object *obj,
 	     const struct drm_i915_gem_pwrite *arg)
 {
 	struct address_space *mapping = obj->base.filp->f_mapping;
+	const struct address_space_operations *aops = mapping->a_ops;
 	char __user *user_data = u64_to_user_ptr(arg->data_ptr);
 	u64 remain, offset;
 	unsigned int pg;
@@ -465,9 +466,8 @@ shmem_pwrite(struct drm_i915_gem_object *obj,
 		if (err)
 			return err;
 
-		err = pagecache_write_begin(obj->base.filp, mapping,
-					    offset, len, 0,
-					    &page, &data);
+		err = aops->write_begin(obj->base.filp, mapping, offset, len,
+					&page, &data);
 		if (err < 0)
 			return err;
 
@@ -477,9 +477,8 @@ shmem_pwrite(struct drm_i915_gem_object *obj,
 						      len);
 		kunmap_atomic(vaddr);
 
-		err = pagecache_write_end(obj->base.filp, mapping,
-					  offset, len, len - unwritten,
-					  page, data);
+		err = aops->write_end(obj->base.filp, mapping, offset, len,
+				      len - unwritten, page, data);
 		if (err < 0)
 			return err;
 
@@ -622,6 +621,7 @@ i915_gem_object_create_shmem_from_data(struct drm_i915_private *dev_priv,
 {
 	struct drm_i915_gem_object *obj;
 	struct file *file;
+	const struct address_space_operations *aops;
 	resource_size_t offset;
 	int err;
 
@@ -633,15 +633,15 @@ i915_gem_object_create_shmem_from_data(struct drm_i915_private *dev_priv,
 	GEM_BUG_ON(obj->write_domain != I915_GEM_DOMAIN_CPU);
 
 	file = obj->base.filp;
+	aops = file->f_mapping->a_ops;
 	offset = 0;
 	do {
 		unsigned int len = min_t(typeof(size), size, PAGE_SIZE);
 		struct page *page;
 		void *pgdata, *vaddr;
 
-		err = pagecache_write_begin(file, file->f_mapping,
-					    offset, len, 0,
-					    &page, &pgdata);
+		err = aops->write_begin(file, file->f_mapping, offset, len,
+					&page, &pgdata);
 		if (err < 0)
 			goto fail;
 
@@ -649,9 +649,8 @@ i915_gem_object_create_shmem_from_data(struct drm_i915_private *dev_priv,
 		memcpy(vaddr, data, len);
 		kunmap(page);
 
-		err = pagecache_write_end(file, file->f_mapping,
-					  offset, len, len,
-					  page, pgdata);
+		err = aops->write_end(file, file->f_mapping, offset, len, len,
+				      page, pgdata);
 		if (err < 0)
 			goto fail;
 
-- 
2.34.1

[PATCH 25/25] fs: Remove pagecache_write_begin() and pagecache_write_end()

[Matthew Wilcox (Oracle)]

These wrappers have no more users; remove them.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 include/linux/fs.h | 12 ------------
 mm/filemap.c       | 20 --------------------
 2 files changed, 32 deletions(-)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index a0e73432526f..b35ce086a7a1 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -380,18 +380,6 @@ struct address_space_operations {
 
 extern const struct address_space_operations empty_aops;
 
-/*
- * pagecache_write_begin/pagecache_write_end must be used by general code
- * to write into the pagecache.
- */
-int pagecache_write_begin(struct file *, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
-				struct page **pagep, void **fsdata);
-
-int pagecache_write_end(struct file *, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned copied,
-				struct page *page, void *fsdata);
-
 /**
  * struct address_space - Contents of a cacheable, mappable object.
  * @host: Owner, either the inode or the block_device.
diff --git a/mm/filemap.c b/mm/filemap.c
index 0751843b052f..c15cfc28f9ce 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3622,26 +3622,6 @@ struct page *read_cache_page_gfp(struct address_space *mapping,
 }
 EXPORT_SYMBOL(read_cache_page_gfp);
 
-int pagecache_write_begin(struct file *file, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned flags,
-				struct page **pagep, void **fsdata)
-{
-	const struct address_space_operations *aops = mapping->a_ops;
-
-	return aops->write_begin(file, mapping, pos, len, pagep, fsdata);
-}
-EXPORT_SYMBOL(pagecache_write_begin);
-
-int pagecache_write_end(struct file *file, struct address_space *mapping,
-				loff_t pos, unsigned len, unsigned copied,
-				struct page *page, void *fsdata)
-{
-	const struct address_space_operations *aops = mapping->a_ops;
-
-	return aops->write_end(file, mapping, pos, len, copied, page, fsdata);
-}
-EXPORT_SYMBOL(pagecache_write_end);
-
 /*
  * Warn about a page cache invalidation failure during a direct I/O write.
  */
-- 
2.34.1

[PATCH 0/3] Pagecache documentation updates

[Matthew Wilcox (Oracle)]

Hopefully the documentation updates here speak for themselves.

Matthew Wilcox (Oracle) (2):
  filemap: Update the folio_lock documentation
  filemap: Update the folio_mark_dirty documentation

Miaohe Lin (1):
  filemap: Remove obsolete comment in lock_page

 include/linux/pagemap.h | 60 ++++++++++++++++++++++++++++++++++++++---
 mm/page-writeback.c     | 10 ++++---
 2 files changed, 62 insertions(+), 8 deletions(-)

-- 
2.34.1

[PATCH 1/3] filemap: Remove obsolete comment in lock_page

[Matthew Wilcox (Oracle)]

From: Miaohe Lin <linmiaohe@huawei.com>

We no longer need the page's inode pinned.  This comment dates back to
commit db37648cd6ce ("[PATCH] mm: non syncing lock_page()") which added
lock_page_nosync().  That was removed by commit 7eaceaccab5f ("block:
remove per-queue plugging") which also made this comment obsolete.

Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 include/linux/pagemap.h | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index 65ae8f96554b..ab47579af434 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -908,9 +908,6 @@ static inline void folio_lock(struct folio *folio)
 		__folio_lock(folio);
 }
 
-/*
- * lock_page may only be called if we have the page's inode pinned.
- */
 static inline void lock_page(struct page *page)
 {
 	struct folio *folio;
-- 
2.34.1

Re: [PATCH 1/3] filemap: Remove obsolete comment in lock_page

[Miaohe Lin]

On 2022/5/9 4:30, Matthew Wilcox (Oracle) wrote:
> From: Miaohe Lin <linmiaohe@huawei.com>
> 
> We no longer need the page's inode pinned.  This comment dates back to
> commit db37648cd6ce ("[PATCH] mm: non syncing lock_page()") which added
> lock_page_nosync().  That was removed by commit 7eaceaccab5f ("block:
> remove per-queue plugging") which also made this comment obsolete.
> 
> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> Reviewed-by: Christoph Hellwig <hch@lst.de>

Many thanks for doing this! :)

> ---
>  include/linux/pagemap.h | 3 ---
>  1 file changed, 3 deletions(-)
> 
> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
> index 65ae8f96554b..ab47579af434 100644
> --- a/include/linux/pagemap.h
> +++ b/include/linux/pagemap.h
> @@ -908,9 +908,6 @@ static inline void folio_lock(struct folio *folio)
>  		__folio_lock(folio);
>  }
>  
> -/*
> - * lock_page may only be called if we have the page's inode pinned.
> - */
>  static inline void lock_page(struct page *page)
>  {
>  	struct folio *folio;
> 

[PATCH 2/3] filemap: Update the folio_lock documentation

[Matthew Wilcox (Oracle)]

Add kernel-doc for several functions relating to take the folio lock.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 include/linux/pagemap.h | 59 +++++++++++++++++++++++++++++++++++++++--
 1 file changed, 57 insertions(+), 2 deletions(-)

diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index ab47579af434..60657132080f 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -888,6 +888,18 @@ bool __folio_lock_or_retry(struct folio *folio, struct mm_struct *mm,
 void unlock_page(struct page *page);
 void folio_unlock(struct folio *folio);
 
+/**
+ * folio_trylock() - Attempt to lock a folio.
+ * @folio: The folio to attempt to lock.
+ *
+ * Sometimes it is undesirable to wait for a folio to be unlocked (eg
+ * when the locks are being taken in the wrong order, or if making
+ * progress through a batch of folios is more important than processing
+ * them in order).  Usually folio_lock() is the correct function to call.
+ *
+ * Context: Any context.
+ * Return: Whether the lock was successfully acquired.
+ */
 static inline bool folio_trylock(struct folio *folio)
 {
 	return likely(!test_and_set_bit_lock(PG_locked, folio_flags(folio, 0)));
@@ -901,6 +913,28 @@ static inline int trylock_page(struct page *page)
 	return folio_trylock(page_folio(page));
 }
 
+/**
+ * folio_lock() - Lock this folio.
+ * @folio: The folio to lock.
+ *
+ * The folio lock protects against many things, probably more than it
+ * should.  It is primarily held while a folio is being brought uptodate,
+ * either from its backing file or from swap.  It is also held while a
+ * folio is being truncated from its address_space, so holding the lock
+ * is sufficient to keep folio->mapping stable.
+ *
+ * The folio lock is also held while write() is modifying the page to
+ * provide POSIX atomicity guarantees (as long as the write does not
+ * cross a page boundary).  Other modifications to the data in the folio
+ * do not hold the folio lock and can race with writes, eg DMA and stores
+ * to mapped pages.
+ *
+ * Context: May sleep.  If you need to acquire the locks of two or
+ * more folios, they must be in order of ascending index, if they are
+ * in the same address_space.  If they are in different address_spaces,
+ * acquire the lock of the folio which belongs to the address_space which
+ * has the lowest address in memory first.
+ */
 static inline void folio_lock(struct folio *folio)
 {
 	might_sleep();
@@ -908,6 +942,17 @@ static inline void folio_lock(struct folio *folio)
 		__folio_lock(folio);
 }
 
+/**
+ * lock_page() - Lock the folio containing this page.
+ * @page: The page to lock.
+ *
+ * See folio_lock() for a description of what the lock protects.
+ * This is a legacy function and new code should probably use folio_lock()
+ * instead.
+ *
+ * Context: May sleep.  Pages in the same folio share a lock, so do not
+ * attempt to lock two pages which share a folio.
+ */
 static inline void lock_page(struct page *page)
 {
 	struct folio *folio;
@@ -918,6 +963,16 @@ static inline void lock_page(struct page *page)
 		__folio_lock(folio);
 }
 
+/**
+ * folio_lock_killable() - Lock this folio, interruptible by a fatal signal.
+ * @folio: The folio to lock.
+ *
+ * Attempts to lock the folio, like folio_lock(), except that the sleep
+ * to acquire the lock is interruptible by a fatal signal.
+ *
+ * Context: May sleep; see folio_lock().
+ * Return: 0 if the lock was acquired; -EINTR if a fatal signal was received.
+ */
 static inline int folio_lock_killable(struct folio *folio)
 {
 	might_sleep();
@@ -964,8 +1019,8 @@ int folio_wait_bit_killable(struct folio *folio, int bit_nr);
  * Wait for a folio to be unlocked.
  *
  * This must be called with the caller "holding" the folio,
- * ie with increased "page->count" so that the folio won't
- * go away during the wait..
+ * ie with increased folio reference count so that the folio won't
+ * go away during the wait.
  */
 static inline void folio_wait_locked(struct folio *folio)
 {
-- 
2.34.1

[PATCH 3/3] filemap: Update the folio_mark_dirty documentation

[Matthew Wilcox (Oracle)]

The previous comment was not terribly helpful.  Be a bit more explicit
about the necessary locking environment.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 mm/page-writeback.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/mm/page-writeback.c b/mm/page-writeback.c
index 7e2da284e427..fa1117db4610 100644
--- a/mm/page-writeback.c
+++ b/mm/page-writeback.c
@@ -2602,10 +2602,12 @@ EXPORT_SYMBOL(folio_redirty_for_writepage);
  * folio_mark_dirty - Mark a folio as being modified.
  * @folio: The folio.
  *
- * For folios with a mapping this should be done with the folio lock held
- * for the benefit of asynchronous memory errors who prefer a consistent
- * dirty state. This rule can be broken in some special cases,
- * but should be better not to.
+ * The folio may not be truncated while this function is running.
+ * Holding the folio lock is sufficient to prevent truncation, but some
+ * callers cannot acquire a sleeping lock.  These callers instead hold
+ * the page table lock for a page table which contains at least one page
+ * in this folio.  Truncation will block on the page table lock as it
+ * unmaps pages before removing the folio from its mapping.
  *
  * Return: True if the folio was newly dirtied, false if it was already dirty.
  */
-- 
2.34.1

[PATCH 00/37] Convert aops->read_page to aops->read_folio

[Matthew Wilcox (Oracle)]

Thanks to Christoph & Kees for pointing out that I'd inadvertently
broken CFI builds; I basically open-coded Kees' suggestion.

The swap_readpage commit is broken out to ease the merge.  Now that
Andrew's using git, I can look into basing on part of his tree, and
that way we can resolve the conflict before Linus sees it.

Matthew Wilcox (Oracle) (37):
  fs: Introduce aops->read_folio
  fs: Add read_folio documentation
  fs: Convert netfs_readpage to netfs_read_folio
  fs: Convert iomap_readpage to iomap_read_folio
  fs: Convert block_read_full_page() to block_read_full_folio()
  fs: Convert mpage_readpage to mpage_read_folio
  fs: Convert simple_readpage to simple_read_folio
  affs: Convert affs to read_folio
  afs: Convert afs_symlink_readpage to afs_symlink_read_folio
  befs: Convert befs to read_folio
  btrfs: Convert btrfs to read_folio
  cifs: Convert cifs to read_folio
  coda: Convert coda to read_folio
  cramfs: Convert cramfs to read_folio
  ecryptfs: Convert ecryptfs to read_folio
  efs: Convert efs symlinks to read_folio
  erofs: Convert erofs zdata to read_folio
  ext4: Convert ext4 to read_folio
  f2fs: Convert f2fs to read_folio
  freevxfs: Convert vxfs_immed to read_folio
  fuse: Convert fuse to read_folio
  hostfs: Convert hostfs to read_folio
  hpfs: Convert symlinks to read_folio
  isofs: Convert symlinks and zisofs to read_folio
  jffs2: Convert jffs2 to read_folio
  jfs: Convert metadata pages to read_folio
  nfs: Convert nfs to read_folio
  ntfs: Convert ntfs to read_folio
  ocfs2: Convert ocfs2 to read_folio
  orangefs: Convert orangefs to read_folio
  romfs: Convert romfs to read_folio
  squashfs: Convert squashfs to read_folio
  ubifs: Convert ubifs to read_folio
  udf: Convert adinicb and symlinks to read_folio
  vboxsf: Convert vboxsf to read_folio
  mm: Convert swap_readpage to call read_folio instead of readpage
  mm,fs: Remove aops->readpage

 Documentation/filesystems/fscrypt.rst       |  2 +-
 Documentation/filesystems/fsverity.rst      |  2 +-
 Documentation/filesystems/locking.rst       | 10 ++--
 Documentation/filesystems/netfs_library.rst |  8 +--
 Documentation/filesystems/vfs.rst           | 20 ++++----
 block/fops.c                                |  6 +--
 fs/9p/vfs_addr.c                            |  2 +-
 fs/adfs/inode.c                             |  6 +--
 fs/affs/file.c                              | 11 +++--
 fs/affs/symlink.c                           |  5 +-
 fs/afs/file.c                               | 17 +++----
 fs/befs/linuxvfs.c                          | 17 ++++---
 fs/bfs/file.c                               |  6 +--
 fs/btrfs/ctree.h                            |  2 +-
 fs/btrfs/file.c                             |  7 +--
 fs/btrfs/free-space-cache.c                 |  2 +-
 fs/btrfs/inode.c                            |  7 +--
 fs/btrfs/ioctl.c                            |  2 +-
 fs/btrfs/relocation.c                       |  8 +--
 fs/btrfs/send.c                             |  2 +-
 fs/buffer.c                                 | 55 +++++++++++----------
 fs/ceph/addr.c                              |  4 +-
 fs/cifs/file.c                              | 13 ++---
 fs/coda/symlink.c                           |  7 +--
 fs/cramfs/README                            |  8 +--
 fs/cramfs/inode.c                           |  7 +--
 fs/ecryptfs/mmap.c                          | 11 +++--
 fs/efs/inode.c                              |  8 +--
 fs/efs/symlink.c                            |  5 +-
 fs/erofs/data.c                             |  6 +--
 fs/erofs/zdata.c                            |  7 +--
 fs/exfat/inode.c                            |  6 +--
 fs/ext2/inode.c                             |  8 +--
 fs/ext4/inode.c                             |  9 ++--
 fs/ext4/move_extent.c                       |  4 +-
 fs/ext4/readpage.c                          |  4 +-
 fs/f2fs/data.c                              |  5 +-
 fs/fat/inode.c                              |  6 +--
 fs/freevxfs/vxfs_immed.c                    | 15 +++---
 fs/freevxfs/vxfs_subr.c                     | 17 +++----
 fs/fuse/dir.c                               | 10 ++--
 fs/fuse/file.c                              |  5 +-
 fs/gfs2/aops.c                              | 18 +++----
 fs/hfs/inode.c                              |  8 +--
 fs/hfsplus/inode.c                          |  8 +--
 fs/hostfs/hostfs_kern.c                     |  5 +-
 fs/hpfs/file.c                              |  6 +--
 fs/hpfs/namei.c                             |  5 +-
 fs/iomap/buffered-io.c                      | 16 +++---
 fs/isofs/compress.c                         |  5 +-
 fs/isofs/inode.c                            |  6 +--
 fs/isofs/rock.c                             |  7 +--
 fs/jffs2/file.c                             | 10 ++--
 fs/jffs2/fs.c                               |  2 +-
 fs/jfs/inode.c                              |  6 +--
 fs/jfs/jfs_metapage.c                       |  5 +-
 fs/libfs.c                                  | 14 +++---
 fs/minix/inode.c                            |  6 +--
 fs/mpage.c                                  | 18 ++++---
 fs/netfs/buffered_read.c                    | 15 +++---
 fs/nfs/file.c                               |  4 +-
 fs/nfs/read.c                               |  3 +-
 fs/nilfs2/inode.c                           | 10 ++--
 fs/ntfs/aops.c                              | 40 ++++++++-------
 fs/ntfs/aops.h                              |  6 +--
 fs/ntfs/attrib.c                            |  2 +-
 fs/ntfs/compress.c                          |  4 +-
 fs/ntfs/file.c                              |  4 +-
 fs/ntfs/inode.c                             |  4 +-
 fs/ntfs/mft.h                               |  2 +-
 fs/ntfs3/inode.c                            |  9 ++--
 fs/ocfs2/alloc.c                            |  2 +-
 fs/ocfs2/aops.c                             | 11 +++--
 fs/ocfs2/file.c                             |  2 +-
 fs/ocfs2/refcounttree.c                     |  6 ++-
 fs/ocfs2/symlink.c                          |  5 +-
 fs/omfs/file.c                              |  6 +--
 fs/orangefs/inode.c                         | 33 ++++++-------
 fs/qnx4/inode.c                             |  7 +--
 fs/qnx6/inode.c                             |  6 +--
 fs/reiserfs/file.c                          |  2 +-
 fs/reiserfs/inode.c                         | 12 ++---
 fs/romfs/super.c                            |  9 ++--
 fs/squashfs/file.c                          |  5 +-
 fs/squashfs/super.c                         |  2 +-
 fs/squashfs/symlink.c                       |  5 +-
 fs/sysv/itree.c                             |  6 +--
 fs/ubifs/file.c                             | 12 +++--
 fs/ubifs/super.c                            |  2 +-
 fs/udf/file.c                               | 10 ++--
 fs/udf/inode.c                              |  6 +--
 fs/udf/symlink.c                            |  5 +-
 fs/ufs/inode.c                              |  8 +--
 fs/vboxsf/file.c                            |  5 +-
 fs/xfs/xfs_aops.c                           |  8 +--
 fs/zonefs/super.c                           |  6 +--
 include/linux/buffer_head.h                 |  2 +-
 include/linux/fs.h                          |  4 +-
 include/linux/iomap.h                       |  2 +-
 include/linux/mpage.h                       |  2 +-
 include/linux/netfs.h                       |  2 +-
 include/linux/nfs_fs.h                      |  2 +-
 kernel/events/uprobes.c                     |  7 +--
 mm/filemap.c                                | 10 ++--
 mm/memory.c                                 |  4 +-
 mm/page_io.c                                |  2 +-
 mm/readahead.c                              | 16 +++---
 mm/shmem.c                                  |  2 +-
 mm/swapfile.c                               |  2 +-
 109 files changed, 444 insertions(+), 404 deletions(-)

-- 
2.34.1

[PATCH 01/37] fs: Introduce aops->read_folio

[Matthew Wilcox (Oracle)]

Change all the callers of ->readpage to call ->read_folio in preference,
if it exists.  This is a transitional duplication, and will be removed
by the end of the series.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/btrfs/file.c         |  2 +-
 fs/buffer.c             |  5 ++++-
 fs/ceph/addr.c          |  2 +-
 include/linux/fs.h      |  1 +
 kernel/events/uprobes.c |  6 ++++--
 mm/filemap.c            |  9 +++++++--
 mm/readahead.c          | 14 +++++++++-----
 mm/swapfile.c           |  2 +-
 8 files changed, 28 insertions(+), 13 deletions(-)

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 380054c94e4b..59510d7b1c65 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -2401,7 +2401,7 @@ static int btrfs_file_mmap(struct file	*filp, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = filp->f_mapping;
 
-	if (!mapping->a_ops->readpage)
+	if (!mapping->a_ops->readpage && !mapping->a_ops->read_folio)
 		return -ENOEXEC;
 
 	file_accessed(filp);
diff --git a/fs/buffer.c b/fs/buffer.c
index 9737e0dbe3ec..225d03cd622d 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2824,7 +2824,10 @@ int nobh_truncate_page(struct address_space *mapping,
 
 	/* Ok, it's mapped. Make sure it's up-to-date */
 	if (!folio_test_uptodate(folio)) {
-		err = mapping->a_ops->readpage(NULL, &folio->page);
+		if (mapping->a_ops->read_folio)
+			err = mapping->a_ops->read_folio(NULL, folio);
+		else
+			err = mapping->a_ops->readpage(NULL, &folio->page);
 		if (err) {
 			folio_put(folio);
 			goto out;
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index e65541a51b68..42bba2b5d98b 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1772,7 +1772,7 @@ int ceph_mmap(struct file *file, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = file->f_mapping;
 
-	if (!mapping->a_ops->readpage)
+	if (!mapping->a_ops->readpage && !mapping->a_ops->read_folio)
 		return -ENOEXEC;
 	file_accessed(file);
 	vma->vm_ops = &ceph_vmops;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 2be852661a29..5ad942183a2c 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -336,6 +336,7 @@ static inline bool is_sync_kiocb(struct kiocb *kiocb)
 struct address_space_operations {
 	int (*writepage)(struct page *page, struct writeback_control *wbc);
 	int (*readpage)(struct file *, struct page *);
+	int (*read_folio)(struct file *, struct folio *);
 
 	/* Write back some dirty pages from this mapping. */
 	int (*writepages)(struct address_space *, struct writeback_control *);
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 6418083901d4..a340b1043efa 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -790,7 +790,7 @@ static int __copy_insn(struct address_space *mapping, struct file *filp,
 	 * and in page-cache. If ->readpage == NULL it must be shmem_mapping(),
 	 * see uprobe_register().
 	 */
-	if (mapping->a_ops->readpage)
+	if (mapping->ops->read_folio || mapping->a_ops->readpage)
 		page = read_mapping_page(mapping, offset >> PAGE_SHIFT, filp);
 	else
 		page = shmem_read_mapping_page(mapping, offset >> PAGE_SHIFT);
@@ -1143,7 +1143,9 @@ static int __uprobe_register(struct inode *inode, loff_t offset,
 		return -EINVAL;
 
 	/* copy_insn() uses read_mapping_page() or shmem_read_mapping_page() */
-	if (!inode->i_mapping->a_ops->readpage && !shmem_mapping(inode->i_mapping))
+	if (!inode->i_mapping->a_ops->read_folio &&
+	    !inode->i_mapping->a_ops->readpage &&
+	    !shmem_mapping(inode->i_mapping))
 		return -EIO;
 	/* Racy, just to catch the obvious mistakes */
 	if (offset > i_size_read(inode))
diff --git a/mm/filemap.c b/mm/filemap.c
index c15cfc28f9ce..96e3d7ffd98e 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2419,7 +2419,10 @@ static int filemap_read_folio(struct file *file, struct address_space *mapping,
 	 */
 	folio_clear_error(folio);
 	/* Start the actual read. The read will unlock the page. */
-	error = mapping->a_ops->readpage(file, &folio->page);
+	if (mapping->a_ops->read_folio)
+		error = mapping->a_ops->read_folio(file, folio);
+	else
+		error = mapping->a_ops->readpage(file, &folio->page);
 	if (error)
 		return error;
 
@@ -3447,7 +3450,7 @@ int generic_file_mmap(struct file *file, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = file->f_mapping;
 
-	if (!mapping->a_ops->readpage)
+	if (!mapping->a_ops->read_folio && !mapping->a_ops->readpage)
 		return -ENOEXEC;
 	file_accessed(file);
 	vma->vm_ops = &generic_file_vm_ops;
@@ -3505,6 +3508,8 @@ static struct folio *do_read_cache_folio(struct address_space *mapping,
 filler:
 		if (filler)
 			err = filler(data, &folio->page);
+		else if (mapping->a_ops->read_folio)
+			err = mapping->a_ops->read_folio(data, folio);
 		else
 			err = mapping->a_ops->readpage(data, &folio->page);
 
diff --git a/mm/readahead.c b/mm/readahead.c
index 60a28af25c4e..76024c20a5a5 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -15,7 +15,7 @@
  * explicitly requested by the application.  Readahead only ever
  * attempts to read folios that are not yet in the page cache.  If a
  * folio is present but not up-to-date, readahead will not try to read
- * it. In that case a simple ->readpage() will be requested.
+ * it. In that case a simple ->read_folio() will be requested.
  *
  * Readahead is triggered when an application read request (whether a
  * system call or a page fault) finds that the requested folio is not in
@@ -78,7 +78,7 @@
  * address space operation, for which mpage_readahead() is a canonical
  * implementation.  ->readahead() should normally initiate reads on all
  * folios, but may fail to read any or all folios without causing an I/O
- * error.  The page cache reading code will issue a ->readpage() request
+ * error.  The page cache reading code will issue a ->read_folio() request
  * for any folio which ->readahead() did not read, and only an error
  * from this will be final.
  *
@@ -110,7 +110,7 @@
  * were not fetched with readahead_folio().  This will allow a
  * subsequent synchronous readahead request to try them again.  If they
  * are left in the page cache, then they will be read individually using
- * ->readpage() which may be less efficient.
+ * ->read_folio() which may be less efficient.
  */
 
 #include <linux/kernel.h>
@@ -170,8 +170,11 @@ static void read_pages(struct readahead_control *rac)
 			}
 			folio_unlock(folio);
 		}
+	} else if (aops->read_folio) {
+		while ((folio = readahead_folio(rac)) != NULL)
+			aops->read_folio(rac->file, folio);
 	} else {
-		while ((folio = readahead_folio(rac)))
+		while ((folio = readahead_folio(rac)) != NULL)
 			aops->readpage(rac->file, &folio->page);
 	}
 
@@ -302,7 +305,8 @@ void force_page_cache_ra(struct readahead_control *ractl,
 	struct backing_dev_info *bdi = inode_to_bdi(mapping->host);
 	unsigned long max_pages, index;
 
-	if (unlikely(!mapping->a_ops->readpage && !mapping->a_ops->readahead))
+	if (unlikely(!mapping->a_ops->read_folio &&
+		     !mapping->a_ops->readpage && !mapping->a_ops->readahead))
 		return;
 
 	/*
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 63c61f8b2611..7c19098b8b45 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -3041,7 +3041,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags)
 	/*
 	 * Read the swap header.
 	 */
-	if (!mapping->a_ops->readpage) {
+	if (!mapping->a_ops->read_folio && !mapping->a_ops->readpage) {
 		error = -EINVAL;
 		goto bad_swap_unlock_inode;
 	}
-- 
2.34.1

[PATCH 02/37] fs: Add read_folio documentation

[Matthew Wilcox (Oracle)]

Convert all the ->readpage documentation to ->read_folio.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 Documentation/filesystems/fscrypt.rst       |  2 +-
 Documentation/filesystems/fsverity.rst      |  2 +-
 Documentation/filesystems/locking.rst       | 10 +++++-----
 Documentation/filesystems/netfs_library.rst |  8 ++++----
 Documentation/filesystems/vfs.rst           | 20 ++++++++++----------
 5 files changed, 21 insertions(+), 21 deletions(-)

diff --git a/Documentation/filesystems/fscrypt.rst b/Documentation/filesystems/fscrypt.rst
index 6ccd5efb25b7..2e9aaa295125 100644
--- a/Documentation/filesystems/fscrypt.rst
+++ b/Documentation/filesystems/fscrypt.rst
@@ -1256,7 +1256,7 @@ inline encryption hardware will encrypt/decrypt the file contents.
 When inline encryption isn't used, filesystems must encrypt/decrypt
 the file contents themselves, as described below:
 
-For the read path (->readpage()) of regular files, filesystems can
+For the read path (->read_folio()) of regular files, filesystems can
 read the ciphertext into the page cache and decrypt it in-place.  The
 page lock must be held until decryption has finished, to prevent the
 page from becoming visible to userspace prematurely.
diff --git a/Documentation/filesystems/fsverity.rst b/Documentation/filesystems/fsverity.rst
index 8cc536d08f51..36290530e194 100644
--- a/Documentation/filesystems/fsverity.rst
+++ b/Documentation/filesystems/fsverity.rst
@@ -548,7 +548,7 @@ already verified).  Below, we describe how filesystems implement this.
 Pagecache
 ~~~~~~~~~
 
-For filesystems using Linux's pagecache, the ``->readpage()`` and
+For filesystems using Linux's pagecache, the ``->read_folio()`` and
 ``->readahead()`` methods must be modified to verify pages before they
 are marked Uptodate.  Merely hooking ``->read_iter()`` would be
 insufficient, since ``->read_iter()`` is not used for memory maps.
diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
index fd9d9caf09ab..aeba2475a53c 100644
--- a/Documentation/filesystems/locking.rst
+++ b/Documentation/filesystems/locking.rst
@@ -237,7 +237,7 @@ address_space_operations
 prototypes::
 
 	int (*writepage)(struct page *page, struct writeback_control *wbc);
-	int (*readpage)(struct file *, struct page *);
+	int (*read_folio)(struct file *, struct folio *);
 	int (*writepages)(struct address_space *, struct writeback_control *);
 	bool (*dirty_folio)(struct address_space *, struct folio *folio);
 	void (*readahead)(struct readahead_control *);
@@ -268,7 +268,7 @@ locking rules:
 ops			PageLocked(page)	 i_rwsem	invalidate_lock
 ======================	======================== =========	===============
 writepage:		yes, unlocks (see below)
-readpage:		yes, unlocks				shared
+read_folio:		yes, unlocks				shared
 writepages:
 dirty_folio		maybe
 readahead:		yes, unlocks				shared
@@ -289,13 +289,13 @@ swap_activate:		no
 swap_deactivate:	no
 ======================	======================== =========	===============
 
-->write_begin(), ->write_end() and ->readpage() may be called from
+->write_begin(), ->write_end() and ->read_folio() may be called from
 the request handler (/dev/loop).
 
-->readpage() unlocks the page, either synchronously or via I/O
+->read_folio() unlocks the folio, either synchronously or via I/O
 completion.
 
-->readahead() unlocks the pages that I/O is attempted on like ->readpage().
+->readahead() unlocks the folios that I/O is attempted on like ->read_folio().
 
 ->writepage() is used for two purposes: for "memory cleansing" and for
 "sync".  These are quite different operations and the behaviour may differ
diff --git a/Documentation/filesystems/netfs_library.rst b/Documentation/filesystems/netfs_library.rst
index d51c2a5ccf57..a80a59941d2f 100644
--- a/Documentation/filesystems/netfs_library.rst
+++ b/Documentation/filesystems/netfs_library.rst
@@ -96,7 +96,7 @@ attached to an inode (or NULL if fscache is disabled)::
 Buffered Read Helpers
 =====================
 
-The library provides a set of read helpers that handle the ->readpage(),
+The library provides a set of read helpers that handle the ->read_folio(),
 ->readahead() and much of the ->write_begin() VM operations and translate them
 into a common call framework.
 
@@ -136,8 +136,8 @@ Read Helper Functions
 Three read helpers are provided::
 
 	void netfs_readahead(struct readahead_control *ractl);
-	int netfs_readpage(struct file *file,
-			   struct page *page);
+	int netfs_read_folio(struct file *file,
+			   struct folio *folio);
 	int netfs_write_begin(struct file *file,
 			      struct address_space *mapping,
 			      loff_t pos,
@@ -148,7 +148,7 @@ Three read helpers are provided::
 Each corresponds to a VM address space operation.  These operations use the
 state in the per-inode context.
 
-For ->readahead() and ->readpage(), the network filesystem just point directly
+For ->readahead() and ->read_folio(), the network filesystem just point directly
 at the corresponding read helper; whereas for ->write_begin(), it may be a
 little more complicated as the network filesystem might want to flush
 conflicting writes or track dirty data and needs to put the acquired folio if
diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
index 469882f72fc1..0919a4ad973a 100644
--- a/Documentation/filesystems/vfs.rst
+++ b/Documentation/filesystems/vfs.rst
@@ -656,7 +656,7 @@ by memory-mapping the page.  Data is written into the address space by
 the application, and then written-back to storage typically in whole
 pages, however the address_space has finer control of write sizes.
 
-The read process essentially only requires 'readpage'.  The write
+The read process essentially only requires 'read_folio'.  The write
 process is more complicated and uses write_begin/write_end or
 dirty_folio to write data into the address_space, and writepage and
 writepages to writeback data to storage.
@@ -722,7 +722,7 @@ cache in your filesystem.  The following members are defined:
 
 	struct address_space_operations {
 		int (*writepage)(struct page *page, struct writeback_control *wbc);
-		int (*readpage)(struct file *, struct page *);
+		int (*read_folio)(struct file *, struct folio *);
 		int (*writepages)(struct address_space *, struct writeback_control *);
 		bool (*dirty_folio)(struct address_space *, struct folio *);
 		void (*readahead)(struct readahead_control *);
@@ -772,14 +772,14 @@ cache in your filesystem.  The following members are defined:
 
 	See the file "Locking" for more details.
 
-``readpage``
-	called by the VM to read a page from backing store.  The page
-	will be Locked when readpage is called, and should be unlocked
-	and marked uptodate once the read completes.  If ->readpage
-	discovers that it needs to unlock the page for some reason, it
-	can do so, and then return AOP_TRUNCATED_PAGE.  In this case,
-	the page will be relocated, relocked and if that all succeeds,
-	->readpage will be called again.
+``read_folio``
+	called by the VM to read a folio from backing store.  The folio
+	will be locked when read_folio is called, and should be unlocked
+	and marked uptodate once the read completes.  If ->read_folio
+	discovers that it cannot perform the I/O at this time, it can
+        unlock the folio and return AOP_TRUNCATED_PAGE.  In this case,
+	the folio will be looked up again, relocked and if that all succeeds,
+	->read_folio will be called again.
 
 ``writepages``
 	called by the VM to write out pages associated with the
-- 
2.34.1

[PATCH 03/37] fs: Convert netfs_readpage to netfs_read_folio

[Matthew Wilcox (Oracle)]

This is straightforward because netfs already worked in terms of folios.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/9p/vfs_addr.c         |  2 +-
 fs/afs/file.c            |  2 +-
 fs/ceph/addr.c           |  2 +-
 fs/netfs/buffered_read.c | 15 +++++++--------
 include/linux/netfs.h    |  2 +-
 5 files changed, 11 insertions(+), 12 deletions(-)

diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c
index a2d57112f53e..3a84167f4893 100644
--- a/fs/9p/vfs_addr.c
+++ b/fs/9p/vfs_addr.c
@@ -336,7 +336,7 @@ static bool v9fs_dirty_folio(struct address_space *mapping, struct folio *folio)
 #endif
 
 const struct address_space_operations v9fs_addr_operations = {
-	.readpage = netfs_readpage,
+	.read_folio = netfs_read_folio,
 	.readahead = netfs_readahead,
 	.dirty_folio = v9fs_dirty_folio,
 	.writepage = v9fs_vfs_writepage,
diff --git a/fs/afs/file.c b/fs/afs/file.c
index 26292a110a8f..e277fbe55262 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -50,7 +50,7 @@ const struct inode_operations afs_file_inode_operations = {
 };
 
 const struct address_space_operations afs_file_aops = {
-	.readpage	= netfs_readpage,
+	.read_folio	= netfs_read_folio,
 	.readahead	= netfs_readahead,
 	.dirty_folio	= afs_dirty_folio,
 	.launder_folio	= afs_launder_folio,
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index 42bba2b5d98b..be3e47784f08 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1372,7 +1372,7 @@ static int ceph_write_end(struct file *file, struct address_space *mapping,
 }
 
 const struct address_space_operations ceph_aops = {
-	.readpage = netfs_readpage,
+	.read_folio = netfs_read_folio,
 	.readahead = netfs_readahead,
 	.writepage = ceph_writepage,
 	.writepages = ceph_writepages_start,
diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index 1d44509455a5..8742d22dfd2b 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -198,22 +198,21 @@ void netfs_readahead(struct readahead_control *ractl)
 EXPORT_SYMBOL(netfs_readahead);
 
 /**
- * netfs_readpage - Helper to manage a readpage request
+ * netfs_read_folio - Helper to manage a read_folio request
  * @file: The file to read from
- * @subpage: A subpage of the folio to read
+ * @folio: The folio to read
  *
- * Fulfil a readpage request by drawing data from the cache if possible, or the
- * netfs if not.  Space beyond the EOF is zero-filled.  Multiple I/O requests
- * from different sources will get munged together.
+ * Fulfil a read_folio request by drawing data from the cache if
+ * possible, or the netfs if not.  Space beyond the EOF is zero-filled.
+ * Multiple I/O requests from different sources will get munged together.
  *
  * The calling netfs must initialise a netfs context contiguous to the vfs
  * inode before calling this.
  *
  * This is usable whether or not caching is enabled.
  */
-int netfs_readpage(struct file *file, struct page *subpage)
+int netfs_read_folio(struct file *file, struct folio *folio)
 {
-	struct folio *folio = page_folio(subpage);
 	struct address_space *mapping = folio_file_mapping(folio);
 	struct netfs_io_request *rreq;
 	struct netfs_i_context *ctx = netfs_i_context(mapping->host);
@@ -245,7 +244,7 @@ int netfs_readpage(struct file *file, struct page *subpage)
 	folio_unlock(folio);
 	return ret;
 }
-EXPORT_SYMBOL(netfs_readpage);
+EXPORT_SYMBOL(netfs_read_folio);
 
 /*
  * Prepare a folio for writing without reading first
diff --git a/include/linux/netfs.h b/include/linux/netfs.h
index 1c29f317d907..4bd5ee709daa 100644
--- a/include/linux/netfs.h
+++ b/include/linux/netfs.h
@@ -274,7 +274,7 @@ struct netfs_cache_ops {
 
 struct readahead_control;
 extern void netfs_readahead(struct readahead_control *);
-extern int netfs_readpage(struct file *, struct page *);
+int netfs_read_folio(struct file *, struct folio *);
 extern int netfs_write_begin(struct file *, struct address_space *,
 			     loff_t, unsigned int, struct folio **,
 			     void **);
-- 
2.34.1

[PATCH 04/37] fs: Convert iomap_readpage to iomap_read_folio

[Matthew Wilcox (Oracle)]

A straightforward conversion as iomap_readpage already worked in folios.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/erofs/data.c        |  6 +++---
 fs/gfs2/aops.c         |  3 ++-
 fs/iomap/buffered-io.c | 12 +++++-------
 fs/xfs/xfs_aops.c      |  8 ++++----
 fs/zonefs/super.c      |  6 +++---
 include/linux/iomap.h  |  2 +-
 6 files changed, 18 insertions(+), 19 deletions(-)

diff --git a/fs/erofs/data.c b/fs/erofs/data.c
index 780db1e5f4b7..2edca5669578 100644
--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -337,9 +337,9 @@ int erofs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
  * since we dont have write or truncate flows, so no inode
  * locking needs to be held at the moment.
  */
-static int erofs_readpage(struct file *file, struct page *page)
+static int erofs_read_folio(struct file *file, struct folio *folio)
 {
-	return iomap_readpage(page, &erofs_iomap_ops);
+	return iomap_read_folio(folio, &erofs_iomap_ops);
 }
 
 static void erofs_readahead(struct readahead_control *rac)
@@ -394,7 +394,7 @@ static ssize_t erofs_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
 
 /* for uncompressed (aligned) files and raw access for other files */
 const struct address_space_operations erofs_raw_access_aops = {
-	.readpage = erofs_readpage,
+	.read_folio = erofs_read_folio,
 	.readahead = erofs_readahead,
 	.bmap = erofs_bmap,
 	.direct_IO = noop_direct_IO,
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 72c9f31ce724..a29eb1e5bfe2 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -467,6 +467,7 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
 
 static int __gfs2_readpage(void *file, struct page *page)
 {
+	struct folio *folio = page_folio(page);
 	struct inode *inode = page->mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
@@ -474,7 +475,7 @@ static int __gfs2_readpage(void *file, struct page *page)
 
 	if (!gfs2_is_jdata(ip) ||
 	    (i_blocksize(inode) == PAGE_SIZE && !page_has_buffers(page))) {
-		error = iomap_readpage(page, &gfs2_iomap_ops);
+		error = iomap_read_folio(folio, &gfs2_iomap_ops);
 	} else if (gfs2_is_stuffed(ip)) {
 		error = stuffed_readpage(ip, page);
 		unlock_page(page);
diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 8ce8720093b9..72f63d719c7c 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -320,10 +320,8 @@ static loff_t iomap_readpage_iter(const struct iomap_iter *iter,
 	return pos - orig_pos + plen;
 }
 
-int
-iomap_readpage(struct page *page, const struct iomap_ops *ops)
+int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops)
 {
-	struct folio *folio = page_folio(page);
 	struct iomap_iter iter = {
 		.inode		= folio->mapping->host,
 		.pos		= folio_pos(folio),
@@ -352,12 +350,12 @@ iomap_readpage(struct page *page, const struct iomap_ops *ops)
 
 	/*
 	 * Just like mpage_readahead and block_read_full_page, we always
-	 * return 0 and just mark the page as PageError on errors.  This
+	 * return 0 and just set the folio error flag on errors.  This
 	 * should be cleaned up throughout the stack eventually.
 	 */
 	return 0;
 }
-EXPORT_SYMBOL_GPL(iomap_readpage);
+EXPORT_SYMBOL_GPL(iomap_read_folio);
 
 static loff_t iomap_readahead_iter(const struct iomap_iter *iter,
 		struct iomap_readpage_ctx *ctx)
@@ -663,10 +661,10 @@ static size_t __iomap_write_end(struct inode *inode, loff_t pos, size_t len,
 
 	/*
 	 * The blocks that were entirely written will now be uptodate, so we
-	 * don't have to worry about a readpage reading them and overwriting a
+	 * don't have to worry about a read_folio reading them and overwriting a
 	 * partial write.  However, if we've encountered a short write and only
 	 * partially written into a block, it will not be marked uptodate, so a
-	 * readpage might come in and destroy our partial write.
+	 * read_folio might come in and destroy our partial write.
 	 *
 	 * Do the simplest thing and just treat any short write to a
 	 * non-uptodate page as a zero-length write, and force the caller to
diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index 90b7f4d127de..a9c4bb500d53 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -538,11 +538,11 @@ xfs_vm_bmap(
 }
 
 STATIC int
-xfs_vm_readpage(
+xfs_vm_read_folio(
 	struct file		*unused,
-	struct page		*page)
+	struct folio		*folio)
 {
-	return iomap_readpage(page, &xfs_read_iomap_ops);
+	return iomap_read_folio(folio, &xfs_read_iomap_ops);
 }
 
 STATIC void
@@ -564,7 +564,7 @@ xfs_iomap_swapfile_activate(
 }
 
 const struct address_space_operations xfs_address_space_operations = {
-	.readpage		= xfs_vm_readpage,
+	.read_folio		= xfs_vm_read_folio,
 	.readahead		= xfs_vm_readahead,
 	.writepages		= xfs_vm_writepages,
 	.dirty_folio		= filemap_dirty_folio,
diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
index e20e7c841489..c3a38f711b24 100644
--- a/fs/zonefs/super.c
+++ b/fs/zonefs/super.c
@@ -124,9 +124,9 @@ static const struct iomap_ops zonefs_iomap_ops = {
 	.iomap_begin	= zonefs_iomap_begin,
 };
 
-static int zonefs_readpage(struct file *unused, struct page *page)
+static int zonefs_read_folio(struct file *unused, struct folio *folio)
 {
-	return iomap_readpage(page, &zonefs_iomap_ops);
+	return iomap_read_folio(folio, &zonefs_iomap_ops);
 }
 
 static void zonefs_readahead(struct readahead_control *rac)
@@ -192,7 +192,7 @@ static int zonefs_swap_activate(struct swap_info_struct *sis,
 }
 
 static const struct address_space_operations zonefs_file_aops = {
-	.readpage		= zonefs_readpage,
+	.read_folio		= zonefs_read_folio,
 	.readahead		= zonefs_readahead,
 	.writepage		= zonefs_writepage,
 	.writepages		= zonefs_writepages,
diff --git a/include/linux/iomap.h b/include/linux/iomap.h
index b76f0dd149fb..5b2aa45ddda3 100644
--- a/include/linux/iomap.h
+++ b/include/linux/iomap.h
@@ -225,7 +225,7 @@ static inline const struct iomap *iomap_iter_srcmap(const struct iomap_iter *i)
 
 ssize_t iomap_file_buffered_write(struct kiocb *iocb, struct iov_iter *from,
 		const struct iomap_ops *ops);
-int iomap_readpage(struct page *page, const struct iomap_ops *ops);
+int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops);
 void iomap_readahead(struct readahead_control *, const struct iomap_ops *ops);
 bool iomap_is_partially_uptodate(struct folio *, size_t from, size_t count);
 int iomap_releasepage(struct page *page, gfp_t gfp_mask);
-- 
2.34.1

[PATCH 05/37] fs: Convert block_read_full_page() to block_read_full_folio()

[Matthew Wilcox (Oracle)]

This function is NOT converted to handle large folios, so include
an assert that the filesystem isn't passing one in.  Otherwise, use
the folio functions instead of the page functions, where they exist.
Convert all filesystems which use block_read_full_page().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 block/fops.c                |  6 ++---
 fs/adfs/inode.c             |  6 ++---
 fs/affs/file.c              |  6 ++---
 fs/befs/linuxvfs.c          | 10 +++----
 fs/bfs/file.c               |  6 ++---
 fs/buffer.c                 | 53 ++++++++++++++++++++-----------------
 fs/efs/inode.c              |  8 +++---
 fs/ext4/readpage.c          |  4 +--
 fs/freevxfs/vxfs_subr.c     | 17 ++++++------
 fs/hfs/inode.c              |  8 +++---
 fs/hfsplus/inode.c          |  8 +++---
 fs/iomap/buffered-io.c      |  2 +-
 fs/minix/inode.c            |  6 ++---
 fs/mpage.c                  | 10 +++----
 fs/ntfs/compress.c          |  4 +--
 fs/ocfs2/aops.c             |  6 ++---
 fs/ocfs2/refcounttree.c     |  6 +++--
 fs/omfs/file.c              |  6 ++---
 fs/qnx4/inode.c             |  7 ++---
 fs/reiserfs/file.c          |  2 +-
 fs/reiserfs/inode.c         | 12 ++++-----
 fs/sysv/itree.c             |  6 ++---
 fs/ufs/inode.c              |  8 +++---
 include/linux/buffer_head.h |  2 +-
 24 files changed, 108 insertions(+), 101 deletions(-)

diff --git a/block/fops.c b/block/fops.c
index 712affe56e29..06feb41d798b 100644
--- a/block/fops.c
+++ b/block/fops.c
@@ -387,9 +387,9 @@ static int blkdev_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, blkdev_get_block, wbc);
 }
 
-static int blkdev_readpage(struct file * file, struct page * page)
+static int blkdev_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, blkdev_get_block);
+	return block_read_full_folio(folio, blkdev_get_block);
 }
 
 static void blkdev_readahead(struct readahead_control *rac)
@@ -425,7 +425,7 @@ static int blkdev_writepages(struct address_space *mapping,
 const struct address_space_operations def_blk_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= blkdev_readpage,
+	.read_folio	= blkdev_read_folio,
 	.readahead	= blkdev_readahead,
 	.writepage	= blkdev_writepage,
 	.write_begin	= blkdev_write_begin,
diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c
index f7959b1a2d52..ee22278b0cfc 100644
--- a/fs/adfs/inode.c
+++ b/fs/adfs/inode.c
@@ -38,9 +38,9 @@ static int adfs_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, adfs_get_block, wbc);
 }
 
-static int adfs_readpage(struct file *file, struct page *page)
+static int adfs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, adfs_get_block);
+	return block_read_full_folio(folio, adfs_get_block);
 }
 
 static void adfs_write_failed(struct address_space *mapping, loff_t to)
@@ -75,7 +75,7 @@ static sector_t _adfs_bmap(struct address_space *mapping, sector_t block)
 static const struct address_space_operations adfs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= adfs_readpage,
+	.read_folio	= adfs_read_folio,
 	.writepage	= adfs_writepage,
 	.write_begin	= adfs_write_begin,
 	.write_end	= generic_write_end,
diff --git a/fs/affs/file.c b/fs/affs/file.c
index b952f65c3f06..5da562cc7fb7 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -375,9 +375,9 @@ static int affs_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, affs_get_block, wbc);
 }
 
-static int affs_readpage(struct file *file, struct page *page)
+static int affs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, affs_get_block);
+	return block_read_full_folio(folio, affs_get_block);
 }
 
 static void affs_write_failed(struct address_space *mapping, loff_t to)
@@ -455,7 +455,7 @@ static sector_t _affs_bmap(struct address_space *mapping, sector_t block)
 const struct address_space_operations affs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = affs_readpage,
+	.read_folio = affs_read_folio,
 	.writepage = affs_writepage,
 	.write_begin = affs_write_begin,
 	.write_end = affs_write_end,
diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c
index b4b3567ac655..25350dd22cda 100644
--- a/fs/befs/linuxvfs.c
+++ b/fs/befs/linuxvfs.c
@@ -40,7 +40,7 @@ MODULE_LICENSE("GPL");
 
 static int befs_readdir(struct file *, struct dir_context *);
 static int befs_get_block(struct inode *, sector_t, struct buffer_head *, int);
-static int befs_readpage(struct file *file, struct page *page);
+static int befs_read_folio(struct file *file, struct folio *folio);
 static sector_t befs_bmap(struct address_space *mapping, sector_t block);
 static struct dentry *befs_lookup(struct inode *, struct dentry *,
 				  unsigned int);
@@ -87,7 +87,7 @@ static const struct inode_operations befs_dir_inode_operations = {
 };
 
 static const struct address_space_operations befs_aops = {
-	.readpage	= befs_readpage,
+	.read_folio	= befs_read_folio,
 	.bmap		= befs_bmap,
 };
 
@@ -102,16 +102,16 @@ static const struct export_operations befs_export_operations = {
 };
 
 /*
- * Called by generic_file_read() to read a page of data
+ * Called by generic_file_read() to read a folio of data
  *
  * In turn, simply calls a generic block read function and
  * passes it the address of befs_get_block, for mapping file
  * positions to disk blocks.
  */
 static int
-befs_readpage(struct file *file, struct page *page)
+befs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, befs_get_block);
+	return block_read_full_folio(folio, befs_get_block);
 }
 
 static sector_t
diff --git a/fs/bfs/file.c b/fs/bfs/file.c
index dc97c9b8f23b..57ae5ee6deec 100644
--- a/fs/bfs/file.c
+++ b/fs/bfs/file.c
@@ -155,9 +155,9 @@ static int bfs_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, bfs_get_block, wbc);
 }
 
-static int bfs_readpage(struct file *file, struct page *page)
+static int bfs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, bfs_get_block);
+	return block_read_full_folio(folio, bfs_get_block);
 }
 
 static void bfs_write_failed(struct address_space *mapping, loff_t to)
@@ -189,7 +189,7 @@ static sector_t bfs_bmap(struct address_space *mapping, sector_t block)
 const struct address_space_operations bfs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= bfs_readpage,
+	.read_folio	= bfs_read_folio,
 	.writepage	= bfs_writepage,
 	.write_begin	= bfs_write_begin,
 	.write_end	= generic_write_end,
diff --git a/fs/buffer.c b/fs/buffer.c
index 225d03cd622d..ec0c52c8848e 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -314,7 +314,7 @@ static void decrypt_bh(struct work_struct *work)
 }
 
 /*
- * I/O completion handler for block_read_full_page() - pages
+ * I/O completion handler for block_read_full_folio() - pages
  * which come unlocked at the end of I/O.
  */
 static void end_buffer_async_read_io(struct buffer_head *bh, int uptodate)
@@ -1060,8 +1060,8 @@ __getblk_slow(struct block_device *bdev, sector_t block,
  * Also.  When blockdev buffers are explicitly read with bread(), they
  * individually become uptodate.  But their backing page remains not
  * uptodate - even if all of its buffers are uptodate.  A subsequent
- * block_read_full_page() against that page will discover all the uptodate
- * buffers, will set the page uptodate and will perform no I/O.
+ * block_read_full_folio() against that folio will discover all the uptodate
+ * buffers, will set the folio uptodate and will perform no I/O.
  */
 
 /**
@@ -2088,7 +2088,7 @@ static int __block_commit_write(struct inode *inode, struct page *page,
 
 	/*
 	 * If this is a partial write which happened to make all buffers
-	 * uptodate then we can optimize away a bogus readpage() for
+	 * uptodate then we can optimize away a bogus read_folio() for
 	 * the next read(). Here we 'discover' whether the page went
 	 * uptodate as a result of this (potentially partial) write.
 	 */
@@ -2137,12 +2137,12 @@ int block_write_end(struct file *file, struct address_space *mapping,
 
 	if (unlikely(copied < len)) {
 		/*
-		 * The buffers that were written will now be uptodate, so we
-		 * don't have to worry about a readpage reading them and
-		 * overwriting a partial write. However if we have encountered
-		 * a short write and only partially written into a buffer, it
-		 * will not be marked uptodate, so a readpage might come in and
-		 * destroy our partial write.
+		 * The buffers that were written will now be uptodate, so
+		 * we don't have to worry about a read_folio reading them
+		 * and overwriting a partial write. However if we have
+		 * encountered a short write and only partially written
+		 * into a buffer, it will not be marked uptodate, so a
+		 * read_folio might come in and destroy our partial write.
 		 *
 		 * Do the simplest thing, and just treat any short write to a
 		 * non uptodate page as a zero-length write, and force the
@@ -2245,26 +2245,28 @@ bool block_is_partially_uptodate(struct folio *folio, size_t from, size_t count)
 EXPORT_SYMBOL(block_is_partially_uptodate);
 
 /*
- * Generic "read page" function for block devices that have the normal
+ * Generic "read_folio" function for block devices that have the normal
  * get_block functionality. This is most of the block device filesystems.
- * Reads the page asynchronously --- the unlock_buffer() and
+ * Reads the folio asynchronously --- the unlock_buffer() and
  * set/clear_buffer_uptodate() functions propagate buffer state into the
- * page struct once IO has completed.
+ * folio once IO has completed.
  */
-int block_read_full_page(struct page *page, get_block_t *get_block)
+int block_read_full_folio(struct folio *folio, get_block_t *get_block)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	sector_t iblock, lblock;
 	struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE];
 	unsigned int blocksize, bbits;
 	int nr, i;
 	int fully_mapped = 1;
 
-	head = create_page_buffers(page, inode, 0);
+	VM_BUG_ON_FOLIO(folio_test_large(folio), folio);
+
+	head = create_page_buffers(&folio->page, inode, 0);
 	blocksize = head->b_size;
 	bbits = block_size_bits(blocksize);
 
-	iblock = (sector_t)page->index << (PAGE_SHIFT - bbits);
+	iblock = (sector_t)folio->index << (PAGE_SHIFT - bbits);
 	lblock = (i_size_read(inode)+blocksize-1) >> bbits;
 	bh = head;
 	nr = 0;
@@ -2282,10 +2284,11 @@ int block_read_full_page(struct page *page, get_block_t *get_block)
 				WARN_ON(bh->b_size != blocksize);
 				err = get_block(inode, iblock, bh, 0);
 				if (err)
-					SetPageError(page);
+					folio_set_error(folio);
 			}
 			if (!buffer_mapped(bh)) {
-				zero_user(page, i * blocksize, blocksize);
+				folio_zero_range(folio, i * blocksize,
+						blocksize);
 				if (!err)
 					set_buffer_uptodate(bh);
 				continue;
@@ -2301,16 +2304,16 @@ int block_read_full_page(struct page *page, get_block_t *get_block)
 	} while (i++, iblock++, (bh = bh->b_this_page) != head);
 
 	if (fully_mapped)
-		SetPageMappedToDisk(page);
+		folio_set_mappedtodisk(folio);
 
 	if (!nr) {
 		/*
-		 * All buffers are uptodate - we can set the page uptodate
+		 * All buffers are uptodate - we can set the folio uptodate
 		 * as well. But not if get_block() returned an error.
 		 */
-		if (!PageError(page))
-			SetPageUptodate(page);
-		unlock_page(page);
+		if (!folio_test_error(folio))
+			folio_mark_uptodate(folio);
+		folio_unlock(folio);
 		return 0;
 	}
 
@@ -2335,7 +2338,7 @@ int block_read_full_page(struct page *page, get_block_t *get_block)
 	}
 	return 0;
 }
-EXPORT_SYMBOL(block_read_full_page);
+EXPORT_SYMBOL(block_read_full_folio);
 
 /* utility function for filesystems that need to do work on expanding
  * truncates.  Uses filesystem pagecache writes to allow the filesystem to
diff --git a/fs/efs/inode.c b/fs/efs/inode.c
index 89e73a6f0d36..3ba94bb005a6 100644
--- a/fs/efs/inode.c
+++ b/fs/efs/inode.c
@@ -14,16 +14,18 @@
 #include "efs.h"
 #include <linux/efs_fs_sb.h>
 
-static int efs_readpage(struct file *file, struct page *page)
+static int efs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page,efs_get_block);
+	return block_read_full_folio(folio, efs_get_block);
 }
+
 static sector_t _efs_bmap(struct address_space *mapping, sector_t block)
 {
 	return generic_block_bmap(mapping,block,efs_get_block);
 }
+
 static const struct address_space_operations efs_aops = {
-	.readpage = efs_readpage,
+	.read_folio = efs_read_folio,
 	.bmap = _efs_bmap
 };
 
diff --git a/fs/ext4/readpage.c b/fs/ext4/readpage.c
index af491e170c4a..e02a5f14e021 100644
--- a/fs/ext4/readpage.c
+++ b/fs/ext4/readpage.c
@@ -163,7 +163,7 @@ static bool bio_post_read_required(struct bio *bio)
  *
  * The mpage code never puts partial pages into a BIO (except for end-of-file).
  * If a page does not map to a contiguous run of blocks then it simply falls
- * back to block_read_full_page().
+ * back to block_read_full_folio().
  *
  * Why is this?  If a page's completion depends on a number of different BIOs
  * which can complete in any order (or at the same time) then determining the
@@ -394,7 +394,7 @@ int ext4_mpage_readpages(struct inode *inode,
 			bio = NULL;
 		}
 		if (!PageUptodate(page))
-			block_read_full_page(page, ext4_get_block);
+			block_read_full_folio(page_folio(page), ext4_get_block);
 		else
 			unlock_page(page);
 	next_page:
diff --git a/fs/freevxfs/vxfs_subr.c b/fs/freevxfs/vxfs_subr.c
index e806694d4145..6143ebab940d 100644
--- a/fs/freevxfs/vxfs_subr.c
+++ b/fs/freevxfs/vxfs_subr.c
@@ -38,11 +38,11 @@
 #include "vxfs_extern.h"
 
 
-static int		vxfs_readpage(struct file *, struct page *);
+static int		vxfs_read_folio(struct file *, struct folio *);
 static sector_t		vxfs_bmap(struct address_space *, sector_t);
 
 const struct address_space_operations vxfs_aops = {
-	.readpage =		vxfs_readpage,
+	.read_folio =		vxfs_read_folio,
 	.bmap =			vxfs_bmap,
 };
 
@@ -141,24 +141,23 @@ vxfs_getblk(struct inode *ip, sector_t iblock,
 }
 
 /**
- * vxfs_readpage - read one page synchronously into the pagecache
+ * vxfs_read_folio - read one page synchronously into the pagecache
  * @file:	file context (unused)
- * @page:	page frame to fill in.
+ * @folio:	folio to fill in.
  *
  * Description:
- *   The vxfs_readpage routine reads @page synchronously into the
+ *   The vxfs_read_folio routine reads @folio synchronously into the
  *   pagecache.
  *
  * Returns:
  *   Zero on success, else a negative error code.
  *
  * Locking status:
- *   @page is locked and will be unlocked.
+ *   @folio is locked and will be unlocked.
  */
-static int
-vxfs_readpage(struct file *file, struct page *page)
+static int vxfs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, vxfs_getblk);
+	return block_read_full_folio(folio, vxfs_getblk);
 }
  
 /**
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 9a26b9510da0..ba3ff9cd7cfc 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -34,9 +34,9 @@ static int hfs_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, hfs_get_block, wbc);
 }
 
-static int hfs_readpage(struct file *file, struct page *page)
+static int hfs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, hfs_get_block);
+	return block_read_full_folio(folio, hfs_get_block);
 }
 
 static void hfs_write_failed(struct address_space *mapping, loff_t to)
@@ -160,7 +160,7 @@ static int hfs_writepages(struct address_space *mapping,
 const struct address_space_operations hfs_btree_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= hfs_readpage,
+	.read_folio	= hfs_read_folio,
 	.writepage	= hfs_writepage,
 	.write_begin	= hfs_write_begin,
 	.write_end	= generic_write_end,
@@ -171,7 +171,7 @@ const struct address_space_operations hfs_btree_aops = {
 const struct address_space_operations hfs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= hfs_readpage,
+	.read_folio	= hfs_read_folio,
 	.writepage	= hfs_writepage,
 	.write_begin	= hfs_write_begin,
 	.write_end	= generic_write_end,
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 905ae3660315..982b34eefec7 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -23,9 +23,9 @@
 #include "hfsplus_raw.h"
 #include "xattr.h"
 
-static int hfsplus_readpage(struct file *file, struct page *page)
+static int hfsplus_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, hfsplus_get_block);
+	return block_read_full_folio(folio, hfsplus_get_block);
 }
 
 static int hfsplus_writepage(struct page *page, struct writeback_control *wbc)
@@ -157,7 +157,7 @@ static int hfsplus_writepages(struct address_space *mapping,
 const struct address_space_operations hfsplus_btree_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= hfsplus_readpage,
+	.read_folio	= hfsplus_read_folio,
 	.writepage	= hfsplus_writepage,
 	.write_begin	= hfsplus_write_begin,
 	.write_end	= generic_write_end,
@@ -168,7 +168,7 @@ const struct address_space_operations hfsplus_btree_aops = {
 const struct address_space_operations hfsplus_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= hfsplus_readpage,
+	.read_folio	= hfsplus_read_folio,
 	.writepage	= hfsplus_writepage,
 	.write_begin	= hfsplus_write_begin,
 	.write_end	= generic_write_end,
diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 72f63d719c7c..75eb0c27a0e8 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -349,7 +349,7 @@ int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops)
 	}
 
 	/*
-	 * Just like mpage_readahead and block_read_full_page, we always
+	 * Just like mpage_readahead and block_read_full_folio, we always
 	 * return 0 and just set the folio error flag on errors.  This
 	 * should be cleaned up throughout the stack eventually.
 	 */
diff --git a/fs/minix/inode.c b/fs/minix/inode.c
index 3add78bccedc..da8bdd1712a7 100644
--- a/fs/minix/inode.c
+++ b/fs/minix/inode.c
@@ -402,9 +402,9 @@ static int minix_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, minix_get_block, wbc);
 }
 
-static int minix_readpage(struct file *file, struct page *page)
+static int minix_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page,minix_get_block);
+	return block_read_full_folio(folio, minix_get_block);
 }
 
 int minix_prepare_chunk(struct page *page, loff_t pos, unsigned len)
@@ -443,7 +443,7 @@ static sector_t minix_bmap(struct address_space *mapping, sector_t block)
 static const struct address_space_operations minix_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = minix_readpage,
+	.read_folio = minix_read_folio,
 	.writepage = minix_writepage,
 	.write_begin = minix_write_begin,
 	.write_end = generic_write_end,
diff --git a/fs/mpage.c b/fs/mpage.c
index 1fe56f8c495f..a04439b84ae2 100644
--- a/fs/mpage.c
+++ b/fs/mpage.c
@@ -36,7 +36,7 @@
  *
  * The mpage code never puts partial pages into a BIO (except for end-of-file).
  * If a page does not map to a contiguous run of blocks then it simply falls
- * back to block_read_full_page().
+ * back to block_read_full_folio().
  *
  * Why is this?  If a page's completion depends on a number of different BIOs
  * which can complete in any order (or at the same time) then determining the
@@ -68,7 +68,7 @@ static struct bio *mpage_bio_submit(struct bio *bio)
 /*
  * support function for mpage_readahead.  The fs supplied get_block might
  * return an up to date buffer.  This is used to map that buffer into
- * the page, which allows readpage to avoid triggering a duplicate call
+ * the page, which allows read_folio to avoid triggering a duplicate call
  * to get_block.
  *
  * The idea is to avoid adding buffers to pages that don't already have
@@ -296,7 +296,7 @@ static struct bio *do_mpage_readpage(struct mpage_readpage_args *args)
 	if (args->bio)
 		args->bio = mpage_bio_submit(args->bio);
 	if (!PageUptodate(page))
-		block_read_full_page(page, args->get_block);
+		block_read_full_folio(page_folio(page), args->get_block);
 	else
 		unlock_page(page);
 	goto out;
@@ -425,7 +425,7 @@ static void clean_buffers(struct page *page, unsigned first_unmapped)
 
 	/*
 	 * we cannot drop the bh if the page is not uptodate or a concurrent
-	 * readpage would fail to serialize with the bh and it would read from
+	 * read_folio would fail to serialize with the bh and it would read from
 	 * disk before we reach the platter.
 	 */
 	if (buffer_heads_over_limit && PageUptodate(page))
@@ -510,7 +510,7 @@ static int __mpage_writepage(struct page *page, struct writeback_control *wbc,
 		/*
 		 * Page has buffers, but they are all unmapped. The page was
 		 * created by pagein or read over a hole which was handled by
-		 * block_read_full_page().  If this address_space is also
+		 * block_read_full_folio().  If this address_space is also
 		 * using mpage_readahead then this can rarely happen.
 		 */
 		goto confused;
diff --git a/fs/ntfs/compress.c b/fs/ntfs/compress.c
index d2f9d6a0ee32..a60f543e7557 100644
--- a/fs/ntfs/compress.c
+++ b/fs/ntfs/compress.c
@@ -780,12 +780,12 @@ int ntfs_read_compressed_block(struct page *page)
 		/* Uncompressed cb, copy it to the destination pages. */
 		/*
 		 * TODO: As a big optimization, we could detect this case
-		 * before we read all the pages and use block_read_full_page()
+		 * before we read all the pages and use block_read_full_folio()
 		 * on all full pages instead (we still have to treat partial
 		 * pages especially but at least we are getting rid of the
 		 * synchronous io for the majority of pages.
 		 * Or if we choose not to do the read-ahead/-behind stuff, we
-		 * could just return block_read_full_page(pages[xpage]) as long
+		 * could just return block_read_full_folio(pages[xpage]) as long
 		 * as PAGE_SIZE <= cb_size.
 		 */
 		if (cb_max_ofs)
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 7cffe9dcad17..7bf4b6fd93bf 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -309,7 +309,7 @@ static int ocfs2_readpage(struct file *file, struct page *page)
 	/*
 	 * i_size might have just been updated as we grabed the meta lock.  We
 	 * might now be discovering a truncate that hit on another node.
-	 * block_read_full_page->get_block freaks out if it is asked to read
+	 * block_read_full_folio->get_block freaks out if it is asked to read
 	 * beyond the end of a file, so we check here.  Callers
 	 * (generic_file_read, vm_ops->fault) are clever enough to check i_size
 	 * and notice that the page they just read isn't needed.
@@ -326,7 +326,7 @@ static int ocfs2_readpage(struct file *file, struct page *page)
 	if (oi->ip_dyn_features & OCFS2_INLINE_DATA_FL)
 		ret = ocfs2_readpage_inline(inode, page);
 	else
-		ret = block_read_full_page(page, ocfs2_get_block);
+		ret = block_read_full_folio(page_folio(page), ocfs2_get_block);
 	unlock = 0;
 
 out_alloc:
@@ -1897,7 +1897,7 @@ static int ocfs2_write_begin(struct file *file, struct address_space *mapping,
 	/*
 	 * Take alloc sem here to prevent concurrent lookups. That way
 	 * the mapping, zeroing and tree manipulation within
-	 * ocfs2_write() will be safe against ->readpage(). This
+	 * ocfs2_write() will be safe against ->read_folio(). This
 	 * should also serve to lock out allocation from a shared
 	 * writeable region.
 	 */
diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c
index 7f6355cbb587..e04358a46b68 100644
--- a/fs/ocfs2/refcounttree.c
+++ b/fs/ocfs2/refcounttree.c
@@ -2961,12 +2961,14 @@ int ocfs2_duplicate_clusters_by_page(handle_t *handle,
 		}
 
 		if (!PageUptodate(page)) {
-			ret = block_read_full_page(page, ocfs2_get_block);
+			struct folio *folio = page_folio(page);
+
+			ret = block_read_full_folio(folio, ocfs2_get_block);
 			if (ret) {
 				mlog_errno(ret);
 				goto unlock;
 			}
-			lock_page(page);
+			folio_lock(folio);
 		}
 
 		if (page_has_buffers(page)) {
diff --git a/fs/omfs/file.c b/fs/omfs/file.c
index 980b0a72c172..fa7fe2393ff6 100644
--- a/fs/omfs/file.c
+++ b/fs/omfs/file.c
@@ -284,9 +284,9 @@ static int omfs_get_block(struct inode *inode, sector_t block,
 	return ret;
 }
 
-static int omfs_readpage(struct file *file, struct page *page)
+static int omfs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page, omfs_get_block);
+	return block_read_full_folio(folio, omfs_get_block);
 }
 
 static void omfs_readahead(struct readahead_control *rac)
@@ -373,7 +373,7 @@ const struct inode_operations omfs_file_inops = {
 const struct address_space_operations omfs_aops = {
 	.dirty_folio = block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = omfs_readpage,
+	.read_folio = omfs_read_folio,
 	.readahead = omfs_readahead,
 	.writepage = omfs_writepage,
 	.writepages = omfs_writepages,
diff --git a/fs/qnx4/inode.c b/fs/qnx4/inode.c
index a635bb6615e9..391ea402920d 100644
--- a/fs/qnx4/inode.c
+++ b/fs/qnx4/inode.c
@@ -245,17 +245,18 @@ static void qnx4_kill_sb(struct super_block *sb)
 	}
 }
 
-static int qnx4_readpage(struct file *file, struct page *page)
+static int qnx4_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page,qnx4_get_block);
+	return block_read_full_folio(folio, qnx4_get_block);
 }
 
 static sector_t qnx4_bmap(struct address_space *mapping, sector_t block)
 {
 	return generic_block_bmap(mapping,block,qnx4_get_block);
 }
+
 static const struct address_space_operations qnx4_aops = {
-	.readpage	= qnx4_readpage,
+	.read_folio	= qnx4_read_folio,
 	.bmap		= qnx4_bmap
 };
 
diff --git a/fs/reiserfs/file.c b/fs/reiserfs/file.c
index 203a47232707..6e228bfbe7ef 100644
--- a/fs/reiserfs/file.c
+++ b/fs/reiserfs/file.c
@@ -227,7 +227,7 @@ int reiserfs_commit_page(struct inode *inode, struct page *page,
 	}
 	/*
 	 * If this is a partial write which happened to make all buffers
-	 * uptodate then we can optimize away a bogus readpage() for
+	 * uptodate then we can optimize away a bogus read_folio() for
 	 * the next read(). Here we 'discover' whether the page went
 	 * uptodate as a result of this (potentially partial) write.
 	 */
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 46ba4892030a..33a9555f77b9 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -167,10 +167,10 @@ inline void make_le_item_head(struct item_head *ih, const struct cpu_key *key,
  * cutting the code is fine, since it really isn't in use yet and is easy
  * to add back in.  But, Vladimir has a really good idea here.  Think
  * about what happens for reading a file.  For each page,
- * The VFS layer calls reiserfs_readpage, who searches the tree to find
+ * The VFS layer calls reiserfs_read_folio, who searches the tree to find
  * an indirect item.  This indirect item has X number of pointers, where
  * X is a big number if we've done the block allocation right.  But,
- * we only use one or two of these pointers during each call to readpage,
+ * we only use one or two of these pointers during each call to read_folio,
  * needlessly researching again later on.
  *
  * The size of the cache could be dynamic based on the size of the file.
@@ -966,7 +966,7 @@ int reiserfs_get_block(struct inode *inode, sector_t block,
 			 * it is important the set_buffer_uptodate is done
 			 * after the direct2indirect.  The buffer might
 			 * contain valid data newer than the data on disk
-			 * (read by readpage, changed, and then sent here by
+			 * (read by read_folio, changed, and then sent here by
 			 * writepage).  direct2indirect needs to know if unbh
 			 * was already up to date, so it can decide if the
 			 * data in unbh needs to be replaced with data from
@@ -2733,9 +2733,9 @@ static int reiserfs_write_full_page(struct page *page,
 	goto done;
 }
 
-static int reiserfs_readpage(struct file *f, struct page *page)
+static int reiserfs_read_folio(struct file *f, struct folio *folio)
 {
-	return block_read_full_page(page, reiserfs_get_block);
+	return block_read_full_folio(folio, reiserfs_get_block);
 }
 
 static int reiserfs_writepage(struct page *page, struct writeback_control *wbc)
@@ -3421,7 +3421,7 @@ int reiserfs_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 
 const struct address_space_operations reiserfs_address_space_operations = {
 	.writepage = reiserfs_writepage,
-	.readpage = reiserfs_readpage,
+	.read_folio = reiserfs_read_folio,
 	.readahead = reiserfs_readahead,
 	.releasepage = reiserfs_releasepage,
 	.invalidate_folio = reiserfs_invalidate_folio,
diff --git a/fs/sysv/itree.c b/fs/sysv/itree.c
index 96ad24fe0ffb..d4ec9bb97de9 100644
--- a/fs/sysv/itree.c
+++ b/fs/sysv/itree.c
@@ -456,9 +456,9 @@ static int sysv_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page,get_block,wbc);
 }
 
-static int sysv_readpage(struct file *file, struct page *page)
+static int sysv_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page,get_block);
+	return block_read_full_folio(folio, get_block);
 }
 
 int sysv_prepare_chunk(struct page *page, loff_t pos, unsigned len)
@@ -497,7 +497,7 @@ static sector_t sysv_bmap(struct address_space *mapping, sector_t block)
 const struct address_space_operations sysv_aops = {
 	.dirty_folio = block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = sysv_readpage,
+	.read_folio = sysv_read_folio,
 	.writepage = sysv_writepage,
 	.write_begin = sysv_write_begin,
 	.write_end = generic_write_end,
diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c
index 6c973b71cab2..a873de7dec1c 100644
--- a/fs/ufs/inode.c
+++ b/fs/ufs/inode.c
@@ -390,7 +390,7 @@ ufs_inode_getblock(struct inode *inode, u64 ind_block,
 
 /**
  * ufs_getfrag_block() - `get_block_t' function, interface between UFS and
- * readpage, writepage and so on
+ * read_folio, writepage and so on
  */
 
 static int ufs_getfrag_block(struct inode *inode, sector_t fragment, struct buffer_head *bh_result, int create)
@@ -472,9 +472,9 @@ static int ufs_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page,ufs_getfrag_block,wbc);
 }
 
-static int ufs_readpage(struct file *file, struct page *page)
+static int ufs_read_folio(struct file *file, struct folio *folio)
 {
-	return block_read_full_page(page,ufs_getfrag_block);
+	return block_read_full_folio(folio, ufs_getfrag_block);
 }
 
 int ufs_prepare_chunk(struct page *page, loff_t pos, unsigned len)
@@ -527,7 +527,7 @@ static sector_t ufs_bmap(struct address_space *mapping, sector_t block)
 const struct address_space_operations ufs_aops = {
 	.dirty_folio = block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = ufs_readpage,
+	.read_folio = ufs_read_folio,
 	.writepage = ufs_writepage,
 	.write_begin = ufs_write_begin,
 	.write_end = ufs_write_end,
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 805c4e12700a..31d82fd9abe8 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -223,7 +223,7 @@ int block_write_full_page(struct page *page, get_block_t *get_block,
 int __block_write_full_page(struct inode *inode, struct page *page,
 			get_block_t *get_block, struct writeback_control *wbc,
 			bh_end_io_t *handler);
-int block_read_full_page(struct page*, get_block_t*);
+int block_read_full_folio(struct folio *, get_block_t *);
 bool block_is_partially_uptodate(struct folio *, size_t from, size_t count);
 int block_write_begin(struct address_space *mapping, loff_t pos, unsigned len,
 		struct page **pagep, get_block_t *get_block);
-- 
2.34.1

[PATCH 06/37] fs: Convert mpage_readpage to mpage_read_folio

[Matthew Wilcox (Oracle)]

mpage_readpage still works in terms of pages, and has not been audited
for correctness with large folios, so include an assertion that the
filesystem is not passing it large folios.  Convert all the filesystems
to call mpage_read_folio() instead of mpage_readpage().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/exfat/inode.c       |  6 +++---
 fs/ext2/inode.c        |  8 ++++----
 fs/fat/inode.c         |  6 +++---
 fs/gfs2/aops.c         | 15 +++++++--------
 fs/hpfs/file.c         |  6 +++---
 fs/iomap/buffered-io.c |  2 +-
 fs/isofs/inode.c       |  6 +++---
 fs/jfs/inode.c         |  6 +++---
 fs/mpage.c             |  8 +++++---
 fs/nilfs2/inode.c      | 10 +++++-----
 fs/ntfs3/inode.c       |  9 +++++----
 fs/qnx6/inode.c        |  6 +++---
 fs/udf/inode.c         |  6 +++---
 include/linux/mpage.h  |  2 +-
 14 files changed, 49 insertions(+), 47 deletions(-)

diff --git a/fs/exfat/inode.c b/fs/exfat/inode.c
index b9f63113db2d..0133d385d8e8 100644
--- a/fs/exfat/inode.c
+++ b/fs/exfat/inode.c
@@ -357,9 +357,9 @@ static int exfat_get_block(struct inode *inode, sector_t iblock,
 	return err;
 }
 
-static int exfat_readpage(struct file *file, struct page *page)
+static int exfat_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, exfat_get_block);
+	return mpage_read_folio(folio, exfat_get_block);
 }
 
 static void exfat_readahead(struct readahead_control *rac)
@@ -492,7 +492,7 @@ int exfat_block_truncate_page(struct inode *inode, loff_t from)
 static const struct address_space_operations exfat_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= exfat_readpage,
+	.read_folio	= exfat_read_folio,
 	.readahead	= exfat_readahead,
 	.writepage	= exfat_writepage,
 	.writepages	= exfat_writepages,
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index d8ca8050945a..9e1ecd89f47f 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -875,9 +875,9 @@ static int ext2_writepage(struct page *page, struct writeback_control *wbc)
 	return block_write_full_page(page, ext2_get_block, wbc);
 }
 
-static int ext2_readpage(struct file *file, struct page *page)
+static int ext2_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, ext2_get_block);
+	return mpage_read_folio(folio, ext2_get_block);
 }
 
 static void ext2_readahead(struct readahead_control *rac)
@@ -966,7 +966,7 @@ ext2_dax_writepages(struct address_space *mapping, struct writeback_control *wbc
 const struct address_space_operations ext2_aops = {
 	.dirty_folio		= block_dirty_folio,
 	.invalidate_folio	= block_invalidate_folio,
-	.readpage		= ext2_readpage,
+	.read_folio		= ext2_read_folio,
 	.readahead		= ext2_readahead,
 	.writepage		= ext2_writepage,
 	.write_begin		= ext2_write_begin,
@@ -982,7 +982,7 @@ const struct address_space_operations ext2_aops = {
 const struct address_space_operations ext2_nobh_aops = {
 	.dirty_folio		= block_dirty_folio,
 	.invalidate_folio	= block_invalidate_folio,
-	.readpage		= ext2_readpage,
+	.read_folio		= ext2_read_folio,
 	.readahead		= ext2_readahead,
 	.writepage		= ext2_nobh_writepage,
 	.write_begin		= ext2_nobh_write_begin,
diff --git a/fs/fat/inode.c b/fs/fat/inode.c
index 1f15b0fd1bb0..8a81017f8d60 100644
--- a/fs/fat/inode.c
+++ b/fs/fat/inode.c
@@ -205,9 +205,9 @@ static int fat_writepages(struct address_space *mapping,
 	return mpage_writepages(mapping, wbc, fat_get_block);
 }
 
-static int fat_readpage(struct file *file, struct page *page)
+static int fat_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, fat_get_block);
+	return mpage_read_folio(folio, fat_get_block);
 }
 
 static void fat_readahead(struct readahead_control *rac)
@@ -344,7 +344,7 @@ int fat_block_truncate_page(struct inode *inode, loff_t from)
 static const struct address_space_operations fat_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= fat_readpage,
+	.read_folio	= fat_read_folio,
 	.readahead	= fat_readahead,
 	.writepage	= fat_writepage,
 	.writepages	= fat_writepages,
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index a29eb1e5bfe2..340bf5d0e835 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -480,7 +480,7 @@ static int __gfs2_readpage(void *file, struct page *page)
 		error = stuffed_readpage(ip, page);
 		unlock_page(page);
 	} else {
-		error = mpage_readpage(page, gfs2_block_map);
+		error = mpage_read_folio(folio, gfs2_block_map);
 	}
 
 	if (unlikely(gfs2_withdrawn(sdp)))
@@ -490,14 +490,13 @@ static int __gfs2_readpage(void *file, struct page *page)
 }
 
 /**
- * gfs2_readpage - read a page of a file
+ * gfs2_read_folio - read a folio from a file
  * @file: The file to read
- * @page: The page of the file
+ * @folio: The folio in the file
  */
-
-static int gfs2_readpage(struct file *file, struct page *page)
+static int gfs2_read_folio(struct file *file, struct folio *folio)
 {
-	return __gfs2_readpage(file, page);
+	return __gfs2_readpage(file, &folio->page);
 }
 
 /**
@@ -773,7 +772,7 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 static const struct address_space_operations gfs2_aops = {
 	.writepage = gfs2_writepage,
 	.writepages = gfs2_writepages,
-	.readpage = gfs2_readpage,
+	.read_folio = gfs2_read_folio,
 	.readahead = gfs2_readahead,
 	.dirty_folio = filemap_dirty_folio,
 	.releasepage = iomap_releasepage,
@@ -788,7 +787,7 @@ static const struct address_space_operations gfs2_aops = {
 static const struct address_space_operations gfs2_jdata_aops = {
 	.writepage = gfs2_jdata_writepage,
 	.writepages = gfs2_jdata_writepages,
-	.readpage = gfs2_readpage,
+	.read_folio = gfs2_read_folio,
 	.readahead = gfs2_readahead,
 	.dirty_folio = jdata_dirty_folio,
 	.bmap = gfs2_bmap,
diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c
index 8b590b3826c3..f7547a62c81f 100644
--- a/fs/hpfs/file.c
+++ b/fs/hpfs/file.c
@@ -158,9 +158,9 @@ static const struct iomap_ops hpfs_iomap_ops = {
 	.iomap_begin		= hpfs_iomap_begin,
 };
 
-static int hpfs_readpage(struct file *file, struct page *page)
+static int hpfs_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, hpfs_get_block);
+	return mpage_read_folio(folio, hpfs_get_block);
 }
 
 static int hpfs_writepage(struct page *page, struct writeback_control *wbc)
@@ -247,7 +247,7 @@ static int hpfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 const struct address_space_operations hpfs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = hpfs_readpage,
+	.read_folio = hpfs_read_folio,
 	.writepage = hpfs_writepage,
 	.readahead = hpfs_readahead,
 	.writepages = hpfs_writepages,
diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 75eb0c27a0e8..2de087ac87b6 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -297,7 +297,7 @@ static loff_t iomap_readpage_iter(const struct iomap_iter *iter,
 		/*
 		 * If the bio_alloc fails, try it again for a single page to
 		 * avoid having to deal with partial page reads.  This emulates
-		 * what do_mpage_readpage does.
+		 * what do_mpage_read_folio does.
 		 */
 		if (!ctx->bio) {
 			ctx->bio = bio_alloc(iomap->bdev, 1, REQ_OP_READ,
diff --git a/fs/isofs/inode.c b/fs/isofs/inode.c
index d7491692aea3..88bf20303466 100644
--- a/fs/isofs/inode.c
+++ b/fs/isofs/inode.c
@@ -1174,9 +1174,9 @@ struct buffer_head *isofs_bread(struct inode *inode, sector_t block)
 	return sb_bread(inode->i_sb, blknr);
 }
 
-static int isofs_readpage(struct file *file, struct page *page)
+static int isofs_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, isofs_get_block);
+	return mpage_read_folio(folio, isofs_get_block);
 }
 
 static void isofs_readahead(struct readahead_control *rac)
@@ -1190,7 +1190,7 @@ static sector_t _isofs_bmap(struct address_space *mapping, sector_t block)
 }
 
 static const struct address_space_operations isofs_aops = {
-	.readpage = isofs_readpage,
+	.read_folio = isofs_read_folio,
 	.readahead = isofs_readahead,
 	.bmap = _isofs_bmap
 };
diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c
index aa9f112107b2..a5dd7e53754a 100644
--- a/fs/jfs/inode.c
+++ b/fs/jfs/inode.c
@@ -293,9 +293,9 @@ static int jfs_writepages(struct address_space *mapping,
 	return mpage_writepages(mapping, wbc, jfs_get_block);
 }
 
-static int jfs_readpage(struct file *file, struct page *page)
+static int jfs_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, jfs_get_block);
+	return mpage_read_folio(folio, jfs_get_block);
 }
 
 static void jfs_readahead(struct readahead_control *rac)
@@ -359,7 +359,7 @@ static ssize_t jfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 const struct address_space_operations jfs_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= jfs_readpage,
+	.read_folio	= jfs_read_folio,
 	.readahead	= jfs_readahead,
 	.writepage	= jfs_writepage,
 	.writepages	= jfs_writepages,
diff --git a/fs/mpage.c b/fs/mpage.c
index a04439b84ae2..6df9c3aa5728 100644
--- a/fs/mpage.c
+++ b/fs/mpage.c
@@ -364,20 +364,22 @@ EXPORT_SYMBOL(mpage_readahead);
 /*
  * This isn't called much at all
  */
-int mpage_readpage(struct page *page, get_block_t get_block)
+int mpage_read_folio(struct folio *folio, get_block_t get_block)
 {
 	struct mpage_readpage_args args = {
-		.page = page,
+		.page = &folio->page,
 		.nr_pages = 1,
 		.get_block = get_block,
 	};
 
+	VM_BUG_ON_FOLIO(folio_test_large(folio), folio);
+
 	args.bio = do_mpage_readpage(&args);
 	if (args.bio)
 		mpage_bio_submit(args.bio);
 	return 0;
 }
-EXPORT_SYMBOL(mpage_readpage);
+EXPORT_SYMBOL(mpage_read_folio);
 
 /*
  * Writing is not so simple.
diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 02297ec8dc55..26b8065401b0 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -140,14 +140,14 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff,
 }
 
 /**
- * nilfs_readpage() - implement readpage() method of nilfs_aops {}
+ * nilfs_read_folio() - implement read_folio() method of nilfs_aops {}
  * address_space_operations.
  * @file - file struct of the file to be read
- * @page - the page to be read
+ * @folio - the folio to be read
  */
-static int nilfs_readpage(struct file *file, struct page *page)
+static int nilfs_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, nilfs_get_block);
+	return mpage_read_folio(folio, nilfs_get_block);
 }
 
 static void nilfs_readahead(struct readahead_control *rac)
@@ -298,7 +298,7 @@ nilfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 
 const struct address_space_operations nilfs_aops = {
 	.writepage		= nilfs_writepage,
-	.readpage		= nilfs_readpage,
+	.read_folio		= nilfs_read_folio,
 	.writepages		= nilfs_writepages,
 	.dirty_folio		= nilfs_dirty_folio,
 	.readahead		= nilfs_readahead,
diff --git a/fs/ntfs3/inode.c b/fs/ntfs3/inode.c
index bfd71f384e21..74f60c457f28 100644
--- a/fs/ntfs3/inode.c
+++ b/fs/ntfs3/inode.c
@@ -676,8 +676,9 @@ static sector_t ntfs_bmap(struct address_space *mapping, sector_t block)
 	return generic_block_bmap(mapping, block, ntfs_get_block_bmap);
 }
 
-static int ntfs_readpage(struct file *file, struct page *page)
+static int ntfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	int err;
 	struct address_space *mapping = page->mapping;
 	struct inode *inode = mapping->host;
@@ -701,7 +702,7 @@ static int ntfs_readpage(struct file *file, struct page *page)
 	}
 
 	/* Normal + sparse files. */
-	return mpage_readpage(page, ntfs_get_block);
+	return mpage_read_folio(folio, ntfs_get_block);
 }
 
 static void ntfs_readahead(struct readahead_control *rac)
@@ -1940,7 +1941,7 @@ const struct inode_operations ntfs_link_inode_operations = {
 };
 
 const struct address_space_operations ntfs_aops = {
-	.readpage	= ntfs_readpage,
+	.read_folio	= ntfs_read_folio,
 	.readahead	= ntfs_readahead,
 	.writepage	= ntfs_writepage,
 	.writepages	= ntfs_writepages,
@@ -1952,7 +1953,7 @@ const struct address_space_operations ntfs_aops = {
 };
 
 const struct address_space_operations ntfs_aops_cmpr = {
-	.readpage	= ntfs_readpage,
+	.read_folio	= ntfs_read_folio,
 	.readahead	= ntfs_readahead,
 };
 // clang-format on
diff --git a/fs/qnx6/inode.c b/fs/qnx6/inode.c
index 9d8e7e9788a1..b9895afca9d1 100644
--- a/fs/qnx6/inode.c
+++ b/fs/qnx6/inode.c
@@ -94,9 +94,9 @@ static int qnx6_check_blockptr(__fs32 ptr)
 	return 1;
 }
 
-static int qnx6_readpage(struct file *file, struct page *page)
+static int qnx6_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, qnx6_get_block);
+	return mpage_read_folio(folio, qnx6_get_block);
 }
 
 static void qnx6_readahead(struct readahead_control *rac)
@@ -496,7 +496,7 @@ static sector_t qnx6_bmap(struct address_space *mapping, sector_t block)
 	return generic_block_bmap(mapping, block, qnx6_get_block);
 }
 static const struct address_space_operations qnx6_aops = {
-	.readpage	= qnx6_readpage,
+	.read_folio	= qnx6_read_folio,
 	.readahead	= qnx6_readahead,
 	.bmap		= qnx6_bmap
 };
diff --git a/fs/udf/inode.c b/fs/udf/inode.c
index 866f9a53248e..edc88716751a 100644
--- a/fs/udf/inode.c
+++ b/fs/udf/inode.c
@@ -193,9 +193,9 @@ static int udf_writepages(struct address_space *mapping,
 	return mpage_writepages(mapping, wbc, udf_get_block);
 }
 
-static int udf_readpage(struct file *file, struct page *page)
+static int udf_read_folio(struct file *file, struct folio *folio)
 {
-	return mpage_readpage(page, udf_get_block);
+	return mpage_read_folio(folio, udf_get_block);
 }
 
 static void udf_readahead(struct readahead_control *rac)
@@ -237,7 +237,7 @@ static sector_t udf_bmap(struct address_space *mapping, sector_t block)
 const struct address_space_operations udf_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= udf_readpage,
+	.read_folio	= udf_read_folio,
 	.readahead	= udf_readahead,
 	.writepage	= udf_writepage,
 	.writepages	= udf_writepages,
diff --git a/include/linux/mpage.h b/include/linux/mpage.h
index f4f5e90a6844..43986f7ec4dd 100644
--- a/include/linux/mpage.h
+++ b/include/linux/mpage.h
@@ -16,7 +16,7 @@ struct writeback_control;
 struct readahead_control;
 
 void mpage_readahead(struct readahead_control *, get_block_t get_block);
-int mpage_readpage(struct page *page, get_block_t get_block);
+int mpage_read_folio(struct folio *folio, get_block_t get_block);
 int mpage_writepages(struct address_space *mapping,
 		struct writeback_control *wbc, get_block_t get_block);
 int mpage_writepage(struct page *page, get_block_t *get_block,
-- 
2.34.1

[PATCH 07/37] fs: Convert simple_readpage to simple_read_folio

[Matthew Wilcox (Oracle)]

This is a full folio conversion; it is prepared to handle folios of
arbitrary size.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/libfs.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/fs/libfs.c b/fs/libfs.c
index a1c10d3163e0..31b0ddf01c31 100644
--- a/fs/libfs.c
+++ b/fs/libfs.c
@@ -539,12 +539,12 @@ int simple_setattr(struct user_namespace *mnt_userns, struct dentry *dentry,
 }
 EXPORT_SYMBOL(simple_setattr);
 
-static int simple_readpage(struct file *file, struct page *page)
+static int simple_read_folio(struct file *file, struct folio *folio)
 {
-	clear_highpage(page);
-	flush_dcache_page(page);
-	SetPageUptodate(page);
-	unlock_page(page);
+	folio_zero_range(folio, 0, folio_size(folio));
+	flush_dcache_folio(folio);
+	folio_mark_uptodate(folio);
+	folio_unlock(folio);
 	return 0;
 }
 
@@ -592,7 +592,7 @@ EXPORT_SYMBOL(simple_write_begin);
  * should extend on what's done here with a call to mark_inode_dirty() in the
  * case that i_size has changed.
  *
- * Use *ONLY* with simple_readpage()
+ * Use *ONLY* with simple_read_folio()
  */
 static int simple_write_end(struct file *file, struct address_space *mapping,
 			loff_t pos, unsigned len, unsigned copied,
@@ -628,7 +628,7 @@ static int simple_write_end(struct file *file, struct address_space *mapping,
  * Provides ramfs-style behavior: data in the pagecache, but no writeback.
  */
 const struct address_space_operations ram_aops = {
-	.readpage	= simple_readpage,
+	.read_folio	= simple_read_folio,
 	.write_begin	= simple_write_begin,
 	.write_end	= simple_write_end,
 	.dirty_folio	= noop_dirty_folio,
-- 
2.34.1

[PATCH 08/37] affs: Convert affs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/affs/file.c    | 5 +++--
 fs/affs/symlink.c | 5 +++--
 2 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/fs/affs/file.c b/fs/affs/file.c
index 5da562cc7fb7..cd00a4c68a12 100644
--- a/fs/affs/file.c
+++ b/fs/affs/file.c
@@ -629,8 +629,9 @@ affs_extent_file_ofs(struct inode *inode, u32 newsize)
 }
 
 static int
-affs_readpage_ofs(struct file *file, struct page *page)
+affs_read_folio_ofs(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	u32 to;
 	int err;
@@ -837,7 +838,7 @@ static int affs_write_end_ofs(struct file *file, struct address_space *mapping,
 const struct address_space_operations affs_aops_ofs = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage = affs_readpage_ofs,
+	.read_folio = affs_read_folio_ofs,
 	//.writepage = affs_writepage_ofs,
 	.write_begin = affs_write_begin_ofs,
 	.write_end = affs_write_end_ofs
diff --git a/fs/affs/symlink.c b/fs/affs/symlink.c
index a7531b26e8f0..31d6446dc166 100644
--- a/fs/affs/symlink.c
+++ b/fs/affs/symlink.c
@@ -11,8 +11,9 @@
 
 #include "affs.h"
 
-static int affs_symlink_readpage(struct file *file, struct page *page)
+static int affs_symlink_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct buffer_head *bh;
 	struct inode *inode = page->mapping->host;
 	char *link = page_address(page);
@@ -67,7 +68,7 @@ static int affs_symlink_readpage(struct file *file, struct page *page)
 }
 
 const struct address_space_operations affs_symlink_aops = {
-	.readpage	= affs_symlink_readpage,
+	.read_folio	= affs_symlink_read_folio,
 };
 
 const struct inode_operations affs_symlink_inode_operations = {
-- 
2.34.1

[PATCH 09/37] afs: Convert afs_symlink_readpage to afs_symlink_read_folio

[Matthew Wilcox (Oracle)]

This function mostly used folios already, and only a few minor changes
were needed.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/afs/file.c | 15 +++++++--------
 1 file changed, 7 insertions(+), 8 deletions(-)

diff --git a/fs/afs/file.c b/fs/afs/file.c
index e277fbe55262..65ef69a1f78e 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -19,7 +19,7 @@
 #include "internal.h"
 
 static int afs_file_mmap(struct file *file, struct vm_area_struct *vma);
-static int afs_symlink_readpage(struct file *file, struct page *page);
+static int afs_symlink_read_folio(struct file *file, struct folio *folio);
 static void afs_invalidate_folio(struct folio *folio, size_t offset,
 			       size_t length);
 static int afs_releasepage(struct page *page, gfp_t gfp_flags);
@@ -63,7 +63,7 @@ const struct address_space_operations afs_file_aops = {
 };
 
 const struct address_space_operations afs_symlink_aops = {
-	.readpage	= afs_symlink_readpage,
+	.read_folio	= afs_symlink_read_folio,
 	.releasepage	= afs_releasepage,
 	.invalidate_folio = afs_invalidate_folio,
 };
@@ -332,11 +332,10 @@ static void afs_issue_read(struct netfs_io_subrequest *subreq)
 	afs_put_read(fsreq);
 }
 
-static int afs_symlink_readpage(struct file *file, struct page *page)
+static int afs_symlink_read_folio(struct file *file, struct folio *folio)
 {
-	struct afs_vnode *vnode = AFS_FS_I(page->mapping->host);
+	struct afs_vnode *vnode = AFS_FS_I(folio->mapping->host);
 	struct afs_read *fsreq;
-	struct folio *folio = page_folio(page);
 	int ret;
 
 	fsreq = afs_alloc_read(GFP_NOFS);
@@ -347,13 +346,13 @@ static int afs_symlink_readpage(struct file *file, struct page *page)
 	fsreq->len	= folio_size(folio);
 	fsreq->vnode	= vnode;
 	fsreq->iter	= &fsreq->def_iter;
-	iov_iter_xarray(&fsreq->def_iter, READ, &page->mapping->i_pages,
+	iov_iter_xarray(&fsreq->def_iter, READ, &folio->mapping->i_pages,
 			fsreq->pos, fsreq->len);
 
 	ret = afs_fetch_data(fsreq->vnode, fsreq);
 	if (ret == 0)
-		SetPageUptodate(page);
-	unlock_page(page);
+		folio_mark_uptodate(folio);
+	folio_unlock(folio);
 	return ret;
 }
 
-- 
2.34.1

[PATCH 10/37] befs: Convert befs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/befs/linuxvfs.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c
index 25350dd22cda..be383fa46b12 100644
--- a/fs/befs/linuxvfs.c
+++ b/fs/befs/linuxvfs.c
@@ -48,7 +48,7 @@ static struct inode *befs_iget(struct super_block *, unsigned long);
 static struct inode *befs_alloc_inode(struct super_block *sb);
 static void befs_free_inode(struct inode *inode);
 static void befs_destroy_inodecache(void);
-static int befs_symlink_readpage(struct file *, struct page *);
+static int befs_symlink_read_folio(struct file *, struct folio *);
 static int befs_utf2nls(struct super_block *sb, const char *in, int in_len,
 			char **out, int *out_len);
 static int befs_nls2utf(struct super_block *sb, const char *in, int in_len,
@@ -92,7 +92,7 @@ static const struct address_space_operations befs_aops = {
 };
 
 static const struct address_space_operations befs_symlink_aops = {
-	.readpage	= befs_symlink_readpage,
+	.read_folio	= befs_symlink_read_folio,
 };
 
 static const struct export_operations befs_export_operations = {
@@ -468,8 +468,9 @@ befs_destroy_inodecache(void)
  * The data stream become link name. Unless the LONG_SYMLINK
  * flag is set.
  */
-static int befs_symlink_readpage(struct file *unused, struct page *page)
+static int befs_symlink_read_folio(struct file *unused, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct super_block *sb = inode->i_sb;
 	struct befs_inode_info *befs_ino = BEFS_I(inode);
-- 
2.34.1

[PATCH 11/37] btrfs: Convert btrfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/btrfs/ctree.h            | 2 +-
 fs/btrfs/file.c             | 5 +++--
 fs/btrfs/free-space-cache.c | 2 +-
 fs/btrfs/inode.c            | 7 ++++---
 fs/btrfs/ioctl.c            | 2 +-
 fs/btrfs/relocation.c       | 8 ++++----
 fs/btrfs/send.c             | 2 +-
 7 files changed, 15 insertions(+), 13 deletions(-)

diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h
index 077c95e9baa5..8d4b5edd4059 100644
--- a/fs/btrfs/ctree.h
+++ b/fs/btrfs/ctree.h
@@ -3269,7 +3269,7 @@ void btrfs_split_delalloc_extent(struct inode *inode,
 				 struct extent_state *orig, u64 split);
 void btrfs_set_range_writeback(struct btrfs_inode *inode, u64 start, u64 end);
 vm_fault_t btrfs_page_mkwrite(struct vm_fault *vmf);
-int btrfs_readpage(struct file *file, struct page *page);
+int btrfs_read_folio(struct file *file, struct folio *folio);
 void btrfs_evict_inode(struct inode *inode);
 int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc);
 struct inode *btrfs_alloc_inode(struct super_block *sb);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 59510d7b1c65..373df5ebaf8d 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1307,11 +1307,12 @@ static int prepare_uptodate_page(struct inode *inode,
 				 struct page *page, u64 pos,
 				 bool force_uptodate)
 {
+	struct folio *folio = page_folio(page);
 	int ret = 0;
 
 	if (((pos & (PAGE_SIZE - 1)) || force_uptodate) &&
 	    !PageUptodate(page)) {
-		ret = btrfs_readpage(NULL, page);
+		ret = btrfs_read_folio(NULL, folio);
 		if (ret)
 			return ret;
 		lock_page(page);
@@ -1321,7 +1322,7 @@ static int prepare_uptodate_page(struct inode *inode,
 		}
 
 		/*
-		 * Since btrfs_readpage() will unlock the page before it
+		 * Since btrfs_read_folio() will unlock the folio before it
 		 * returns, there is a window where btrfs_releasepage() can be
 		 * called to release the page.  Here we check both inode
 		 * mapping and PagePrivate() to make sure the page was not
diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c
index 01a408db5683..829a414a7ecb 100644
--- a/fs/btrfs/free-space-cache.c
+++ b/fs/btrfs/free-space-cache.c
@@ -465,7 +465,7 @@ static int io_ctl_prepare_pages(struct btrfs_io_ctl *io_ctl, bool uptodate)
 
 		io_ctl->pages[i] = page;
 		if (uptodate && !PageUptodate(page)) {
-			btrfs_readpage(NULL, page);
+			btrfs_read_folio(NULL, page_folio(page));
 			lock_page(page);
 			if (page->mapping != inode->i_mapping) {
 				btrfs_err(BTRFS_I(inode)->root->fs_info,
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 95c499b8424e..9a2530b1695d 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -4714,7 +4714,7 @@ int btrfs_truncate_block(struct btrfs_inode *inode, loff_t from, loff_t len,
 		goto out_unlock;
 
 	if (!PageUptodate(page)) {
-		ret = btrfs_readpage(NULL, page);
+		ret = btrfs_read_folio(NULL, page_folio(page));
 		lock_page(page);
 		if (page->mapping != mapping) {
 			unlock_page(page);
@@ -8113,8 +8113,9 @@ static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	return extent_fiemap(BTRFS_I(inode), fieinfo, start, len);
 }
 
-int btrfs_readpage(struct file *file, struct page *page)
+int btrfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct btrfs_inode *inode = BTRFS_I(page->mapping->host);
 	u64 start = page_offset(page);
 	u64 end = start + PAGE_SIZE - 1;
@@ -11357,7 +11358,7 @@ static const struct file_operations btrfs_dir_file_operations = {
  * For now we're avoiding this by dropping bmap.
  */
 static const struct address_space_operations btrfs_aops = {
-	.readpage	= btrfs_readpage,
+	.read_folio	= btrfs_read_folio,
 	.writepage	= btrfs_writepage,
 	.writepages	= btrfs_writepages,
 	.readahead	= btrfs_readahead,
diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c
index be6c24577dbe..8d0c4d23b743 100644
--- a/fs/btrfs/ioctl.c
+++ b/fs/btrfs/ioctl.c
@@ -1359,7 +1359,7 @@ static struct page *defrag_prepare_one_page(struct btrfs_inode *inode,
 	 * make it uptodate.
 	 */
 	if (!PageUptodate(page)) {
-		btrfs_readpage(NULL, page);
+		btrfs_read_folio(NULL, page_folio(page));
 		lock_page(page);
 		if (page->mapping != mapping || !PagePrivate(page)) {
 			unlock_page(page);
diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index 9ae06895ffc9..fb16c484bbae 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -1101,7 +1101,7 @@ int replace_file_extents(struct btrfs_trans_handle *trans,
 			continue;
 
 		/*
-		 * if we are modifying block in fs tree, wait for readpage
+		 * if we are modifying block in fs tree, wait for read_folio
 		 * to complete and drop the extent cache
 		 */
 		if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) {
@@ -1563,7 +1563,7 @@ static int invalidate_extent_cache(struct btrfs_root *root,
 			end = (u64)-1;
 		}
 
-		/* the lock_extent waits for readpage to complete */
+		/* the lock_extent waits for read_folio to complete */
 		lock_extent(&BTRFS_I(inode)->io_tree, start, end);
 		btrfs_drop_extent_cache(BTRFS_I(inode), start, end, 1);
 		unlock_extent(&BTRFS_I(inode)->io_tree, start, end);
@@ -2818,7 +2818,7 @@ static noinline_for_stack int prealloc_file_extent_cluster(
 		 * Subpage can't handle page with DIRTY but without UPTODATE
 		 * bit as it can lead to the following deadlock:
 		 *
-		 * btrfs_readpage()
+		 * btrfs_read_folio()
 		 * | Page already *locked*
 		 * |- btrfs_lock_and_flush_ordered_range()
 		 *    |- btrfs_start_ordered_extent()
@@ -2972,7 +2972,7 @@ static int relocate_one_page(struct inode *inode, struct file_ra_state *ra,
 				last_index + 1 - page_index);
 
 	if (!PageUptodate(page)) {
-		btrfs_readpage(NULL, page);
+		btrfs_read_folio(NULL, page_folio(page));
 		lock_page(page);
 		if (!PageUptodate(page)) {
 			ret = -EIO;
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index b327dbe0cbf5..8985d115559d 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -4991,7 +4991,7 @@ static int put_file_data(struct send_ctx *sctx, u64 offset, u32 len)
 		}
 
 		if (!PageUptodate(page)) {
-			btrfs_readpage(NULL, page);
+			btrfs_read_folio(NULL, page_folio(page));
 			lock_page(page);
 			if (!PageUptodate(page)) {
 				unlock_page(page);
-- 
2.34.1

[PATCH 12/37] cifs: Convert cifs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
CIFS should probably be converted to use netfs_read_folio() by someone
familiar with it.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/cifs/file.c | 13 +++++++------
 1 file changed, 7 insertions(+), 6 deletions(-)

diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index da362b5a0c96..bc6d88e2e672 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -4612,8 +4612,9 @@ static int cifs_readpage_worker(struct file *file, struct page *page,
 	return rc;
 }
 
-static int cifs_readpage(struct file *file, struct page *page)
+static int cifs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	loff_t offset = page_file_offset(page);
 	int rc = -EACCES;
 	unsigned int xid;
@@ -4626,7 +4627,7 @@ static int cifs_readpage(struct file *file, struct page *page)
 		return rc;
 	}
 
-	cifs_dbg(FYI, "readpage %p at offset %d 0x%x\n",
+	cifs_dbg(FYI, "read_folio %p at offset %d 0x%x\n",
 		 page, (int)offset, (int)offset);
 
 	rc = cifs_readpage_worker(file, page, &offset);
@@ -4965,7 +4966,7 @@ static bool cifs_dirty_folio(struct address_space *mapping, struct folio *folio)
 #endif
 
 const struct address_space_operations cifs_addr_ops = {
-	.readpage = cifs_readpage,
+	.read_folio = cifs_read_folio,
 	.readahead = cifs_readahead,
 	.writepage = cifs_writepage,
 	.writepages = cifs_writepages,
@@ -4986,12 +4987,12 @@ const struct address_space_operations cifs_addr_ops = {
 };
 
 /*
- * cifs_readpages requires the server to support a buffer large enough to
+ * cifs_readahead requires the server to support a buffer large enough to
  * contain the header plus one complete page of data.  Otherwise, we need
- * to leave cifs_readpages out of the address space operations.
+ * to leave cifs_readahead out of the address space operations.
  */
 const struct address_space_operations cifs_addr_ops_smallbuf = {
-	.readpage = cifs_readpage,
+	.read_folio = cifs_read_folio,
 	.writepage = cifs_writepage,
 	.writepages = cifs_writepages,
 	.write_begin = cifs_write_begin,
-- 
2.34.1

[PATCH 13/37] coda: Convert coda to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/coda/symlink.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/coda/symlink.c b/fs/coda/symlink.c
index 8907d0508198..8adf81042498 100644
--- a/fs/coda/symlink.c
+++ b/fs/coda/symlink.c
@@ -20,9 +20,10 @@
 #include "coda_psdev.h"
 #include "coda_linux.h"
 
-static int coda_symlink_filler(struct file *file, struct page *page)
+static int coda_symlink_filler(struct file *file, struct folio *folio)
 {
-	struct inode *inode = page->mapping->host;
+	struct page *page = &folio->page;
+	struct inode *inode = folio->mapping->host;
 	int error;
 	struct coda_inode_info *cii;
 	unsigned int len = PAGE_SIZE;
@@ -44,5 +45,5 @@ static int coda_symlink_filler(struct file *file, struct page *page)
 }
 
 const struct address_space_operations coda_symlink_aops = {
-	.readpage	= coda_symlink_filler,
+	.read_folio	= coda_symlink_filler,
 };
-- 
2.34.1

[PATCH 14/37] cramfs: Convert cramfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/cramfs/README  | 8 ++++----
 fs/cramfs/inode.c | 7 ++++---
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/cramfs/README b/fs/cramfs/README
index d71b27e0ff15..778df5c4d70b 100644
--- a/fs/cramfs/README
+++ b/fs/cramfs/README
@@ -115,7 +115,7 @@ Block Size
 
 (Block size in cramfs refers to the size of input data that is
 compressed at a time.  It's intended to be somewhere around
-PAGE_SIZE for cramfs_readpage's convenience.)
+PAGE_SIZE for cramfs_read_folio's convenience.)
 
 The superblock ought to indicate the block size that the fs was
 written for, since comments in <linux/pagemap.h> indicate that
@@ -161,7 +161,7 @@ size.  The options are:
      PAGE_SIZE.
 
 It's easy enough to change the kernel to use a smaller value than
-PAGE_SIZE: just make cramfs_readpage read multiple blocks.
+PAGE_SIZE: just make cramfs_read_folio read multiple blocks.
 
 The cost of option 1 is that kernels with a larger PAGE_SIZE
 value don't get as good compression as they can.
@@ -173,9 +173,9 @@ they don't mind their cramfs being inaccessible to kernels with
 smaller PAGE_SIZE values.
 
 Option 3 is easy to implement if we don't mind being CPU-inefficient:
-e.g. get readpage to decompress to a buffer of size MAX_BLKSIZE (which
+e.g. get read_folio to decompress to a buffer of size MAX_BLKSIZE (which
 must be no larger than 32KB) and discard what it doesn't need.
-Getting readpage to read into all the covered pages is harder.
+Getting read_folio to read into all the covered pages is harder.
 
 The main advantage of option 3 over 1, 2, is better compression.  The
 cost is greater complexity.  Probably not worth it, but I hope someone
diff --git a/fs/cramfs/inode.c b/fs/cramfs/inode.c
index 666aa380011e..7ae59a6afc5c 100644
--- a/fs/cramfs/inode.c
+++ b/fs/cramfs/inode.c
@@ -414,7 +414,7 @@ static int cramfs_physmem_mmap(struct file *file, struct vm_area_struct *vma)
 		/*
 		 * Let's create a mixed map if we can't map it all.
 		 * The normal paging machinery will take care of the
-		 * unpopulated ptes via cramfs_readpage().
+		 * unpopulated ptes via cramfs_read_folio().
 		 */
 		int i;
 		vma->vm_flags |= VM_MIXEDMAP;
@@ -814,8 +814,9 @@ static struct dentry *cramfs_lookup(struct inode *dir, struct dentry *dentry, un
 	return d_splice_alias(inode, dentry);
 }
 
-static int cramfs_readpage(struct file *file, struct page *page)
+static int cramfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	u32 maxblock;
 	int bytes_filled;
@@ -925,7 +926,7 @@ static int cramfs_readpage(struct file *file, struct page *page)
 }
 
 static const struct address_space_operations cramfs_aops = {
-	.readpage = cramfs_readpage
+	.read_folio = cramfs_read_folio
 };
 
 /*
-- 
2.34.1

[PATCH 15/37] ecryptfs: Convert ecryptfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ecryptfs/mmap.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c
index 47904d40ef88..19af229eb7ca 100644
--- a/fs/ecryptfs/mmap.c
+++ b/fs/ecryptfs/mmap.c
@@ -170,16 +170,17 @@ ecryptfs_copy_up_encrypted_with_header(struct page *page,
 }
 
 /**
- * ecryptfs_readpage
+ * ecryptfs_read_folio
  * @file: An eCryptfs file
- * @page: Page from eCryptfs inode mapping into which to stick the read data
+ * @folio: Folio from eCryptfs inode mapping into which to stick the read data
  *
- * Read in a page, decrypting if necessary.
+ * Read in a folio, decrypting if necessary.
  *
  * Returns zero on success; non-zero on error.
  */
-static int ecryptfs_readpage(struct file *file, struct page *page)
+static int ecryptfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct ecryptfs_crypt_stat *crypt_stat =
 		&ecryptfs_inode_to_private(page->mapping->host)->crypt_stat;
 	int rc = 0;
@@ -549,7 +550,7 @@ const struct address_space_operations ecryptfs_aops = {
 	.invalidate_folio = block_invalidate_folio,
 #endif
 	.writepage = ecryptfs_writepage,
-	.readpage = ecryptfs_readpage,
+	.read_folio = ecryptfs_read_folio,
 	.write_begin = ecryptfs_write_begin,
 	.write_end = ecryptfs_write_end,
 	.bmap = ecryptfs_bmap,
-- 
2.34.1

[PATCH 16/37] efs: Convert efs symlinks to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/efs/symlink.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/efs/symlink.c b/fs/efs/symlink.c
index 923eb91654d5..3b03a573cb1a 100644
--- a/fs/efs/symlink.c
+++ b/fs/efs/symlink.c
@@ -12,8 +12,9 @@
 #include <linux/buffer_head.h>
 #include "efs.h"
 
-static int efs_symlink_readpage(struct file *file, struct page *page)
+static int efs_symlink_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	char *link = page_address(page);
 	struct buffer_head * bh;
 	struct inode * inode = page->mapping->host;
@@ -49,5 +50,5 @@ static int efs_symlink_readpage(struct file *file, struct page *page)
 }
 
 const struct address_space_operations efs_symlink_aops = {
-	.readpage	= efs_symlink_readpage
+	.read_folio	= efs_symlink_read_folio
 };
-- 
2.34.1

[PATCH 17/37] erofs: Convert erofs zdata to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/erofs/zdata.c | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/fs/erofs/zdata.c b/fs/erofs/zdata.c
index e6dea6dfca16..95efc127b2ba 100644
--- a/fs/erofs/zdata.c
+++ b/fs/erofs/zdata.c
@@ -791,7 +791,7 @@ static int z_erofs_do_read_page(struct z_erofs_decompress_frontend *fe,
 static bool z_erofs_get_sync_decompress_policy(struct erofs_sb_info *sbi,
 				       unsigned int readahead_pages)
 {
-	/* auto: enable for readpage, disable for readahead */
+	/* auto: enable for read_folio, disable for readahead */
 	if ((sbi->opt.sync_decompress == EROFS_SYNC_DECOMPRESS_AUTO) &&
 	    !readahead_pages)
 		return true;
@@ -1488,8 +1488,9 @@ static void z_erofs_pcluster_readmore(struct z_erofs_decompress_frontend *f,
 	}
 }
 
-static int z_erofs_readpage(struct file *file, struct page *page)
+static int z_erofs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *const inode = page->mapping->host;
 	struct erofs_sb_info *const sbi = EROFS_I_SB(inode);
 	struct z_erofs_decompress_frontend f = DECOMPRESS_FRONTEND_INIT(inode);
@@ -1563,6 +1564,6 @@ static void z_erofs_readahead(struct readahead_control *rac)
 }
 
 const struct address_space_operations z_erofs_aops = {
-	.readpage = z_erofs_readpage,
+	.read_folio = z_erofs_read_folio,
 	.readahead = z_erofs_readahead,
 };
-- 
2.34.1

[PATCH 18/37] ext4: Convert ext4 to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ext4/inode.c       | 9 +++++----
 fs/ext4/move_extent.c | 4 ++--
 2 files changed, 7 insertions(+), 6 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index d3a7e8581291..c6b8cb4949f1 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3180,8 +3180,9 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block)
 	return iomap_bmap(mapping, block, &ext4_iomap_ops);
 }
 
-static int ext4_readpage(struct file *file, struct page *page)
+static int ext4_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	int ret = -EAGAIN;
 	struct inode *inode = page->mapping->host;
 
@@ -3608,7 +3609,7 @@ static int ext4_iomap_swap_activate(struct swap_info_struct *sis,
 }
 
 static const struct address_space_operations ext4_aops = {
-	.readpage		= ext4_readpage,
+	.read_folio		= ext4_read_folio,
 	.readahead		= ext4_readahead,
 	.writepage		= ext4_writepage,
 	.writepages		= ext4_writepages,
@@ -3626,7 +3627,7 @@ static const struct address_space_operations ext4_aops = {
 };
 
 static const struct address_space_operations ext4_journalled_aops = {
-	.readpage		= ext4_readpage,
+	.read_folio		= ext4_read_folio,
 	.readahead		= ext4_readahead,
 	.writepage		= ext4_writepage,
 	.writepages		= ext4_writepages,
@@ -3643,7 +3644,7 @@ static const struct address_space_operations ext4_journalled_aops = {
 };
 
 static const struct address_space_operations ext4_da_aops = {
-	.readpage		= ext4_readpage,
+	.read_folio		= ext4_read_folio,
 	.readahead		= ext4_readahead,
 	.writepage		= ext4_writepage,
 	.writepages		= ext4_writepages,
diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c
index 4172a7d22471..701f1d6a217f 100644
--- a/fs/ext4/move_extent.c
+++ b/fs/ext4/move_extent.c
@@ -669,8 +669,8 @@ ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 orig_blk,
 		 * Up semaphore to avoid following problems:
 		 * a. transaction deadlock among ext4_journal_start,
 		 *    ->write_begin via pagefault, and jbd2_journal_commit
-		 * b. racing with ->readpage, ->write_begin, and ext4_get_block
-		 *    in move_extent_per_page
+		 * b. racing with ->read_folio, ->write_begin, and
+		 *    ext4_get_block in move_extent_per_page
 		 */
 		ext4_double_up_write_data_sem(orig_inode, donor_inode);
 		/* Swap original branches with new branches */
-- 
2.34.1

Re: [PATCH 18/37] ext4: Convert ext4 to read_folio

[Theodore Ts'o]

On Sun, May 08, 2022 at 09:31:12PM +0100, Matthew Wilcox (Oracle) wrote:
> This is a "weak" conversion which converts straight back to using pages.
> A full conversion should be performed at some point, hopefully by
> someone familiar with the filesystem.

What worries me about the "weak" conversion is that ext4_read_folio()
(formerly ext4_read_page()) is going to completely the wrong thing if
the folio contains more than a single page.  This seems to
be... scary.  How are callers of aops->read_folio() supposed to know
whether the right thing or the wrong thing will happen if a random
folio is passed to aops->read_folio()?

						- Ted

Re: [PATCH 18/37] ext4: Convert ext4 to read_folio

[Matthew Wilcox]

On Mon, May 09, 2022 at 09:30:52AM -0400, Theodore Ts'o wrote:
> On Sun, May 08, 2022 at 09:31:12PM +0100, Matthew Wilcox (Oracle) wrote:
> > This is a "weak" conversion which converts straight back to using pages.
> > A full conversion should be performed at some point, hopefully by
> > someone familiar with the filesystem.
> 
> What worries me about the "weak" conversion is that ext4_read_folio()
> (formerly ext4_read_page()) is going to completely the wrong thing if
> the folio contains more than a single page.  This seems to
> be... scary.  How are callers of aops->read_folio() supposed to know
> whether the right thing or the wrong thing will happen if a random
> folio is passed to aops->read_folio()?

I'm probably answering these emails out of order, but the page
cache is absolutely not supposed to be creating large folios for
filesystems that haven't indicated their support for such by calling
mapping_set_large_folios().  We could assert that in every filesystem,
but it's on the same scale as asserting that page (or folio) != NULL
and page->mapping != NULL.  Filesystems should trust the VFS to obey
its rules.  Obviously this is a new rule, so people aren't as familiar
with it as "you can rely on the page to not be truncated while we're
calling readpage" or "the page is locked on entry to these aops".

Re: [PATCH 18/37] ext4: Convert ext4 to read_folio

[Theodore Ts'o]

On Mon, May 09, 2022 at 03:07:20PM +0100, Matthew Wilcox wrote:
> 
> I'm probably answering these emails out of order,

No worries, I've been reviewing these patches out of order myself.
:-)

> but the page
> cache is absolutely not supposed to be creating large folios for
> filesystems that haven't indicated their support for such by calling
> mapping_set_large_folios().

I think my concern is that at some point in the future, ext4 probably
*will* want to enable large folios --- and we may want to do so
selectively.  e.g., just on the read-path, and assume that someone
will break apart large folios to individual pages on the write path,
for example.

The question is when do we add all of these sanity check asserts ---
at the point when ext4 starts making the transition from large folio
unaware, to large folio kind-of-aware, and hope we don't miss any of
these interfaces?  Or add those sanity check asserts now, so we get
reminded that some of these functions may need fixing up when we start
adding large folio support to the file system?

Also, what's the intent for when the MM layer would call
aops->read_folio() with the intent to fill a huge folio, versus
calling aops->readahead()?  After all, when we take a page fault,
it'll be either a 4k page, right?  We currently don't support
file-backed huge pages; is there a plan to change this?

		    	  	    	    - Ted

P.S.  On a somewhat unrelated issue, if we have a really large folio
caused by a 4MB readahead because CIFS really wanted a huge readahead
size because of the network setup overhead --- and then a single 4k
page gets dirtied, I imagine the VM subsystem *want* to break apart
that 4MB folio so that we know that only that single 4k page was
dirtied, and not require writing back a huge amount of clean 4k pages
just because we didn't track dirtiness at the right granularity,
right?

Re: [PATCH 18/37] ext4: Convert ext4 to read_folio

[Matthew Wilcox]

On Mon, May 09, 2022 at 04:16:19PM -0400, Theodore Ts'o wrote:
> > but the page
> > cache is absolutely not supposed to be creating large folios for
> > filesystems that haven't indicated their support for such by calling
> > mapping_set_large_folios().
> 
> I think my concern is that at some point in the future, ext4 probably
> *will* want to enable large folios --- and we may want to do so
> selectively.  e.g., just on the read-path, and assume that someone
> will break apart large folios to individual pages on the write path,
> for example.

My mental model doesn't include that last part -- if the filesystem says
it supports large folios, then it supports large folios, including read,
write-into, writeback, etc.  The filesystem can (of course) split the
page itself, but page splitting can fail, and part of the point of all
this is to keep the memory allocation as a single object throughout its
allocation-usage-free lifecycle.

> The question is when do we add all of these sanity check asserts ---
> at the point when ext4 starts making the transition from large folio
> unaware, to large folio kind-of-aware, and hope we don't miss any of
> these interfaces?  Or add those sanity check asserts now, so we get
> reminded that some of these functions may need fixing up when we start
> adding large folio support to the file system?

Right!  I went through the same questions when starting to work on iomap
-- how do we know which of these functions is really capable of dealing
with large folios.  The good news is that xfstests blows up in very
entertaining and obvious ways, so most of this you really can do by
trial and error.  That's obviously a bit unsatisfying ...

As I convert utility functions to take a folio, I try not to leave
landmines; I either convert the function to work on a folio of arbitrary
size, or I leave a VM_BUG_ON_FOLIO in there to help the first person
who tries to use it with a large folio.  Of course, there can always
be places I missed.  If there's anywhere still using a struct page,
that's always cause for greater scrutiny.

> Also, what's the intent for when the MM layer would call
> aops->read_folio() with the intent to fill a huge folio, versus
> calling aops->readahead()?  After all, when we take a page fault,
> it'll be either a 4k page, right?  We currently don't support
> file-backed huge pages; is there a plan to change this?

One of the recent changes to the pagecache (from Jens, iirc) is that
we _always_ call ->readahead() before calling ->read_folio().  That's a
really good thing because it means you can now make your ->read_folio()
synchronous and return the exact error instead of sending off an I/O
and signalling the page "had an error" and returning -EIO to userspace.

The decision about whether to use THP for file-backed faults is
currently left up to userspace.  Commit 4687fdbb805a added support for
the existing MADV_HUGEPAGE to file-backed pages.  I'm not inclined to
leave the decision entirely up to userspace ... I think we should notice
that userspace is behaving in a way that using larger folios would be
beneficial, and eventually end up going all the way to THPs.  However,
the vast majority of applications do not use mmap() on files, so I'm
not too enthusiastic about adding that support until someone comes to
me with a use case.

I did just notice that I need to update the manpage for madvise(2).

> P.S.  On a somewhat unrelated issue, if we have a really large folio
> caused by a 4MB readahead because CIFS really wanted a huge readahead
> size because of the network setup overhead --- and then a single 4k
> page gets dirtied, I imagine the VM subsystem *want* to break apart
> that 4MB folio so that we know that only that single 4k page was
> dirtied, and not require writing back a huge amount of clean 4k pages
> just because we didn't track dirtiness at the right granularity,
> right?

Well ... we don't necessarily grow the folio size all the way up to
the size of the readahead window.  The two are decoupled to a certain
extent (obviously the folio size won't exceed the readahead size!)  And I
currently limit the readahead folio size to PMD_SIZE because I don't want
to track down all the places that assume a PMD page is exactly PMD_SIZE.

But no, the VM doesn't want to break up 4MB pages just because we only
dirtied 4KB of it.  We do tell the FS which parts of it are dirtied, so if
the FS wants to keep track of what is dirty at a sub-folio level, it can.
But it doesn't have to; choosing the granularity of dirtiness is a job
for the filesystem, not the VM.  It can be beneficial to do a larger
write than what is dirty, or some filesystems want to be byte-precise
in what they write back.

If an application has exhibited good locality in reads (enough to get
readahead growing to 4MB), the hope is that it will also exhibit good
locality in writes, so absorbing a lot of writes to the same 4MB page
before it gets written back in a big chunk will actually be a good thing.

Obviously, that's something we'll only know is true once users start
banging on this feature in earnest.  We may need to adjust how we handle
large folios to make users happier.  I don't think we can reasonably
anticipate the problems they'll see.

[PATCH 19/37] f2fs: Convert f2fs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/f2fs/data.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index b3cf49136b9f..f894267f0722 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -2372,8 +2372,9 @@ static int f2fs_mpage_readpages(struct inode *inode,
 	return ret;
 }
 
-static int f2fs_read_data_page(struct file *file, struct page *page)
+static int f2fs_read_data_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page_file_mapping(page)->host;
 	int ret = -EAGAIN;
 
@@ -3935,7 +3936,7 @@ static void f2fs_swap_deactivate(struct file *file)
 #endif
 
 const struct address_space_operations f2fs_dblock_aops = {
-	.readpage	= f2fs_read_data_page,
+	.read_folio	= f2fs_read_data_folio,
 	.readahead	= f2fs_readahead,
 	.writepage	= f2fs_write_data_page,
 	.writepages	= f2fs_write_data_pages,
-- 
2.34.1

[PATCH 20/37] freevxfs: Convert vxfs_immed to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/freevxfs/vxfs_immed.c | 15 ++++++++-------
 1 file changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/freevxfs/vxfs_immed.c b/fs/freevxfs/vxfs_immed.c
index bfc780c682fb..a37431e443d3 100644
--- a/fs/freevxfs/vxfs_immed.c
+++ b/fs/freevxfs/vxfs_immed.c
@@ -38,33 +38,34 @@
 #include "vxfs_inode.h"
 
 
-static int	vxfs_immed_readpage(struct file *, struct page *);
+static int	vxfs_immed_read_folio(struct file *, struct folio *);
 
 /*
  * Address space operations for immed files and directories.
  */
 const struct address_space_operations vxfs_immed_aops = {
-	.readpage =		vxfs_immed_readpage,
+	.read_folio =		vxfs_immed_read_folio,
 };
 
 /**
- * vxfs_immed_readpage - read part of an immed inode into pagecache
+ * vxfs_immed_read_folio - read part of an immed inode into pagecache
  * @file:	file context (unused)
- * @page:	page frame to fill in.
+ * @folio:	folio to fill in.
  *
  * Description:
- *   vxfs_immed_readpage reads a part of the immed area of the
+ *   vxfs_immed_read_folio reads a part of the immed area of the
  *   file that hosts @pp into the pagecache.
  *
  * Returns:
  *   Zero on success, else a negative error code.
  *
  * Locking status:
- *   @page is locked and will be unlocked.
+ *   @folio is locked and will be unlocked.
  */
 static int
-vxfs_immed_readpage(struct file *fp, struct page *pp)
+vxfs_immed_read_folio(struct file *fp, struct folio *folio)
 {
+	struct page *pp = &folio->page;
 	struct vxfs_inode_info	*vip = VXFS_INO(pp->mapping->host);
 	u_int64_t	offset = (u_int64_t)pp->index << PAGE_SHIFT;
 	caddr_t		kaddr;
-- 
2.34.1

[PATCH 21/37] fuse: Convert fuse to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/fuse/dir.c  | 10 +++++-----
 fs/fuse/file.c |  5 +++--
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c
index 9ff27b8a9782..74303d6e987b 100644
--- a/fs/fuse/dir.c
+++ b/fs/fuse/dir.c
@@ -1957,20 +1957,20 @@ void fuse_init_dir(struct inode *inode)
 	fi->rdc.version = 0;
 }
 
-static int fuse_symlink_readpage(struct file *null, struct page *page)
+static int fuse_symlink_read_folio(struct file *null, struct folio *folio)
 {
-	int err = fuse_readlink_page(page->mapping->host, page);
+	int err = fuse_readlink_page(folio->mapping->host, &folio->page);
 
 	if (!err)
-		SetPageUptodate(page);
+		folio_mark_uptodate(folio);
 
-	unlock_page(page);
+	folio_unlock(folio);
 
 	return err;
 }
 
 static const struct address_space_operations fuse_symlink_aops = {
-	.readpage	= fuse_symlink_readpage,
+	.read_folio	= fuse_symlink_read_folio,
 };
 
 void fuse_init_symlink(struct inode *inode)
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index bca8c2135ec5..05caa2b9272e 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -857,8 +857,9 @@ static int fuse_do_readpage(struct file *file, struct page *page)
 	return 0;
 }
 
-static int fuse_readpage(struct file *file, struct page *page)
+static int fuse_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	int err;
 
@@ -3174,7 +3175,7 @@ static const struct file_operations fuse_file_operations = {
 };
 
 static const struct address_space_operations fuse_file_aops  = {
-	.readpage	= fuse_readpage,
+	.read_folio	= fuse_read_folio,
 	.readahead	= fuse_readahead,
 	.writepage	= fuse_writepage,
 	.writepages	= fuse_writepages,
-- 
2.34.1

[PATCH 22/37] hostfs: Convert hostfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/hostfs/hostfs_kern.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c
index e658d8edde35..cc1bc6f93a01 100644
--- a/fs/hostfs/hostfs_kern.c
+++ b/fs/hostfs/hostfs_kern.c
@@ -434,8 +434,9 @@ static int hostfs_writepage(struct page *page, struct writeback_control *wbc)
 	return err;
 }
 
-static int hostfs_readpage(struct file *file, struct page *page)
+static int hostfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	char *buffer;
 	loff_t start = page_offset(page);
 	int bytes_read, ret = 0;
@@ -504,7 +505,7 @@ static int hostfs_write_end(struct file *file, struct address_space *mapping,
 
 static const struct address_space_operations hostfs_aops = {
 	.writepage 	= hostfs_writepage,
-	.readpage	= hostfs_readpage,
+	.read_folio	= hostfs_read_folio,
 	.dirty_folio	= filemap_dirty_folio,
 	.write_begin	= hostfs_write_begin,
 	.write_end	= hostfs_write_end,
-- 
2.34.1

[PATCH 23/37] hpfs: Convert symlinks to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/hpfs/namei.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c
index d73f8a67168e..15fc63276caa 100644
--- a/fs/hpfs/namei.c
+++ b/fs/hpfs/namei.c
@@ -479,8 +479,9 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry)
 	return err;
 }
 
-static int hpfs_symlink_readpage(struct file *file, struct page *page)
+static int hpfs_symlink_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	char *link = page_address(page);
 	struct inode *i = page->mapping->host;
 	struct fnode *fnode;
@@ -508,7 +509,7 @@ static int hpfs_symlink_readpage(struct file *file, struct page *page)
 }
 
 const struct address_space_operations hpfs_symlink_aops = {
-	.readpage	= hpfs_symlink_readpage
+	.read_folio	= hpfs_symlink_read_folio
 };
 
 static int hpfs_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
-- 
2.34.1

[PATCH 24/37] isofs: Convert symlinks and zisofs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/isofs/compress.c | 5 +++--
 fs/isofs/rock.c     | 7 ++++---
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/fs/isofs/compress.c b/fs/isofs/compress.c
index bc12ac7e2312..95a19f25d61c 100644
--- a/fs/isofs/compress.c
+++ b/fs/isofs/compress.c
@@ -296,8 +296,9 @@ static int zisofs_fill_pages(struct inode *inode, int full_page, int pcount,
  * per reference.  We inject the additional pages into the page
  * cache as a form of readahead.
  */
-static int zisofs_readpage(struct file *file, struct page *page)
+static int zisofs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = file_inode(file);
 	struct address_space *mapping = inode->i_mapping;
 	int err;
@@ -369,7 +370,7 @@ static int zisofs_readpage(struct file *file, struct page *page)
 }
 
 const struct address_space_operations zisofs_aops = {
-	.readpage = zisofs_readpage,
+	.read_folio = zisofs_read_folio,
 	/* No bmap operation supported */
 };
 
diff --git a/fs/isofs/rock.c b/fs/isofs/rock.c
index 4880146babaf..48f58c6c9e69 100644
--- a/fs/isofs/rock.c
+++ b/fs/isofs/rock.c
@@ -687,11 +687,12 @@ int parse_rock_ridge_inode(struct iso_directory_record *de, struct inode *inode,
 }
 
 /*
- * readpage() for symlinks: reads symlink contents into the page and either
+ * read_folio() for symlinks: reads symlink contents into the folio and either
  * makes it uptodate and returns 0 or returns error (-EIO)
  */
-static int rock_ridge_symlink_readpage(struct file *file, struct page *page)
+static int rock_ridge_symlink_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct iso_inode_info *ei = ISOFS_I(inode);
 	struct isofs_sb_info *sbi = ISOFS_SB(inode->i_sb);
@@ -804,5 +805,5 @@ static int rock_ridge_symlink_readpage(struct file *file, struct page *page)
 }
 
 const struct address_space_operations isofs_symlink_aops = {
-	.readpage = rock_ridge_symlink_readpage
+	.read_folio = rock_ridge_symlink_read_folio
 };
-- 
2.34.1

[PATCH 25/37] jffs2: Convert jffs2 to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/jffs2/file.c | 10 +++++-----
 fs/jffs2/fs.c   |  2 +-
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index 2b35811772de..f8616683fbee 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -27,7 +27,7 @@ static int jffs2_write_end(struct file *filp, struct address_space *mapping,
 static int jffs2_write_begin(struct file *filp, struct address_space *mapping,
 			loff_t pos, unsigned len,
 			struct page **pagep, void **fsdata);
-static int jffs2_readpage (struct file *filp, struct page *pg);
+static int jffs2_read_folio(struct file *filp, struct folio *folio);
 
 int jffs2_fsync(struct file *filp, loff_t start, loff_t end, int datasync)
 {
@@ -72,7 +72,7 @@ const struct inode_operations jffs2_file_inode_operations =
 
 const struct address_space_operations jffs2_file_address_operations =
 {
-	.readpage =	jffs2_readpage,
+	.read_folio =	jffs2_read_folio,
 	.write_begin =	jffs2_write_begin,
 	.write_end =	jffs2_write_end,
 };
@@ -118,13 +118,13 @@ int jffs2_do_readpage_unlock(void *data, struct page *pg)
 }
 
 
-static int jffs2_readpage (struct file *filp, struct page *pg)
+static int jffs2_read_folio(struct file *file, struct folio *folio)
 {
-	struct jffs2_inode_info *f = JFFS2_INODE_INFO(pg->mapping->host);
+	struct jffs2_inode_info *f = JFFS2_INODE_INFO(folio->mapping->host);
 	int ret;
 
 	mutex_lock(&f->sem);
-	ret = jffs2_do_readpage_unlock(pg->mapping->host, pg);
+	ret = jffs2_do_readpage_unlock(folio->mapping->host, &folio->page);
 	mutex_unlock(&f->sem);
 	return ret;
 }
diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c
index 71f03a5d36ed..00a110f40e10 100644
--- a/fs/jffs2/fs.c
+++ b/fs/jffs2/fs.c
@@ -178,7 +178,7 @@ int jffs2_do_setattr (struct inode *inode, struct iattr *iattr)
 	jffs2_complete_reservation(c);
 
 	/* We have to do the truncate_setsize() without f->sem held, since
-	   some pages may be locked and waiting for it in readpage().
+	   some pages may be locked and waiting for it in read_folio().
 	   We are protected from a simultaneous write() extending i_size
 	   back past iattr->ia_size, because do_truncate() holds the
 	   generic inode semaphore. */
-- 
2.34.1

[PATCH 26/37] jfs: Convert metadata pages to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/jfs/jfs_metapage.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c
index c4220ccdedef..2fc78405b3f2 100644
--- a/fs/jfs/jfs_metapage.c
+++ b/fs/jfs/jfs_metapage.c
@@ -467,8 +467,9 @@ static int metapage_writepage(struct page *page, struct writeback_control *wbc)
 	return -EIO;
 }
 
-static int metapage_readpage(struct file *fp, struct page *page)
+static int metapage_read_folio(struct file *fp, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct bio *bio = NULL;
 	int block_offset;
@@ -563,7 +564,7 @@ static void metapage_invalidate_folio(struct folio *folio, size_t offset,
 }
 
 const struct address_space_operations jfs_metapage_aops = {
-	.readpage	= metapage_readpage,
+	.read_folio	= metapage_read_folio,
 	.writepage	= metapage_writepage,
 	.releasepage	= metapage_releasepage,
 	.invalidate_folio = metapage_invalidate_folio,
-- 
2.34.1

[PATCH 27/37] nfs: Convert nfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/nfs/file.c          | 4 ++--
 fs/nfs/read.c          | 3 ++-
 include/linux/nfs_fs.h | 2 +-
 3 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index f05c4b18b681..4f6d1f90b87f 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -337,7 +337,7 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,
 	} else if (!once_thru &&
 		   nfs_want_read_modify_write(file, page, pos, len)) {
 		once_thru = 1;
-		ret = nfs_readpage(file, page);
+		ret = nfs_read_folio(file, page_folio(page));
 		put_page(page);
 		if (!ret)
 			goto start;
@@ -514,7 +514,7 @@ static void nfs_swap_deactivate(struct file *file)
 }
 
 const struct address_space_operations nfs_file_aops = {
-	.readpage = nfs_readpage,
+	.read_folio = nfs_read_folio,
 	.readahead = nfs_readahead,
 	.dirty_folio = filemap_dirty_folio,
 	.writepage = nfs_writepage,
diff --git a/fs/nfs/read.c b/fs/nfs/read.c
index 5e7657374bc3..5a9b043662e9 100644
--- a/fs/nfs/read.c
+++ b/fs/nfs/read.c
@@ -333,8 +333,9 @@ readpage_async_filler(struct nfs_readdesc *desc, struct page *page)
  *  -	The error flag is set for this page. This happens only when a
  *	previous async read operation failed.
  */
-int nfs_readpage(struct file *file, struct page *page)
+int nfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct nfs_readdesc desc;
 	struct inode *inode = page_file_mapping(page)->host;
 	int ret;
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index b48b9259e02c..1bba71757d62 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -594,7 +594,7 @@ static inline bool nfs_have_writebacks(const struct inode *inode)
 /*
  * linux/fs/nfs/read.c
  */
-extern int  nfs_readpage(struct file *, struct page *);
+int  nfs_read_folio(struct file *, struct folio *);
 void nfs_readahead(struct readahead_control *);
 
 /*
-- 
2.34.1

[PATCH 28/37] ntfs: Convert ntfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ntfs/aops.c   | 40 +++++++++++++++++++++-------------------
 fs/ntfs/aops.h   |  6 +++---
 fs/ntfs/attrib.c |  2 +-
 fs/ntfs/file.c   |  4 ++--
 fs/ntfs/inode.c  |  4 ++--
 fs/ntfs/mft.h    |  2 +-
 6 files changed, 30 insertions(+), 28 deletions(-)

diff --git a/fs/ntfs/aops.c b/fs/ntfs/aops.c
index 90e3dad8ee45..9e3964ea2ea0 100644
--- a/fs/ntfs/aops.c
+++ b/fs/ntfs/aops.c
@@ -159,7 +159,7 @@ static void ntfs_end_buffer_async_read(struct buffer_head *bh, int uptodate)
  *
  * Return 0 on success and -errno on error.
  *
- * Contains an adapted version of fs/buffer.c::block_read_full_page().
+ * Contains an adapted version of fs/buffer.c::block_read_full_folio().
  */
 static int ntfs_read_block(struct page *page)
 {
@@ -358,16 +358,16 @@ static int ntfs_read_block(struct page *page)
 }
 
 /**
- * ntfs_readpage - fill a @page of a @file with data from the device
- * @file:	open file to which the page @page belongs or NULL
- * @page:	page cache page to fill with data
+ * ntfs_read_folio - fill a @folio of a @file with data from the device
+ * @file:	open file to which the folio @folio belongs or NULL
+ * @folio:	page cache folio to fill with data
  *
- * For non-resident attributes, ntfs_readpage() fills the @page of the open
- * file @file by calling the ntfs version of the generic block_read_full_page()
+ * For non-resident attributes, ntfs_read_folio() fills the @folio of the open
+ * file @file by calling the ntfs version of the generic block_read_full_folio()
  * function, ntfs_read_block(), which in turn creates and reads in the buffers
- * associated with the page asynchronously.
+ * associated with the folio asynchronously.
  *
- * For resident attributes, OTOH, ntfs_readpage() fills @page by copying the
+ * For resident attributes, OTOH, ntfs_read_folio() fills @folio by copying the
  * data from the mft record (which at this stage is most likely in memory) and
  * fills the remainder with zeroes. Thus, in this case, I/O is synchronous, as
  * even if the mft record is not cached at this point in time, we need to wait
@@ -375,8 +375,9 @@ static int ntfs_read_block(struct page *page)
  *
  * Return 0 on success and -errno on error.
  */
-static int ntfs_readpage(struct file *file, struct page *page)
+static int ntfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	loff_t i_size;
 	struct inode *vi;
 	ntfs_inode *ni, *base_ni;
@@ -458,7 +459,7 @@ static int ntfs_readpage(struct file *file, struct page *page)
 	}
 	/*
 	 * If a parallel write made the attribute non-resident, drop the mft
-	 * record and retry the readpage.
+	 * record and retry the read_folio.
 	 */
 	if (unlikely(NInoNonResident(ni))) {
 		unmap_mft_record(base_ni);
@@ -637,10 +638,11 @@ static int ntfs_write_block(struct page *page, struct writeback_control *wbc)
 		if (unlikely((block >= iblock) &&
 				(initialized_size < i_size))) {
 			/*
-			 * If this page is fully outside initialized size, zero
-			 * out all pages between the current initialized size
-			 * and the current page. Just use ntfs_readpage() to do
-			 * the zeroing transparently.
+			 * If this page is fully outside initialized
+			 * size, zero out all pages between the current
+			 * initialized size and the current page. Just
+			 * use ntfs_read_folio() to do the zeroing
+			 * transparently.
 			 */
 			if (block > iblock) {
 				// TODO:
@@ -798,7 +800,7 @@ static int ntfs_write_block(struct page *page, struct writeback_control *wbc)
 	/* For the error case, need to reset bh to the beginning. */
 	bh = head;
 
-	/* Just an optimization, so ->readpage() is not called later. */
+	/* Just an optimization, so ->read_folio() is not called later. */
 	if (unlikely(!PageUptodate(page))) {
 		int uptodate = 1;
 		do {
@@ -1329,7 +1331,7 @@ static int ntfs_write_mst_block(struct page *page,
  * vfs inode dirty code path for the inode the mft record belongs to or via the
  * vm page dirty code path for the page the mft record is in.
  *
- * Based on ntfs_readpage() and fs/buffer.c::block_write_full_page().
+ * Based on ntfs_read_folio() and fs/buffer.c::block_write_full_page().
  *
  * Return 0 on success and -errno on error.
  */
@@ -1651,7 +1653,7 @@ static sector_t ntfs_bmap(struct address_space *mapping, sector_t block)
  * attributes.
  */
 const struct address_space_operations ntfs_normal_aops = {
-	.readpage	= ntfs_readpage,
+	.read_folio	= ntfs_read_folio,
 #ifdef NTFS_RW
 	.writepage	= ntfs_writepage,
 	.dirty_folio	= block_dirty_folio,
@@ -1666,7 +1668,7 @@ const struct address_space_operations ntfs_normal_aops = {
  * ntfs_compressed_aops - address space operations for compressed inodes
  */
 const struct address_space_operations ntfs_compressed_aops = {
-	.readpage	= ntfs_readpage,
+	.read_folio	= ntfs_read_folio,
 #ifdef NTFS_RW
 	.writepage	= ntfs_writepage,
 	.dirty_folio	= block_dirty_folio,
@@ -1681,7 +1683,7 @@ const struct address_space_operations ntfs_compressed_aops = {
  *		   and attributes
  */
 const struct address_space_operations ntfs_mst_aops = {
-	.readpage	= ntfs_readpage,	/* Fill page with data. */
+	.read_folio	= ntfs_read_folio,	/* Fill page with data. */
 #ifdef NTFS_RW
 	.writepage	= ntfs_writepage,	/* Write dirty page to disk. */
 	.dirty_folio	= filemap_dirty_folio,
diff --git a/fs/ntfs/aops.h b/fs/ntfs/aops.h
index f0962d46bd67..934d5f79b9e7 100644
--- a/fs/ntfs/aops.h
+++ b/fs/ntfs/aops.h
@@ -37,9 +37,9 @@ static inline void ntfs_unmap_page(struct page *page)
  * Read a page from the page cache of the address space @mapping at position
  * @index, where @index is in units of PAGE_SIZE, and not in bytes.
  *
- * If the page is not in memory it is loaded from disk first using the readpage
- * method defined in the address space operations of @mapping and the page is
- * added to the page cache of @mapping in the process.
+ * If the page is not in memory it is loaded from disk first using the
+ * read_folio method defined in the address space operations of @mapping
+ * and the page is added to the page cache of @mapping in the process.
  *
  * If the page belongs to an mst protected attribute and it is marked as such
  * in its ntfs inode (NInoMstProtected()) the mst fixups are applied but no
diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c
index 2911c04a33e0..4de597a83b88 100644
--- a/fs/ntfs/attrib.c
+++ b/fs/ntfs/attrib.c
@@ -1719,7 +1719,7 @@ int ntfs_attr_make_non_resident(ntfs_inode *ni, const u32 data_size)
 		vi->i_blocks = ni->allocated_size >> 9;
 	write_unlock_irqrestore(&ni->size_lock, flags);
 	/*
-	 * This needs to be last since the address space operations ->readpage
+	 * This needs to be last since the address space operations ->read_folio
 	 * and ->writepage can run concurrently with us as they are not
 	 * serialized on i_mutex.  Note, we are not allowed to fail once we flip
 	 * this switch, which is another reason to do this last.
diff --git a/fs/ntfs/file.c b/fs/ntfs/file.c
index 2ae25e48a41a..e1392a9b8ceb 100644
--- a/fs/ntfs/file.c
+++ b/fs/ntfs/file.c
@@ -251,14 +251,14 @@ static int ntfs_attr_extend_initialized(ntfs_inode *ni, const s64 new_init_size)
 		 *
 		 * TODO: For sparse pages could optimize this workload by using
 		 * the FsMisc / MiscFs page bit as a "PageIsSparse" bit.  This
-		 * would be set in readpage for sparse pages and here we would
+		 * would be set in read_folio for sparse pages and here we would
 		 * not need to mark dirty any pages which have this bit set.
 		 * The only caveat is that we have to clear the bit everywhere
 		 * where we allocate any clusters that lie in the page or that
 		 * contain the page.
 		 *
 		 * TODO: An even greater optimization would be for us to only
-		 * call readpage() on pages which are not in sparse regions as
+		 * call read_folio() on pages which are not in sparse regions as
 		 * determined from the runlist.  This would greatly reduce the
 		 * number of pages we read and make dirty in the case of sparse
 		 * files.
diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c
index efe0602b4e51..db0f1995aedd 100644
--- a/fs/ntfs/inode.c
+++ b/fs/ntfs/inode.c
@@ -1832,7 +1832,7 @@ int ntfs_read_inode_mount(struct inode *vi)
 	/* Need this to sanity check attribute list references to $MFT. */
 	vi->i_generation = ni->seq_no = le16_to_cpu(m->sequence_number);
 
-	/* Provides readpage() for map_mft_record(). */
+	/* Provides read_folio() for map_mft_record(). */
 	vi->i_mapping->a_ops = &ntfs_mst_aops;
 
 	ctx = ntfs_attr_get_search_ctx(ni, m);
@@ -2503,7 +2503,7 @@ int ntfs_truncate(struct inode *vi)
 		 * between the old data_size, i.e. old_size, and the new_size
 		 * has not been zeroed.  Fortunately, we do not need to zero it
 		 * either since on one hand it will either already be zero due
-		 * to both readpage and writepage clearing partial page data
+		 * to both read_folio and writepage clearing partial page data
 		 * beyond i_size in which case there is nothing to do or in the
 		 * case of the file being mmap()ped at the same time, POSIX
 		 * specifies that the behaviour is unspecified thus we do not
diff --git a/fs/ntfs/mft.h b/fs/ntfs/mft.h
index 17bfefc30271..49c001af16ed 100644
--- a/fs/ntfs/mft.h
+++ b/fs/ntfs/mft.h
@@ -79,7 +79,7 @@ extern int write_mft_record_nolock(ntfs_inode *ni, MFT_RECORD *m, int sync);
  * paths and via the page cache write back code paths or between writing
  * neighbouring mft records residing in the same page.
  *
- * Locking the page also serializes us against ->readpage() if the page is not
+ * Locking the page also serializes us against ->read_folio() if the page is not
  * uptodate.
  *
  * On success, clean the mft record and return 0.  On error, leave the mft
-- 
2.34.1

[PATCH 29/37] ocfs2: Convert ocfs2 to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ocfs2/alloc.c   | 2 +-
 fs/ocfs2/aops.c    | 5 +++--
 fs/ocfs2/file.c    | 2 +-
 fs/ocfs2/symlink.c | 5 +++--
 4 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c
index 49f41074baad..51c93929a146 100644
--- a/fs/ocfs2/alloc.c
+++ b/fs/ocfs2/alloc.c
@@ -7427,7 +7427,7 @@ int ocfs2_truncate_inline(struct inode *inode, struct buffer_head *di_bh,
 	/*
 	 * No need to worry about the data page here - it's been
 	 * truncated already and inline data doesn't need it for
-	 * pushing zero's to disk, so we'll let readpage pick it up
+	 * pushing zero's to disk, so we'll let read_folio pick it up
 	 * later.
 	 */
 	if (trunc) {
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 7bf4b6fd93bf..6b1679db9636 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -275,8 +275,9 @@ static int ocfs2_readpage_inline(struct inode *inode, struct page *page)
 	return ret;
 }
 
-static int ocfs2_readpage(struct file *file, struct page *page)
+static int ocfs2_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct ocfs2_inode_info *oi = OCFS2_I(inode);
 	loff_t start = (loff_t)page->index << PAGE_SHIFT;
@@ -2454,7 +2455,7 @@ static ssize_t ocfs2_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
 
 const struct address_space_operations ocfs2_aops = {
 	.dirty_folio		= block_dirty_folio,
-	.readpage		= ocfs2_readpage,
+	.read_folio		= ocfs2_read_folio,
 	.readahead		= ocfs2_readahead,
 	.writepage		= ocfs2_writepage,
 	.write_begin		= ocfs2_write_begin,
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 01b7407a8893..7497cd592258 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2526,7 +2526,7 @@ static ssize_t ocfs2_file_read_iter(struct kiocb *iocb,
 		return -EOPNOTSUPP;
 
 	/*
-	 * buffered reads protect themselves in ->readpage().  O_DIRECT reads
+	 * buffered reads protect themselves in ->read_folio().  O_DIRECT reads
 	 * need locks to protect pending reads from racing with truncate.
 	 */
 	if (direct_io) {
diff --git a/fs/ocfs2/symlink.c b/fs/ocfs2/symlink.c
index f755a4985821..d4c5fdcfa1e4 100644
--- a/fs/ocfs2/symlink.c
+++ b/fs/ocfs2/symlink.c
@@ -52,8 +52,9 @@
 #include "buffer_head_io.h"
 
 
-static int ocfs2_fast_symlink_readpage(struct file *unused, struct page *page)
+static int ocfs2_fast_symlink_read_folio(struct file *f, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct buffer_head *bh = NULL;
 	int status = ocfs2_read_inode_block(inode, &bh);
@@ -81,7 +82,7 @@ static int ocfs2_fast_symlink_readpage(struct file *unused, struct page *page)
 }
 
 const struct address_space_operations ocfs2_fast_symlink_aops = {
-	.readpage		= ocfs2_fast_symlink_readpage,
+	.read_folio		= ocfs2_fast_symlink_read_folio,
 };
 
 const struct inode_operations ocfs2_symlink_inode_operations = {
-- 
2.34.1

[PATCH 30/37] orangefs: Convert orangefs to read_folio

[Matthew Wilcox (Oracle)]

This is a full conversion which should be large folio ready, although
I have not tested it.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/orangefs/inode.c | 33 ++++++++++++++++-----------------
 1 file changed, 16 insertions(+), 17 deletions(-)

diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index bc7ccd15d7a3..241ac21f527b 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -288,40 +288,39 @@ static void orangefs_readahead(struct readahead_control *rac)
 	}
 }
 
-static int orangefs_readpage(struct file *file, struct page *page)
+static int orangefs_read_folio(struct file *file, struct folio *folio)
 {
-	struct folio *folio = page_folio(page);
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct iov_iter iter;
 	struct bio_vec bv;
 	ssize_t ret;
-	loff_t off; /* offset into this page */
+	loff_t off; /* offset of this folio in the file */
 
 	if (folio_test_dirty(folio))
 		orangefs_launder_folio(folio);
 
-	off = page_offset(page);
-	bv.bv_page = page;
-	bv.bv_len = PAGE_SIZE;
+	off = folio_pos(folio);
+	bv.bv_page = &folio->page;
+	bv.bv_len = folio_size(folio);
 	bv.bv_offset = 0;
-	iov_iter_bvec(&iter, READ, &bv, 1, PAGE_SIZE);
+	iov_iter_bvec(&iter, READ, &bv, 1, folio_size(folio));
 
 	ret = wait_for_direct_io(ORANGEFS_IO_READ, inode, &off, &iter,
-	    PAGE_SIZE, inode->i_size, NULL, NULL, file);
+			folio_size(folio), inode->i_size, NULL, NULL, file);
 	/* this will only zero remaining unread portions of the page data */
 	iov_iter_zero(~0U, &iter);
 	/* takes care of potential aliasing */
-	flush_dcache_page(page);
+	flush_dcache_folio(folio);
 	if (ret < 0) {
-		SetPageError(page);
+		folio_set_error(folio);
 	} else {
-		SetPageUptodate(page);
-		if (PageError(page))
-			ClearPageError(page);
+		folio_mark_uptodate(folio);
+		if (folio_test_error(folio))
+			folio_clear_error(folio);
 		ret = 0;
 	}
-	/* unlock the page after the ->readpage() routine completes */
-	unlock_page(page);
+	/* unlock the folio after the ->read_folio() routine completes */
+	folio_unlock(folio);
         return ret;
 }
 
@@ -631,7 +630,7 @@ static ssize_t orangefs_direct_IO(struct kiocb *iocb,
 static const struct address_space_operations orangefs_address_operations = {
 	.writepage = orangefs_writepage,
 	.readahead = orangefs_readahead,
-	.readpage = orangefs_readpage,
+	.read_folio = orangefs_read_folio,
 	.writepages = orangefs_writepages,
 	.dirty_folio = filemap_dirty_folio,
 	.write_begin = orangefs_write_begin,
-- 
2.34.1

[PATCH 31/37] romfs: Convert romfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/romfs/super.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/fs/romfs/super.c b/fs/romfs/super.c
index 9e6bbb4219de..c59b230d55b4 100644
--- a/fs/romfs/super.c
+++ b/fs/romfs/super.c
@@ -18,7 +18,7 @@
  *					Changed for 2.1.19 modules
  *	Jan 1997			Initial release
  *	Jun 1997			2.1.43+ changes
- *					Proper page locking in readpage
+ *					Proper page locking in read_folio
  *					Changed to work with 2.1.45+ fs
  *	Jul 1997			Fixed follow_link
  *			2.1.47
@@ -41,7 +41,7 @@
  *					  dentries in lookup
  *					clean up page flags setting
  *					  (error, uptodate, locking) in
- *					  in readpage
+ *					  in read_folio
  *					use init_special_inode for
  *					  fifos/sockets (and streamline) in
  *					  read_inode, fix _ops table order
@@ -99,8 +99,9 @@ static struct inode *romfs_iget(struct super_block *sb, unsigned long pos);
 /*
  * read a page worth of data from the image
  */
-static int romfs_readpage(struct file *file, struct page *page)
+static int romfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	loff_t offset, size;
 	unsigned long fillsize, pos;
@@ -142,7 +143,7 @@ static int romfs_readpage(struct file *file, struct page *page)
 }
 
 static const struct address_space_operations romfs_aops = {
-	.readpage	= romfs_readpage
+	.read_folio	= romfs_read_folio
 };
 
 /*
-- 
2.34.1

[PATCH 32/37] squashfs: Convert squashfs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/squashfs/file.c    | 5 +++--
 fs/squashfs/super.c   | 2 +-
 fs/squashfs/symlink.c | 5 +++--
 3 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/fs/squashfs/file.c b/fs/squashfs/file.c
index 89d492916dea..a8e495d8eb86 100644
--- a/fs/squashfs/file.c
+++ b/fs/squashfs/file.c
@@ -444,8 +444,9 @@ static int squashfs_readpage_sparse(struct page *page, int expected)
 	return 0;
 }
 
-static int squashfs_readpage(struct file *file, struct page *page)
+static int squashfs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct squashfs_sb_info *msblk = inode->i_sb->s_fs_info;
 	int index = page->index >> (msblk->block_log - PAGE_SHIFT);
@@ -496,5 +497,5 @@ static int squashfs_readpage(struct file *file, struct page *page)
 
 
 const struct address_space_operations squashfs_aops = {
-	.readpage = squashfs_readpage
+	.read_folio = squashfs_read_folio
 };
diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c
index 4f74abbc1a54..6d594ba2ed28 100644
--- a/fs/squashfs/super.c
+++ b/fs/squashfs/super.c
@@ -148,7 +148,7 @@ static int squashfs_fill_super(struct super_block *sb, struct fs_context *fc)
 
 	/*
 	 * squashfs provides 'backing_dev_info' in order to disable read-ahead. For
-	 * squashfs, I/O is not deferred, it is done immediately in readpage,
+	 * squashfs, I/O is not deferred, it is done immediately in read_folio,
 	 * which means the user would always have to wait their own I/O. So the effect
 	 * of readahead is very weak for squashfs. squashfs_bdi_init will set
 	 * sb->s_bdi->ra_pages and sb->s_bdi->io_pages to 0 and close readahead for
diff --git a/fs/squashfs/symlink.c b/fs/squashfs/symlink.c
index 1430613183e6..2bf977a52c2c 100644
--- a/fs/squashfs/symlink.c
+++ b/fs/squashfs/symlink.c
@@ -30,8 +30,9 @@
 #include "squashfs.h"
 #include "xattr.h"
 
-static int squashfs_symlink_readpage(struct file *file, struct page *page)
+static int squashfs_symlink_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct super_block *sb = inode->i_sb;
 	struct squashfs_sb_info *msblk = sb->s_fs_info;
@@ -101,7 +102,7 @@ static int squashfs_symlink_readpage(struct file *file, struct page *page)
 
 
 const struct address_space_operations squashfs_symlink_aops = {
-	.readpage = squashfs_symlink_readpage
+	.read_folio = squashfs_symlink_read_folio
 };
 
 const struct inode_operations squashfs_symlink_inode_ops = {
-- 
2.34.1

[PATCH 33/37] ubifs: Convert ubifs to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ubifs/file.c  | 12 +++++++-----
 fs/ubifs/super.c |  2 +-
 2 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 81c085c4decf..7cbf2edf8907 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -31,9 +31,9 @@
  * in the "sys_write -> alloc_pages -> direct reclaim path". So, in
  * 'ubifs_writepage()' we are only guaranteed that the page is locked.
  *
- * Similarly, @i_mutex is not always locked in 'ubifs_readpage()', e.g., the
+ * Similarly, @i_mutex is not always locked in 'ubifs_read_folio()', e.g., the
  * read-ahead path does not lock it ("sys_read -> generic_file_aio_read ->
- * ondemand_readahead -> readpage"). In case of readahead, @I_SYNC flag is not
+ * ondemand_readahead -> read_folio"). In case of readahead, @I_SYNC flag is not
  * set as well. However, UBIFS disables readahead.
  */
 
@@ -889,12 +889,14 @@ static int ubifs_bulk_read(struct page *page)
 	return err;
 }
 
-static int ubifs_readpage(struct file *file, struct page *page)
+static int ubifs_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
+
 	if (ubifs_bulk_read(page))
 		return 0;
 	do_readpage(page);
-	unlock_page(page);
+	folio_unlock(folio);
 	return 0;
 }
 
@@ -1641,7 +1643,7 @@ static int ubifs_symlink_getattr(struct user_namespace *mnt_userns,
 }
 
 const struct address_space_operations ubifs_file_address_operations = {
-	.readpage       = ubifs_readpage,
+	.read_folio     = ubifs_read_folio,
 	.writepage      = ubifs_writepage,
 	.write_begin    = ubifs_write_begin,
 	.write_end      = ubifs_write_end,
diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
index bad67455215f..0978d01b0ea4 100644
--- a/fs/ubifs/super.c
+++ b/fs/ubifs/super.c
@@ -2191,7 +2191,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent)
 
 	/*
 	 * UBIFS provides 'backing_dev_info' in order to disable read-ahead. For
-	 * UBIFS, I/O is not deferred, it is done immediately in readpage,
+	 * UBIFS, I/O is not deferred, it is done immediately in read_folio,
 	 * which means the user would have to wait not just for their own I/O
 	 * but the read-ahead I/O as well i.e. completely pointless.
 	 *
-- 
2.34.1

[PATCH 34/37] udf: Convert adinicb and symlinks to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/udf/file.c    | 10 +++++-----
 fs/udf/symlink.c |  5 +++--
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/fs/udf/file.c b/fs/udf/file.c
index 3f4d5c44c784..09aef77269fe 100644
--- a/fs/udf/file.c
+++ b/fs/udf/file.c
@@ -57,11 +57,11 @@ static void __udf_adinicb_readpage(struct page *page)
 	kunmap_atomic(kaddr);
 }
 
-static int udf_adinicb_readpage(struct file *file, struct page *page)
+static int udf_adinicb_read_folio(struct file *file, struct folio *folio)
 {
-	BUG_ON(!PageLocked(page));
-	__udf_adinicb_readpage(page);
-	unlock_page(page);
+	BUG_ON(!folio_test_locked(folio));
+	__udf_adinicb_readpage(&folio->page);
+	folio_unlock(folio);
 
 	return 0;
 }
@@ -127,7 +127,7 @@ static int udf_adinicb_write_end(struct file *file, struct address_space *mappin
 const struct address_space_operations udf_adinicb_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
-	.readpage	= udf_adinicb_readpage,
+	.read_folio	= udf_adinicb_read_folio,
 	.writepage	= udf_adinicb_writepage,
 	.write_begin	= udf_adinicb_write_begin,
 	.write_end	= udf_adinicb_write_end,
diff --git a/fs/udf/symlink.c b/fs/udf/symlink.c
index 9b223421a3c5..f3642f9c23f8 100644
--- a/fs/udf/symlink.c
+++ b/fs/udf/symlink.c
@@ -101,8 +101,9 @@ static int udf_pc_to_char(struct super_block *sb, unsigned char *from,
 	return 0;
 }
 
-static int udf_symlink_filler(struct file *file, struct page *page)
+static int udf_symlink_filler(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct inode *inode = page->mapping->host;
 	struct buffer_head *bh = NULL;
 	unsigned char *symlink;
@@ -183,7 +184,7 @@ static int udf_symlink_getattr(struct user_namespace *mnt_userns,
  * symlinks can't do much...
  */
 const struct address_space_operations udf_symlink_aops = {
-	.readpage		= udf_symlink_filler,
+	.read_folio		= udf_symlink_filler,
 };
 
 const struct inode_operations udf_symlink_inode_operations = {
-- 
2.34.1

[PATCH 35/37] vboxsf: Convert vboxsf to read_folio

[Matthew Wilcox (Oracle)]

This is a "weak" conversion which converts straight back to using pages.
A full conversion should be performed at some point, hopefully by
someone familiar with the filesystem.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/vboxsf/file.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/fs/vboxsf/file.c b/fs/vboxsf/file.c
index d74e0d336995..572aa1c43b37 100644
--- a/fs/vboxsf/file.c
+++ b/fs/vboxsf/file.c
@@ -225,8 +225,9 @@ const struct inode_operations vboxsf_reg_iops = {
 	.setattr = vboxsf_setattr
 };
 
-static int vboxsf_readpage(struct file *file, struct page *page)
+static int vboxsf_read_folio(struct file *file, struct folio *folio)
 {
+	struct page *page = &folio->page;
 	struct vboxsf_handle *sf_handle = file->private_data;
 	loff_t off = page_offset(page);
 	u32 nread = PAGE_SIZE;
@@ -352,7 +353,7 @@ static int vboxsf_write_end(struct file *file, struct address_space *mapping,
  * page and it does not call SetPageUptodate for partial writes.
  */
 const struct address_space_operations vboxsf_reg_aops = {
-	.readpage = vboxsf_readpage,
+	.read_folio = vboxsf_read_folio,
 	.writepage = vboxsf_writepage,
 	.dirty_folio = filemap_dirty_folio,
 	.write_begin = simple_write_begin,
-- 
2.34.1

[PATCH 36/37] mm: Convert swap_readpage to call read_folio instead of readpage

[Matthew Wilcox (Oracle)]

This commit is split out so it can be dropped when resolving
conflicts with Neil Brown's series to stop calling ->readpage in
the swap code.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 mm/page_io.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/page_io.c b/mm/page_io.c
index 89fbf3cae30f..1ae4be14f9d3 100644
--- a/mm/page_io.c
+++ b/mm/page_io.c
@@ -336,7 +336,7 @@ int swap_readpage(struct page *page, bool synchronous)
 		struct file *swap_file = sis->swap_file;
 		struct address_space *mapping = swap_file->f_mapping;
 
-		ret = mapping->a_ops->readpage(swap_file, page);
+		ret = mapping->a_ops->read_folio(swap_file, page_folio(page));
 		if (!ret)
 			count_vm_event(PSWPIN);
 		goto out;
-- 
2.34.1

[PATCH 37/37] mm,fs: Remove aops->readpage

[Matthew Wilcox (Oracle)]

With all implementations of aops->readpage converted to aops->read_folio,
we can stop checking whether it's set and remove the member from aops.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/btrfs/file.c         |  2 +-
 fs/buffer.c             |  5 +----
 fs/ceph/addr.c          |  2 +-
 include/linux/fs.h      |  3 +--
 kernel/events/uprobes.c |  5 ++---
 mm/filemap.c            | 11 ++++-------
 mm/memory.c             |  4 ++--
 mm/readahead.c          | 12 ++++--------
 mm/shmem.c              |  2 +-
 mm/swapfile.c           |  2 +-
 10 files changed, 18 insertions(+), 30 deletions(-)

diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 373df5ebaf8d..57fba5abb059 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -2402,7 +2402,7 @@ static int btrfs_file_mmap(struct file	*filp, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = filp->f_mapping;
 
-	if (!mapping->a_ops->readpage && !mapping->a_ops->read_folio)
+	if (!mapping->a_ops->read_folio)
 		return -ENOEXEC;
 
 	file_accessed(filp);
diff --git a/fs/buffer.c b/fs/buffer.c
index ec0c52c8848e..786ef5b98c80 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2827,10 +2827,7 @@ int nobh_truncate_page(struct address_space *mapping,
 
 	/* Ok, it's mapped. Make sure it's up-to-date */
 	if (!folio_test_uptodate(folio)) {
-		if (mapping->a_ops->read_folio)
-			err = mapping->a_ops->read_folio(NULL, folio);
-		else
-			err = mapping->a_ops->readpage(NULL, &folio->page);
+		err = mapping->a_ops->read_folio(NULL, folio);
 		if (err) {
 			folio_put(folio);
 			goto out;
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index be3e47784f08..e040b92bb17c 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -1772,7 +1772,7 @@ int ceph_mmap(struct file *file, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = file->f_mapping;
 
-	if (!mapping->a_ops->readpage && !mapping->a_ops->read_folio)
+	if (!mapping->a_ops->read_folio)
 		return -ENOEXEC;
 	file_accessed(file);
 	vma->vm_ops = &ceph_vmops;
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 5ad942183a2c..f812f5aa07dd 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -262,7 +262,7 @@ struct iattr {
  *  			trying again.  The aop will be taking reasonable
  *  			precautions not to livelock.  If the caller held a page
  *  			reference, it should drop it before retrying.  Returned
- *  			by readpage().
+ *  			by read_folio().
  *
  * address_space_operation functions return these large constants to indicate
  * special semantics to the caller.  These are much larger than the bytes in a
@@ -335,7 +335,6 @@ static inline bool is_sync_kiocb(struct kiocb *kiocb)
 
 struct address_space_operations {
 	int (*writepage)(struct page *page, struct writeback_control *wbc);
-	int (*readpage)(struct file *, struct page *);
 	int (*read_folio)(struct file *, struct folio *);
 
 	/* Write back some dirty pages from this mapping. */
diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index a340b1043efa..a9bc3c98f76a 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -787,10 +787,10 @@ static int __copy_insn(struct address_space *mapping, struct file *filp,
 	struct page *page;
 	/*
 	 * Ensure that the page that has the original instruction is populated
-	 * and in page-cache. If ->readpage == NULL it must be shmem_mapping(),
+	 * and in page-cache. If ->read_folio == NULL it must be shmem_mapping(),
 	 * see uprobe_register().
 	 */
-	if (mapping->ops->read_folio || mapping->a_ops->readpage)
+	if (mapping->a_ops->read_folio)
 		page = read_mapping_page(mapping, offset >> PAGE_SHIFT, filp);
 	else
 		page = shmem_read_mapping_page(mapping, offset >> PAGE_SHIFT);
@@ -1144,7 +1144,6 @@ static int __uprobe_register(struct inode *inode, loff_t offset,
 
 	/* copy_insn() uses read_mapping_page() or shmem_read_mapping_page() */
 	if (!inode->i_mapping->a_ops->read_folio &&
-	    !inode->i_mapping->a_ops->readpage &&
 	    !shmem_mapping(inode->i_mapping))
 		return -EIO;
 	/* Racy, just to catch the obvious mistakes */
diff --git a/mm/filemap.c b/mm/filemap.c
index 96e3d7ffd98e..6ede3ab76a5a 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -2414,15 +2414,12 @@ static int filemap_read_folio(struct file *file, struct address_space *mapping,
 
 	/*
 	 * A previous I/O error may have been due to temporary failures,
-	 * eg. multipath errors.  PG_error will be set again if readpage
+	 * eg. multipath errors.  PG_error will be set again if read_folio
 	 * fails.
 	 */
 	folio_clear_error(folio);
 	/* Start the actual read. The read will unlock the page. */
-	if (mapping->a_ops->read_folio)
-		error = mapping->a_ops->read_folio(file, folio);
-	else
-		error = mapping->a_ops->readpage(file, &folio->page);
+	error = mapping->a_ops->read_folio(file, folio);
 	if (error)
 		return error;
 
@@ -2639,7 +2636,7 @@ static int filemap_get_pages(struct kiocb *iocb, struct iov_iter *iter,
  * @already_read: Number of bytes already read by the caller.
  *
  * Copies data from the page cache.  If the data is not currently present,
- * uses the readahead and readpage address_space operations to fetch it.
+ * uses the readahead and read_folio address_space operations to fetch it.
  *
  * Return: Total number of bytes copied, including those already read by
  * the caller.  If an error happens before any bytes are copied, returns
@@ -3450,7 +3447,7 @@ int generic_file_mmap(struct file *file, struct vm_area_struct *vma)
 {
 	struct address_space *mapping = file->f_mapping;
 
-	if (!mapping->a_ops->read_folio && !mapping->a_ops->readpage)
+	if (!mapping->a_ops->read_folio)
 		return -ENOEXEC;
 	file_accessed(file);
 	vma->vm_ops = &generic_file_vm_ops;
diff --git a/mm/memory.c b/mm/memory.c
index 76e3af9639d9..2a12028a3749 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -555,11 +555,11 @@ static void print_bad_pte(struct vm_area_struct *vma, unsigned long addr,
 		dump_page(page, "bad pte");
 	pr_alert("addr:%px vm_flags:%08lx anon_vma:%px mapping:%px index:%lx\n",
 		 (void *)addr, vma->vm_flags, vma->anon_vma, mapping, index);
-	pr_alert("file:%pD fault:%ps mmap:%ps readpage:%ps\n",
+	pr_alert("file:%pD fault:%ps mmap:%ps read_folio:%ps\n",
 		 vma->vm_file,
 		 vma->vm_ops ? vma->vm_ops->fault : NULL,
 		 vma->vm_file ? vma->vm_file->f_op->mmap : NULL,
-		 mapping ? mapping->a_ops->readpage : NULL);
+		 mapping ? mapping->a_ops->read_folio : NULL);
 	dump_stack();
 	add_taint(TAINT_BAD_PAGE, LOCKDEP_NOW_UNRELIABLE);
 }
diff --git a/mm/readahead.c b/mm/readahead.c
index 76024c20a5a5..39983a3a93f0 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -170,12 +170,9 @@ static void read_pages(struct readahead_control *rac)
 			}
 			folio_unlock(folio);
 		}
-	} else if (aops->read_folio) {
-		while ((folio = readahead_folio(rac)) != NULL)
-			aops->read_folio(rac->file, folio);
 	} else {
 		while ((folio = readahead_folio(rac)) != NULL)
-			aops->readpage(rac->file, &folio->page);
+			aops->read_folio(rac->file, folio);
 	}
 
 	blk_finish_plug(&plug);
@@ -256,8 +253,8 @@ void page_cache_ra_unbounded(struct readahead_control *ractl,
 	}
 
 	/*
-	 * Now start the IO.  We ignore I/O errors - if the page is not
-	 * uptodate then the caller will launch readpage again, and
+	 * Now start the IO.  We ignore I/O errors - if the folio is not
+	 * uptodate then the caller will launch read_folio again, and
 	 * will then handle the error.
 	 */
 	read_pages(ractl);
@@ -305,8 +302,7 @@ void force_page_cache_ra(struct readahead_control *ractl,
 	struct backing_dev_info *bdi = inode_to_bdi(mapping->host);
 	unsigned long max_pages, index;
 
-	if (unlikely(!mapping->a_ops->read_folio &&
-		     !mapping->a_ops->readpage && !mapping->a_ops->readahead))
+	if (unlikely(!mapping->a_ops->read_folio && !mapping->a_ops->readahead))
 		return;
 
 	/*
diff --git a/mm/shmem.c b/mm/shmem.c
index 0f557a512171..f3e8de8ff75c 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -4162,7 +4162,7 @@ int shmem_zero_setup(struct vm_area_struct *vma)
  *
  * This behaves as a tmpfs "read_cache_page_gfp(mapping, index, gfp)",
  * with any new page allocations done using the specified allocation flags.
- * But read_cache_page_gfp() uses the ->readpage() method: which does not
+ * But read_cache_page_gfp() uses the ->read_folio() method: which does not
  * suit tmpfs, since it may have pages in swapcache, and needs to find those
  * for itself; although drivers/gpu/drm i915 and ttm rely upon this support.
  *
diff --git a/mm/swapfile.c b/mm/swapfile.c
index 7c19098b8b45..ecd45bdbad9b 100644
--- a/mm/swapfile.c
+++ b/mm/swapfile.c
@@ -3041,7 +3041,7 @@ SYSCALL_DEFINE2(swapon, const char __user *, specialfile, int, swap_flags)
 	/*
 	 * Read the swap header.
 	 */
-	if (!mapping->a_ops->read_folio && !mapping->a_ops->readpage) {
+	if (!mapping->a_ops->read_folio) {
 		error = -EINVAL;
 		goto bad_swap_unlock_inode;
 	}
-- 
2.34.1

[PATCH 0/4] Miscellaneous folio conversions

[Matthew Wilcox (Oracle)]

Mostly this is prep work for the following patches.

Matthew Wilcox (Oracle) (4):
  readahead: Use a folio in read_pages()
  fs: Convert is_dirty_writeback() to take a folio
  mm/readahead: Convert page_cache_async_readahead to take a folio
  buffer: Rewrite nobh_truncate_page() to use folios

 Documentation/filesystems/vfs.rst | 10 ++--
 fs/btrfs/relocation.c             |  5 +-
 fs/btrfs/send.c                   |  3 +-
 fs/buffer.c                       | 80 ++++++++++++++-----------------
 fs/nfs/file.c                     | 21 ++++----
 fs/verity/enable.c                | 29 ++++++-----
 include/linux/buffer_head.h       |  2 +-
 include/linux/fs.h                |  2 +-
 include/linux/pagemap.h           |  6 +--
 mm/readahead.c                    | 25 +++++-----
 mm/vmscan.c                       |  2 +-
 11 files changed, 86 insertions(+), 99 deletions(-)

-- 
2.34.1

[PATCH 1/4] readahead: Use a folio in read_pages()

[Matthew Wilcox (Oracle)]

Handle multi-page folios correctly and removes a few calls to
compound_head().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 mm/readahead.c | 25 ++++++++++++-------------
 1 file changed, 12 insertions(+), 13 deletions(-)

diff --git a/mm/readahead.c b/mm/readahead.c
index 4a60cdb64262..60a28af25c4e 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -145,7 +145,7 @@ EXPORT_SYMBOL_GPL(file_ra_state_init);
 static void read_pages(struct readahead_control *rac)
 {
 	const struct address_space_operations *aops = rac->mapping->a_ops;
-	struct page *page;
+	struct folio *folio;
 	struct blk_plug plug;
 
 	if (!readahead_count(rac))
@@ -156,24 +156,23 @@ static void read_pages(struct readahead_control *rac)
 	if (aops->readahead) {
 		aops->readahead(rac);
 		/*
-		 * Clean up the remaining pages.  The sizes in ->ra
+		 * Clean up the remaining folios.  The sizes in ->ra
 		 * may be used to size the next readahead, so make sure
 		 * they accurately reflect what happened.
 		 */
-		while ((page = readahead_page(rac))) {
-			rac->ra->size -= 1;
-			if (rac->ra->async_size > 0) {
-				rac->ra->async_size -= 1;
-				delete_from_page_cache(page);
+		while ((folio = readahead_folio(rac)) != NULL) {
+			unsigned long nr = folio_nr_pages(folio);
+
+			rac->ra->size -= nr;
+			if (rac->ra->async_size >= nr) {
+				rac->ra->async_size -= nr;
+				filemap_remove_folio(folio);
 			}
-			unlock_page(page);
-			put_page(page);
+			folio_unlock(folio);
 		}
 	} else {
-		while ((page = readahead_page(rac))) {
-			aops->readpage(rac->file, page);
-			put_page(page);
-		}
+		while ((folio = readahead_folio(rac)))
+			aops->readpage(rac->file, &folio->page);
 	}
 
 	blk_finish_plug(&plug);
-- 
2.34.1

[PATCH 2/4] fs: Convert is_dirty_writeback() to take a folio

[Matthew Wilcox (Oracle)]

Pass a folio instead of a page to aops->is_dirty_writeback().
Convert both implementations and the caller.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 Documentation/filesystems/vfs.rst | 10 +++++-----
 fs/buffer.c                       | 16 ++++++++--------
 fs/nfs/file.c                     | 21 +++++++++------------
 include/linux/buffer_head.h       |  2 +-
 include/linux/fs.h                |  2 +-
 mm/vmscan.c                       |  2 +-
 6 files changed, 25 insertions(+), 28 deletions(-)

diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
index 30f303180a7d..469882f72fc1 100644
--- a/Documentation/filesystems/vfs.rst
+++ b/Documentation/filesystems/vfs.rst
@@ -747,7 +747,7 @@ cache in your filesystem.  The following members are defined:
 
 		bool (*is_partially_uptodate) (struct folio *, size_t from,
 					       size_t count);
-		void (*is_dirty_writeback) (struct page *, bool *, bool *);
+		void (*is_dirty_writeback)(struct folio *, bool *, bool *);
 		int (*error_remove_page) (struct mapping *mapping, struct page *page);
 		int (*swap_activate)(struct file *);
 		int (*swap_deactivate)(struct file *);
@@ -932,14 +932,14 @@ cache in your filesystem.  The following members are defined:
 	without needing I/O to bring the whole page up to date.
 
 ``is_dirty_writeback``
-	Called by the VM when attempting to reclaim a page.  The VM uses
+	Called by the VM when attempting to reclaim a folio.  The VM uses
 	dirty and writeback information to determine if it needs to
 	stall to allow flushers a chance to complete some IO.
-	Ordinarily it can use PageDirty and PageWriteback but some
-	filesystems have more complex state (unstable pages in NFS
+	Ordinarily it can use folio_test_dirty and folio_test_writeback but
+	some filesystems have more complex state (unstable folios in NFS
 	prevent reclaim) or do not set those flags due to locking
 	problems.  This callback allows a filesystem to indicate to the
-	VM if a page should be treated as dirty or writeback for the
+	VM if a folio should be treated as dirty or writeback for the
 	purposes of stalling.
 
 ``error_remove_page``
diff --git a/fs/buffer.c b/fs/buffer.c
index d538495a0553..fb4df259c92d 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -79,26 +79,26 @@ void unlock_buffer(struct buffer_head *bh)
 EXPORT_SYMBOL(unlock_buffer);
 
 /*
- * Returns if the page has dirty or writeback buffers. If all the buffers
- * are unlocked and clean then the PageDirty information is stale. If
- * any of the pages are locked, it is assumed they are locked for IO.
+ * Returns if the folio has dirty or writeback buffers. If all the buffers
+ * are unlocked and clean then the folio_test_dirty information is stale. If
+ * any of the buffers are locked, it is assumed they are locked for IO.
  */
-void buffer_check_dirty_writeback(struct page *page,
+void buffer_check_dirty_writeback(struct folio *folio,
 				     bool *dirty, bool *writeback)
 {
 	struct buffer_head *head, *bh;
 	*dirty = false;
 	*writeback = false;
 
-	BUG_ON(!PageLocked(page));
+	BUG_ON(!folio_test_locked(folio));
 
-	if (!page_has_buffers(page))
+	head = folio_buffers(folio);
+	if (!head)
 		return;
 
-	if (PageWriteback(page))
+	if (folio_test_writeback(folio))
 		*writeback = true;
 
-	head = page_buffers(page);
 	bh = head;
 	do {
 		if (buffer_locked(bh))
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 314d2d7ba84a..f05c4b18b681 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -430,19 +430,16 @@ static int nfs_release_page(struct page *page, gfp_t gfp)
 	return nfs_fscache_release_page(page, gfp);
 }
 
-static void nfs_check_dirty_writeback(struct page *page,
+static void nfs_check_dirty_writeback(struct folio *folio,
 				bool *dirty, bool *writeback)
 {
 	struct nfs_inode *nfsi;
-	struct address_space *mapping = page_file_mapping(page);
-
-	if (!mapping || PageSwapCache(page))
-		return;
+	struct address_space *mapping = folio->mapping;
 
 	/*
-	 * Check if an unstable page is currently being committed and
-	 * if so, have the VM treat it as if the page is under writeback
-	 * so it will not block due to pages that will shortly be freeable.
+	 * Check if an unstable folio is currently being committed and
+	 * if so, have the VM treat it as if the folio is under writeback
+	 * so it will not block due to folios that will shortly be freeable.
 	 */
 	nfsi = NFS_I(mapping->host);
 	if (atomic_read(&nfsi->commit_info.rpcs_out)) {
@@ -451,11 +448,11 @@ static void nfs_check_dirty_writeback(struct page *page,
 	}
 
 	/*
-	 * If PagePrivate() is set, then the page is not freeable and as the
-	 * inode is not being committed, it's not going to be cleaned in the
-	 * near future so treat it as dirty
+	 * If the private flag is set, then the folio is not freeable
+	 * and as the inode is not being committed, it's not going to
+	 * be cleaned in the near future so treat it as dirty
 	 */
-	if (PagePrivate(page))
+	if (folio_test_private(folio))
 		*dirty = true;
 }
 
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 6e5a64005fef..805c4e12700a 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -146,7 +146,7 @@ BUFFER_FNS(Defer_Completion, defer_completion)
 #define page_has_buffers(page)	PagePrivate(page)
 #define folio_buffers(folio)		folio_get_private(folio)
 
-void buffer_check_dirty_writeback(struct page *page,
+void buffer_check_dirty_writeback(struct folio *folio,
 				     bool *dirty, bool *writeback);
 
 /*
diff --git a/include/linux/fs.h b/include/linux/fs.h
index b35ce086a7a1..2be852661a29 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -369,7 +369,7 @@ struct address_space_operations {
 	int (*launder_folio)(struct folio *);
 	bool (*is_partially_uptodate) (struct folio *, size_t from,
 			size_t count);
-	void (*is_dirty_writeback) (struct page *, bool *, bool *);
+	void (*is_dirty_writeback) (struct folio *, bool *dirty, bool *wb);
 	int (*error_remove_page)(struct address_space *, struct page *);
 
 	/* swapfile support */
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 1678802e03e7..27851232e00c 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1451,7 +1451,7 @@ static void folio_check_dirty_writeback(struct folio *folio,
 
 	mapping = folio_mapping(folio);
 	if (mapping && mapping->a_ops->is_dirty_writeback)
-		mapping->a_ops->is_dirty_writeback(&folio->page, dirty, writeback);
+		mapping->a_ops->is_dirty_writeback(folio, dirty, writeback);
 }
 
 static struct page *alloc_demote_page(struct page *page, unsigned long node)
-- 
2.34.1

[PATCH 3/4] mm/readahead: Convert page_cache_async_readahead to take a folio

[Matthew Wilcox (Oracle)]

Removes a couple of calls to compound_head and saves a few bytes.
Also convert verity's read_file_data_page() to be folio-based.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/btrfs/relocation.c   |  5 +++--
 fs/btrfs/send.c         |  3 ++-
 fs/verity/enable.c      | 29 ++++++++++++++---------------
 include/linux/pagemap.h |  6 +++---
 4 files changed, 22 insertions(+), 21 deletions(-)

diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c
index fdc2c4b411f0..9ae06895ffc9 100644
--- a/fs/btrfs/relocation.c
+++ b/fs/btrfs/relocation.c
@@ -2967,8 +2967,9 @@ static int relocate_one_page(struct inode *inode, struct file_ra_state *ra,
 		goto release_page;
 
 	if (PageReadahead(page))
-		page_cache_async_readahead(inode->i_mapping, ra, NULL, page,
-				   page_index, last_index + 1 - page_index);
+		page_cache_async_readahead(inode->i_mapping, ra, NULL,
+				page_folio(page), page_index,
+				last_index + 1 - page_index);
 
 	if (!PageUptodate(page)) {
 		btrfs_readpage(NULL, page);
diff --git a/fs/btrfs/send.c b/fs/btrfs/send.c
index 7d1642937274..b327dbe0cbf5 100644
--- a/fs/btrfs/send.c
+++ b/fs/btrfs/send.c
@@ -4986,7 +4986,8 @@ static int put_file_data(struct send_ctx *sctx, u64 offset, u32 len)
 
 		if (PageReadahead(page)) {
 			page_cache_async_readahead(inode->i_mapping, &sctx->ra,
-				NULL, page, index, last_index + 1 - index);
+						NULL, page_folio(page), index,
+						last_index + 1 - index);
 		}
 
 		if (!PageUptodate(page)) {
diff --git a/fs/verity/enable.c b/fs/verity/enable.c
index 60a4372aa4d7..f75d2c010f36 100644
--- a/fs/verity/enable.c
+++ b/fs/verity/enable.c
@@ -18,27 +18,26 @@
  * Read a file data page for Merkle tree construction.  Do aggressive readahead,
  * since we're sequentially reading the entire file.
  */
-static struct page *read_file_data_page(struct file *filp, pgoff_t index,
+static struct page *read_file_data_page(struct file *file, pgoff_t index,
 					struct file_ra_state *ra,
 					unsigned long remaining_pages)
 {
-	struct page *page;
+	DEFINE_READAHEAD(ractl, file, ra, file->f_mapping, index);
+	struct folio *folio;
 
-	page = find_get_page_flags(filp->f_mapping, index, FGP_ACCESSED);
-	if (!page || !PageUptodate(page)) {
-		if (page)
-			put_page(page);
+	folio = __filemap_get_folio(ractl.mapping, index, FGP_ACCESSED, 0);
+	if (!folio || !folio_test_uptodate(folio)) {
+		if (folio)
+			folio_put(folio);
 		else
-			page_cache_sync_readahead(filp->f_mapping, ra, filp,
-						  index, remaining_pages);
-		page = read_mapping_page(filp->f_mapping, index, NULL);
-		if (IS_ERR(page))
-			return page;
+			page_cache_sync_ra(&ractl, remaining_pages);
+		folio = read_cache_folio(ractl.mapping, index, NULL, file);
+		if (IS_ERR(folio))
+			return &folio->page;
 	}
-	if (PageReadahead(page))
-		page_cache_async_readahead(filp->f_mapping, ra, filp, page,
-					   index, remaining_pages);
-	return page;
+	if (folio_test_readahead(folio))
+		page_cache_async_ra(&ractl, folio, remaining_pages);
+	return folio_file_page(folio, index);
 }
 
 static int build_merkle_tree_level(struct file *filp, unsigned int level,
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index 60657132080f..b70192f56454 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -1242,7 +1242,7 @@ void page_cache_sync_readahead(struct address_space *mapping,
  * @mapping: address_space which holds the pagecache and I/O vectors
  * @ra: file_ra_state which holds the readahead state
  * @file: Used by the filesystem for authentication.
- * @page: The page at @index which triggered the readahead call.
+ * @folio: The folio at @index which triggered the readahead call.
  * @index: Index of first page to be read.
  * @req_count: Total number of pages being read by the caller.
  *
@@ -1254,10 +1254,10 @@ void page_cache_sync_readahead(struct address_space *mapping,
 static inline
 void page_cache_async_readahead(struct address_space *mapping,
 		struct file_ra_state *ra, struct file *file,
-		struct page *page, pgoff_t index, unsigned long req_count)
+		struct folio *folio, pgoff_t index, unsigned long req_count)
 {
 	DEFINE_READAHEAD(ractl, file, ra, mapping, index);
-	page_cache_async_ra(&ractl, page_folio(page), req_count);
+	page_cache_async_ra(&ractl, folio, req_count);
 }
 
 static inline struct folio *__readahead_folio(struct readahead_control *ractl)
-- 
2.34.1

[PATCH 4/4] buffer: Rewrite nobh_truncate_page() to use folios

[Matthew Wilcox (Oracle)]

 - Calculate iblock directly instead of using a while loop
 - Move has_buffers to the end to remove a backwards jump
 - Use __filemap_get_folio() instead of grab_cache_page(), which
   removes a spurious FGP_ACCESSED flag.
 - Eliminate length and pos variables
 - Use folio APIs where they exist

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/buffer.c | 64 ++++++++++++++++++++++-------------------------------
 1 file changed, 27 insertions(+), 37 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index fb4df259c92d..9737e0dbe3ec 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -2791,44 +2791,28 @@ int nobh_truncate_page(struct address_space *mapping,
 			loff_t from, get_block_t *get_block)
 {
 	pgoff_t index = from >> PAGE_SHIFT;
-	unsigned offset = from & (PAGE_SIZE-1);
-	unsigned blocksize;
-	sector_t iblock;
-	unsigned length, pos;
 	struct inode *inode = mapping->host;
-	struct page *page;
+	unsigned blocksize = i_blocksize(inode);
+	struct folio *folio;
 	struct buffer_head map_bh;
+	size_t offset;
+	sector_t iblock;
 	int err;
 
-	blocksize = i_blocksize(inode);
-	length = offset & (blocksize - 1);
-
 	/* Block boundary? Nothing to do */
-	if (!length)
+	if (!(from & (blocksize - 1)))
 		return 0;
 
-	length = blocksize - length;
-	iblock = (sector_t)index << (PAGE_SHIFT - inode->i_blkbits);
-
-	page = grab_cache_page(mapping, index);
+	folio = __filemap_get_folio(mapping, index, FGP_LOCK | FGP_CREAT,
+			mapping_gfp_mask(mapping));
 	err = -ENOMEM;
-	if (!page)
+	if (!folio)
 		goto out;
 
-	if (page_has_buffers(page)) {
-has_buffers:
-		unlock_page(page);
-		put_page(page);
-		return block_truncate_page(mapping, from, get_block);
-	}
-
-	/* Find the buffer that contains "offset" */
-	pos = blocksize;
-	while (offset >= pos) {
-		iblock++;
-		pos += blocksize;
-	}
+	if (folio_buffers(folio))
+		goto has_buffers;
 
+	iblock = from >> inode->i_blkbits;
 	map_bh.b_size = blocksize;
 	map_bh.b_state = 0;
 	err = get_block(inode, iblock, &map_bh, 0);
@@ -2839,29 +2823,35 @@ int nobh_truncate_page(struct address_space *mapping,
 		goto unlock;
 
 	/* Ok, it's mapped. Make sure it's up-to-date */
-	if (!PageUptodate(page)) {
-		err = mapping->a_ops->readpage(NULL, page);
+	if (!folio_test_uptodate(folio)) {
+		err = mapping->a_ops->readpage(NULL, &folio->page);
 		if (err) {
-			put_page(page);
+			folio_put(folio);
 			goto out;
 		}
-		lock_page(page);
-		if (!PageUptodate(page)) {
+		folio_lock(folio);
+		if (!folio_test_uptodate(folio)) {
 			err = -EIO;
 			goto unlock;
 		}
-		if (page_has_buffers(page))
+		if (folio_buffers(folio))
 			goto has_buffers;
 	}
-	zero_user(page, offset, length);
-	set_page_dirty(page);
+	offset = offset_in_folio(folio, from);
+	folio_zero_segment(folio, offset, round_up(offset, blocksize));
+	folio_mark_dirty(folio);
 	err = 0;
 
 unlock:
-	unlock_page(page);
-	put_page(page);
+	folio_unlock(folio);
+	folio_put(folio);
 out:
 	return err;
+
+has_buffers:
+	folio_unlock(folio);
+	folio_put(folio);
+	return block_truncate_page(mapping, from, get_block);
 }
 EXPORT_SYMBOL(nobh_truncate_page);
 
-- 
2.34.1

[PATCH 00/26] Convert aops->releasepage to aops->release_folio

[Matthew Wilcox (Oracle)]

Convert all filesystems.  The last five patches are further cleanups
that are enabled by this change.

Matthew Wilcox (Oracle) (26):
  fs: Add aops->release_folio
  iomap: Convert to release_folio
  9p: Convert to release_folio
  afs: Convert to release_folio
  btrfs: Convert to release_folio
  ceph: Convert to release_folio
  cifs: Convert to release_folio
  erofs: Convert to release_folio
  ext4: Convert to release_folio
  f2fs: Convert to release_folio
  gfs2: Convert to release_folio
  hfs: Convert to release_folio
  hfsplus: Convert to release_folio
  jfs: Convert to release_folio
  nfs: Convert to release_folio
  nilfs2: Remove comment about releasepage
  ocfs2: Convert to release_folio
  orangefs: Convert to release_folio
  reiserfs: Convert to release_folio
  ubifs: Convert to release_folio
  fs: Remove last vestiges of releasepage
  reiserfs: Convert release_buffer_page() to use a folio
  jbd2: Convert jbd2_journal_try_to_free_buffers to take a folio
  jbd2: Convert release_buffer_page() to use a folio
  fs: Change try_to_free_buffers() to take a folio
  fs: Convert drop_buffers() to use a folio

 .../filesystems/caching/netfs-api.rst         |  4 +-
 Documentation/filesystems/locking.rst         | 14 ++---
 Documentation/filesystems/vfs.rst             | 45 ++++++++--------
 fs/9p/vfs_addr.c                              | 17 +++---
 fs/afs/dir.c                                  |  7 ++-
 fs/afs/file.c                                 | 11 ++--
 fs/afs/internal.h                             |  2 +-
 fs/btrfs/disk-io.c                            | 12 ++---
 fs/btrfs/extent_io.c                          | 14 ++---
 fs/btrfs/file.c                               |  2 +-
 fs/btrfs/inode.c                              | 24 ++++-----
 fs/buffer.c                                   | 54 +++++++++----------
 fs/ceph/addr.c                                | 24 ++++-----
 fs/cifs/file.c                                | 14 ++---
 fs/erofs/super.c                              | 16 +++---
 fs/ext4/inode.c                               | 20 +++----
 fs/f2fs/checkpoint.c                          |  2 +-
 fs/f2fs/compress.c                            |  2 +-
 fs/f2fs/data.c                                | 32 +++++------
 fs/f2fs/f2fs.h                                |  2 +-
 fs/f2fs/node.c                                |  2 +-
 fs/gfs2/aops.c                                | 44 +++++++--------
 fs/gfs2/inode.h                               |  2 +-
 fs/gfs2/meta_io.c                             |  4 +-
 fs/hfs/inode.c                                | 23 ++++----
 fs/hfsplus/inode.c                            | 23 ++++----
 fs/iomap/buffered-io.c                        | 22 ++++----
 fs/iomap/trace.h                              |  2 +-
 fs/jbd2/commit.c                              | 14 ++---
 fs/jbd2/transaction.c                         | 14 ++---
 fs/jfs/jfs_metapage.c                         | 16 +++---
 fs/mpage.c                                    |  2 +-
 fs/nfs/file.c                                 | 22 ++++----
 fs/nfs/fscache.h                              | 14 ++---
 fs/nilfs2/inode.c                             |  1 -
 fs/ocfs2/aops.c                               | 10 ++--
 fs/orangefs/inode.c                           |  6 +--
 fs/reiserfs/inode.c                           | 20 +++----
 fs/reiserfs/journal.c                         | 14 ++---
 fs/ubifs/file.c                               | 18 +++----
 fs/xfs/xfs_aops.c                             |  2 +-
 fs/zonefs/super.c                             |  2 +-
 include/linux/buffer_head.h                   |  4 +-
 include/linux/fs.h                            |  2 +-
 include/linux/iomap.h                         |  2 +-
 include/linux/jbd2.h                          |  2 +-
 include/linux/page-flags.h                    |  2 +-
 include/linux/pagemap.h                       |  4 --
 mm/filemap.c                                  |  6 +--
 mm/migrate.c                                  |  2 +-
 mm/vmscan.c                                   |  2 +-
 51 files changed, 309 insertions(+), 312 deletions(-)

-- 
2.34.1

[PATCH 01/26] fs: Add aops->release_folio

[Matthew Wilcox (Oracle)]

This replaces aops->releasepage.  Update the documentation, and call it
if it exists.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
---
 .../filesystems/caching/netfs-api.rst         |  4 +-
 Documentation/filesystems/locking.rst         | 14 +++---
 Documentation/filesystems/vfs.rst             | 45 +++++++++----------
 include/linux/fs.h                            |  1 +
 mm/filemap.c                                  |  2 +
 5 files changed, 34 insertions(+), 32 deletions(-)

diff --git a/Documentation/filesystems/caching/netfs-api.rst b/Documentation/filesystems/caching/netfs-api.rst
index 7308d76a29dc..1d18e9def183 100644
--- a/Documentation/filesystems/caching/netfs-api.rst
+++ b/Documentation/filesystems/caching/netfs-api.rst
@@ -433,11 +433,11 @@ has done a write and then the page it wrote from has been released by the VM,
 after which it *has* to look in the cache.
 
 To inform fscache that a page might now be in the cache, the following function
-should be called from the ``releasepage`` address space op::
+should be called from the ``release_folio`` address space op::
 
 	void fscache_note_page_release(struct fscache_cookie *cookie);
 
-if the page has been released (ie. releasepage returned true).
+if the page has been released (ie. release_folio returned true).
 
 Page release and page invalidation should also wait for any mark left on the
 page to say that a DIO write is underway from that page::
diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
index aeba2475a53c..2a295bb72dbc 100644
--- a/Documentation/filesystems/locking.rst
+++ b/Documentation/filesystems/locking.rst
@@ -249,7 +249,7 @@ prototypes::
 				struct page *page, void *fsdata);
 	sector_t (*bmap)(struct address_space *, sector_t);
 	void (*invalidate_folio) (struct folio *, size_t start, size_t len);
-	int (*releasepage) (struct page *, int);
+	int (*release_folio)(struct folio *, gfp_t);
 	void (*freepage)(struct page *);
 	int (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 	bool (*isolate_page) (struct page *, isolate_mode_t);
@@ -270,13 +270,13 @@ ops			PageLocked(page)	 i_rwsem	invalidate_lock
 writepage:		yes, unlocks (see below)
 read_folio:		yes, unlocks				shared
 writepages:
-dirty_folio		maybe
+dirty_folio:		maybe
 readahead:		yes, unlocks				shared
 write_begin:		locks the page		 exclusive
 write_end:		yes, unlocks		 exclusive
 bmap:
 invalidate_folio:	yes					exclusive
-releasepage:		yes
+release_folio:		yes
 freepage:		yes
 direct_IO:
 isolate_page:		yes
@@ -372,10 +372,10 @@ invalidate_lock before invalidating page cache in truncate / hole punch
 path (and thus calling into ->invalidate_folio) to block races between page
 cache invalidation and page cache filling functions (fault, read, ...).
 
-->releasepage() is called when the kernel is about to try to drop the
-buffers from the page in preparation for freeing it.  It returns zero to
-indicate that the buffers are (or may be) freeable.  If ->releasepage is zero,
-the kernel assumes that the fs has no private interest in the buffers.
+->release_folio() is called when the kernel is about to try to drop the
+buffers from the folio in preparation for freeing it.  It returns false to
+indicate that the buffers are (or may be) freeable.  If ->release_folio is
+NULL, the kernel assumes that the fs has no private interest in the buffers.
 
 ->freepage() is called when the kernel is done dropping the page
 from the page cache.
diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
index 0919a4ad973a..679887b5c8fc 100644
--- a/Documentation/filesystems/vfs.rst
+++ b/Documentation/filesystems/vfs.rst
@@ -620,9 +620,9 @@ Writeback.
 The first can be used independently to the others.  The VM can try to
 either write dirty pages in order to clean them, or release clean pages
 in order to reuse them.  To do this it can call the ->writepage method
-on dirty pages, and ->releasepage on clean pages with PagePrivate set.
-Clean pages without PagePrivate and with no external references will be
-released without notice being given to the address_space.
+on dirty pages, and ->release_folio on clean folios with the private
+flag set.  Clean pages without PagePrivate and with no external references
+will be released without notice being given to the address_space.
 
 To achieve this functionality, pages need to be placed on an LRU with
 lru_cache_add and mark_page_active needs to be called whenever the page
@@ -734,7 +734,7 @@ cache in your filesystem.  The following members are defined:
 				 struct page *page, void *fsdata);
 		sector_t (*bmap)(struct address_space *, sector_t);
 		void (*invalidate_folio) (struct folio *, size_t start, size_t len);
-		int (*releasepage) (struct page *, int);
+		bool (*release_folio)(struct folio *, gfp_t);
 		void (*freepage)(struct page *);
 		ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 		/* isolate a page for migration */
@@ -864,33 +864,32 @@ cache in your filesystem.  The following members are defined:
 	address space.  This generally corresponds to either a
 	truncation, punch hole or a complete invalidation of the address
 	space (in the latter case 'offset' will always be 0 and 'length'
-	will be folio_size()).  Any private data associated with the page
+	will be folio_size()).  Any private data associated with the folio
 	should be updated to reflect this truncation.  If offset is 0
 	and length is folio_size(), then the private data should be
-	released, because the page must be able to be completely
-	discarded.  This may be done by calling the ->releasepage
+	released, because the folio must be able to be completely
+	discarded.  This may be done by calling the ->release_folio
 	function, but in this case the release MUST succeed.
 
-``releasepage``
-	releasepage is called on PagePrivate pages to indicate that the
-	page should be freed if possible.  ->releasepage should remove
-	any private data from the page and clear the PagePrivate flag.
-	If releasepage() fails for some reason, it must indicate failure
-	with a 0 return value.  releasepage() is used in two distinct
-	though related cases.  The first is when the VM finds a clean
-	page with no active users and wants to make it a free page.  If
-	->releasepage succeeds, the page will be removed from the
-	address_space and become free.
+``release_folio``
+	release_folio is called on folios with private data to tell the
+	filesystem that the folio is about to be freed.  ->release_folio
+	should remove any private data from the folio and clear the
+	private flag.  If release_folio() fails, it should return false.
+	release_folio() is used in two distinct though related cases.
+	The first is when the VM wants to free a clean folio with no
+	active users.  If ->release_folio succeeds, the folio will be
+	removed from the address_space and be freed.
 
 	The second case is when a request has been made to invalidate
-	some or all pages in an address_space.  This can happen through
-	the fadvise(POSIX_FADV_DONTNEED) system call or by the
-	filesystem explicitly requesting it as nfs and 9fs do (when they
+	some or all folios in an address_space.  This can happen
+	through the fadvise(POSIX_FADV_DONTNEED) system call or by the
+	filesystem explicitly requesting it as nfs and 9p do (when they
 	believe the cache may be out of date with storage) by calling
 	invalidate_inode_pages2().  If the filesystem makes such a call,
-	and needs to be certain that all pages are invalidated, then its
-	releasepage will need to ensure this.  Possibly it can clear the
-	PageUptodate bit if it cannot free private data yet.
+	and needs to be certain that all folios are invalidated, then
+	its release_folio will need to ensure this.  Possibly it can
+	clear the uptodate flag if it cannot free private data yet.
 
 ``freepage``
 	freepage is called once the page is no longer visible in the
diff --git a/include/linux/fs.h b/include/linux/fs.h
index f812f5aa07dd..ad768f13f485 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -355,6 +355,7 @@ struct address_space_operations {
 	/* Unfortunately this kludge is needed for FIBMAP. Don't use it */
 	sector_t (*bmap)(struct address_space *, sector_t);
 	void (*invalidate_folio) (struct folio *, size_t offset, size_t len);
+	bool (*release_folio)(struct folio *, gfp_t);
 	int (*releasepage) (struct page *, gfp_t);
 	void (*freepage)(struct page *);
 	ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
diff --git a/mm/filemap.c b/mm/filemap.c
index 9b7fa47feb5e..78e4a7dc3a56 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3955,6 +3955,8 @@ bool filemap_release_folio(struct folio *folio, gfp_t gfp)
 	if (folio_test_writeback(folio))
 		return false;
 
+	if (mapping && mapping->a_ops->release_folio)
+		return mapping->a_ops->release_folio(folio, gfp);
 	if (mapping && mapping->a_ops->releasepage)
 		return mapping->a_ops->releasepage(&folio->page, gfp);
 	return try_to_free_buffers(&folio->page);
-- 
2.34.1

Re: [PATCH 01/26] fs: Add aops->release_folio

[Jeff Layton]

On Sun, 2022-05-08 at 21:32 +0100, Matthew Wilcox (Oracle) wrote:
> This replaces aops->releasepage.  Update the documentation, and call it
> if it exists.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> Reviewed-by: Jeff Layton <jlayton@kernel.org>
> ---
>  .../filesystems/caching/netfs-api.rst         |  4 +-
>  Documentation/filesystems/locking.rst         | 14 +++---
>  Documentation/filesystems/vfs.rst             | 45 +++++++++----------
>  include/linux/fs.h                            |  1 +
>  mm/filemap.c                                  |  2 +
>  5 files changed, 34 insertions(+), 32 deletions(-)
> 
> diff --git a/Documentation/filesystems/caching/netfs-api.rst b/Documentation/filesystems/caching/netfs-api.rst
> index 7308d76a29dc..1d18e9def183 100644
> --- a/Documentation/filesystems/caching/netfs-api.rst
> +++ b/Documentation/filesystems/caching/netfs-api.rst
> @@ -433,11 +433,11 @@ has done a write and then the page it wrote from has been released by the VM,
>  after which it *has* to look in the cache.
>  
>  To inform fscache that a page might now be in the cache, the following function
> -should be called from the ``releasepage`` address space op::
> +should be called from the ``release_folio`` address space op::
>  
>  	void fscache_note_page_release(struct fscache_cookie *cookie);
>  
> -if the page has been released (ie. releasepage returned true).
> +if the page has been released (ie. release_folio returned true).
>  
>  Page release and page invalidation should also wait for any mark left on the
>  page to say that a DIO write is underway from that page::
> diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
> index aeba2475a53c..2a295bb72dbc 100644
> --- a/Documentation/filesystems/locking.rst
> +++ b/Documentation/filesystems/locking.rst
> @@ -249,7 +249,7 @@ prototypes::
>  				struct page *page, void *fsdata);
>  	sector_t (*bmap)(struct address_space *, sector_t);
>  	void (*invalidate_folio) (struct folio *, size_t start, size_t len);
> -	int (*releasepage) (struct page *, int);
> +	int (*release_folio)(struct folio *, gfp_t);

Shouldn't that be a bool return?

>  	void (*freepage)(struct page *);
>  	int (*direct_IO)(struct kiocb *, struct iov_iter *iter);
>  	bool (*isolate_page) (struct page *, isolate_mode_t);
> @@ -270,13 +270,13 @@ ops			PageLocked(page)	 i_rwsem	invalidate_lock
>  writepage:		yes, unlocks (see below)
>  read_folio:		yes, unlocks				shared
>  writepages:
> -dirty_folio		maybe
> +dirty_folio:		maybe
>  readahead:		yes, unlocks				shared
>  write_begin:		locks the page		 exclusive
>  write_end:		yes, unlocks		 exclusive
>  bmap:
>  invalidate_folio:	yes					exclusive
> -releasepage:		yes
> +release_folio:		yes
>  freepage:		yes
>  direct_IO:
>  isolate_page:		yes
> @@ -372,10 +372,10 @@ invalidate_lock before invalidating page cache in truncate / hole punch
>  path (and thus calling into ->invalidate_folio) to block races between page
>  cache invalidation and page cache filling functions (fault, read, ...).
>  
> -->releasepage() is called when the kernel is about to try to drop the
> -buffers from the page in preparation for freeing it.  It returns zero to
> -indicate that the buffers are (or may be) freeable.  If ->releasepage is zero,
> -the kernel assumes that the fs has no private interest in the buffers.
> +->release_folio() is called when the kernel is about to try to drop the
> +buffers from the folio in preparation for freeing it.  It returns false to
> +indicate that the buffers are (or may be) freeable.  If ->release_folio is
> +NULL, the kernel assumes that the fs has no private interest in the buffers.
>  
>  ->freepage() is called when the kernel is done dropping the page
>  from the page cache.
> diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
> index 0919a4ad973a..679887b5c8fc 100644
> --- a/Documentation/filesystems/vfs.rst
> +++ b/Documentation/filesystems/vfs.rst
> @@ -620,9 +620,9 @@ Writeback.
>  The first can be used independently to the others.  The VM can try to
>  either write dirty pages in order to clean them, or release clean pages
>  in order to reuse them.  To do this it can call the ->writepage method
> -on dirty pages, and ->releasepage on clean pages with PagePrivate set.
> -Clean pages without PagePrivate and with no external references will be
> -released without notice being given to the address_space.
> +on dirty pages, and ->release_folio on clean folios with the private
> +flag set.  Clean pages without PagePrivate and with no external references
> +will be released without notice being given to the address_space.
>  
>  To achieve this functionality, pages need to be placed on an LRU with
>  lru_cache_add and mark_page_active needs to be called whenever the page
> @@ -734,7 +734,7 @@ cache in your filesystem.  The following members are defined:
>  				 struct page *page, void *fsdata);
>  		sector_t (*bmap)(struct address_space *, sector_t);
>  		void (*invalidate_folio) (struct folio *, size_t start, size_t len);
> -		int (*releasepage) (struct page *, int);
> +		bool (*release_folio)(struct folio *, gfp_t);
>  		void (*freepage)(struct page *);
>  		ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
>  		/* isolate a page for migration */
> @@ -864,33 +864,32 @@ cache in your filesystem.  The following members are defined:
>  	address space.  This generally corresponds to either a
>  	truncation, punch hole or a complete invalidation of the address
>  	space (in the latter case 'offset' will always be 0 and 'length'
> -	will be folio_size()).  Any private data associated with the page
> +	will be folio_size()).  Any private data associated with the folio
>  	should be updated to reflect this truncation.  If offset is 0
>  	and length is folio_size(), then the private data should be
> -	released, because the page must be able to be completely
> -	discarded.  This may be done by calling the ->releasepage
> +	released, because the folio must be able to be completely
> +	discarded.  This may be done by calling the ->release_folio
>  	function, but in this case the release MUST succeed.
>  
> -``releasepage``
> -	releasepage is called on PagePrivate pages to indicate that the
> -	page should be freed if possible.  ->releasepage should remove
> -	any private data from the page and clear the PagePrivate flag.
> -	If releasepage() fails for some reason, it must indicate failure
> -	with a 0 return value.  releasepage() is used in two distinct
> -	though related cases.  The first is when the VM finds a clean
> -	page with no active users and wants to make it a free page.  If
> -	->releasepage succeeds, the page will be removed from the
> -	address_space and become free.
> +``release_folio``
> +	release_folio is called on folios with private data to tell the
> +	filesystem that the folio is about to be freed.  ->release_folio
> +	should remove any private data from the folio and clear the
> +	private flag.  If release_folio() fails, it should return false.
> +	release_folio() is used in two distinct though related cases.
> +	The first is when the VM wants to free a clean folio with no
> +	active users.  If ->release_folio succeeds, the folio will be
> +	removed from the address_space and be freed.
>  
>  	The second case is when a request has been made to invalidate
> -	some or all pages in an address_space.  This can happen through
> -	the fadvise(POSIX_FADV_DONTNEED) system call or by the
> -	filesystem explicitly requesting it as nfs and 9fs do (when they
> +	some or all folios in an address_space.  This can happen
> +	through the fadvise(POSIX_FADV_DONTNEED) system call or by the
> +	filesystem explicitly requesting it as nfs and 9p do (when they
>  	believe the cache may be out of date with storage) by calling
>  	invalidate_inode_pages2().  If the filesystem makes such a call,
> -	and needs to be certain that all pages are invalidated, then its
> -	releasepage will need to ensure this.  Possibly it can clear the
> -	PageUptodate bit if it cannot free private data yet.
> +	and needs to be certain that all folios are invalidated, then
> +	its release_folio will need to ensure this.  Possibly it can
> +	clear the uptodate flag if it cannot free private data yet.
>  
>  ``freepage``
>  	freepage is called once the page is no longer visible in the
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index f812f5aa07dd..ad768f13f485 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -355,6 +355,7 @@ struct address_space_operations {
>  	/* Unfortunately this kludge is needed for FIBMAP. Don't use it */
>  	sector_t (*bmap)(struct address_space *, sector_t);
>  	void (*invalidate_folio) (struct folio *, size_t offset, size_t len);
> +	bool (*release_folio)(struct folio *, gfp_t);
>  	int (*releasepage) (struct page *, gfp_t);
>  	void (*freepage)(struct page *);
>  	ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
> diff --git a/mm/filemap.c b/mm/filemap.c
> index 9b7fa47feb5e..78e4a7dc3a56 100644
> --- a/mm/filemap.c
> +++ b/mm/filemap.c
> @@ -3955,6 +3955,8 @@ bool filemap_release_folio(struct folio *folio, gfp_t gfp)
>  	if (folio_test_writeback(folio))
>  		return false;
>  
> +	if (mapping && mapping->a_ops->release_folio)
> +		return mapping->a_ops->release_folio(folio, gfp);
>  	if (mapping && mapping->a_ops->releasepage)
>  		return mapping->a_ops->releasepage(&folio->page, gfp);
>  	return try_to_free_buffers(&folio->page);

-- 
Jeff Layton <jlayton@kernel.org>

Re: [PATCH 01/26] fs: Add aops->release_folio

[Matthew Wilcox]

On Mon, May 09, 2022 at 06:33:55AM -0400, Jeff Layton wrote:
> On Sun, 2022-05-08 at 21:32 +0100, Matthew Wilcox (Oracle) wrote:
> > +++ b/Documentation/filesystems/locking.rst
> > @@ -249,7 +249,7 @@ prototypes::
> >  				struct page *page, void *fsdata);
> >  	sector_t (*bmap)(struct address_space *, sector_t);
> >  	void (*invalidate_folio) (struct folio *, size_t start, size_t len);
> > -	int (*releasepage) (struct page *, int);
> > +	int (*release_folio)(struct folio *, gfp_t);
> 
> Shouldn't that be a bool return?

Oh, good catch.  This whole documentation section needs a re-do in
kernel-doc, so we don't have to duplicate this.  Thanks; I'll fix that
up.

[PATCH 02/26] iomap: Convert to release_folio

[Matthew Wilcox (Oracle)]

Change all the filesystems which used iomap_releasepage to use the
new function.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/gfs2/aops.c         |  2 +-
 fs/iomap/buffered-io.c | 22 ++++++++++------------
 fs/iomap/trace.h       |  2 +-
 fs/xfs/xfs_aops.c      |  2 +-
 fs/zonefs/super.c      |  2 +-
 include/linux/iomap.h  |  2 +-
 6 files changed, 15 insertions(+), 17 deletions(-)

diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 1016631bcbdc..3d6c5c5eb4f1 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -768,7 +768,7 @@ static const struct address_space_operations gfs2_aops = {
 	.read_folio = gfs2_read_folio,
 	.readahead = gfs2_readahead,
 	.dirty_folio = filemap_dirty_folio,
-	.releasepage = iomap_releasepage,
+	.release_folio = iomap_release_folio,
 	.invalidate_folio = iomap_invalidate_folio,
 	.bmap = gfs2_bmap,
 	.direct_IO = noop_direct_IO,
diff --git a/fs/iomap/buffered-io.c b/fs/iomap/buffered-io.c
index 2de087ac87b6..8532f0e2e2d6 100644
--- a/fs/iomap/buffered-io.c
+++ b/fs/iomap/buffered-io.c
@@ -452,25 +452,23 @@ bool iomap_is_partially_uptodate(struct folio *folio, size_t from, size_t count)
 }
 EXPORT_SYMBOL_GPL(iomap_is_partially_uptodate);
 
-int
-iomap_releasepage(struct page *page, gfp_t gfp_mask)
+bool iomap_release_folio(struct folio *folio, gfp_t gfp_flags)
 {
-	struct folio *folio = page_folio(page);
-
-	trace_iomap_releasepage(folio->mapping->host, folio_pos(folio),
+	trace_iomap_release_folio(folio->mapping->host, folio_pos(folio),
 			folio_size(folio));
 
 	/*
-	 * mm accommodates an old ext3 case where clean pages might not have had
-	 * the dirty bit cleared. Thus, it can send actual dirty pages to
-	 * ->releasepage() via shrink_active_list(); skip those here.
+	 * mm accommodates an old ext3 case where clean folios might
+	 * not have had the dirty bit cleared.  Thus, it can send actual
+	 * dirty folios to ->release_folio() via shrink_active_list();
+	 * skip those here.
 	 */
 	if (folio_test_dirty(folio) || folio_test_writeback(folio))
-		return 0;
+		return false;
 	iomap_page_release(folio);
-	return 1;
+	return true;
 }
-EXPORT_SYMBOL_GPL(iomap_releasepage);
+EXPORT_SYMBOL_GPL(iomap_release_folio);
 
 void iomap_invalidate_folio(struct folio *folio, size_t offset, size_t len)
 {
@@ -1483,7 +1481,7 @@ iomap_do_writepage(struct page *page, struct writeback_control *wbc, void *data)
 		 * Skip the page if it's fully outside i_size, e.g. due to a
 		 * truncate operation that's in progress. We must redirty the
 		 * page so that reclaim stops reclaiming it. Otherwise
-		 * iomap_vm_releasepage() is called on it and gets confused.
+		 * iomap_release_folio() is called on it and gets confused.
 		 *
 		 * Note that the end_index is unsigned long.  If the given
 		 * offset is greater than 16TB on a 32-bit system then if we
diff --git a/fs/iomap/trace.h b/fs/iomap/trace.h
index a6689a563c6e..d48868fc40d7 100644
--- a/fs/iomap/trace.h
+++ b/fs/iomap/trace.h
@@ -80,7 +80,7 @@ DEFINE_EVENT(iomap_range_class, name,	\
 	TP_PROTO(struct inode *inode, loff_t off, u64 len),\
 	TP_ARGS(inode, off, len))
 DEFINE_RANGE_EVENT(iomap_writepage);
-DEFINE_RANGE_EVENT(iomap_releasepage);
+DEFINE_RANGE_EVENT(iomap_release_folio);
 DEFINE_RANGE_EVENT(iomap_invalidate_folio);
 DEFINE_RANGE_EVENT(iomap_dio_invalidate_fail);
 
diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c
index a9c4bb500d53..2acbfc6925dd 100644
--- a/fs/xfs/xfs_aops.c
+++ b/fs/xfs/xfs_aops.c
@@ -568,7 +568,7 @@ const struct address_space_operations xfs_address_space_operations = {
 	.readahead		= xfs_vm_readahead,
 	.writepages		= xfs_vm_writepages,
 	.dirty_folio		= filemap_dirty_folio,
-	.releasepage		= iomap_releasepage,
+	.release_folio		= iomap_release_folio,
 	.invalidate_folio	= iomap_invalidate_folio,
 	.bmap			= xfs_vm_bmap,
 	.direct_IO		= noop_direct_IO,
diff --git a/fs/zonefs/super.c b/fs/zonefs/super.c
index c3a38f711b24..b1a428f860b3 100644
--- a/fs/zonefs/super.c
+++ b/fs/zonefs/super.c
@@ -197,7 +197,7 @@ static const struct address_space_operations zonefs_file_aops = {
 	.writepage		= zonefs_writepage,
 	.writepages		= zonefs_writepages,
 	.dirty_folio		= filemap_dirty_folio,
-	.releasepage		= iomap_releasepage,
+	.release_folio		= iomap_release_folio,
 	.invalidate_folio	= iomap_invalidate_folio,
 	.migratepage		= iomap_migrate_page,
 	.is_partially_uptodate	= iomap_is_partially_uptodate,
diff --git a/include/linux/iomap.h b/include/linux/iomap.h
index 5b2aa45ddda3..0d674695b6d3 100644
--- a/include/linux/iomap.h
+++ b/include/linux/iomap.h
@@ -228,7 +228,7 @@ ssize_t iomap_file_buffered_write(struct kiocb *iocb, struct iov_iter *from,
 int iomap_read_folio(struct folio *folio, const struct iomap_ops *ops);
 void iomap_readahead(struct readahead_control *, const struct iomap_ops *ops);
 bool iomap_is_partially_uptodate(struct folio *, size_t from, size_t count);
-int iomap_releasepage(struct page *page, gfp_t gfp_mask);
+bool iomap_release_folio(struct folio *folio, gfp_t gfp_flags);
 void iomap_invalidate_folio(struct folio *folio, size_t offset, size_t len);
 #ifdef CONFIG_MIGRATION
 int iomap_migrate_page(struct address_space *mapping, struct page *newpage,
-- 
2.34.1

[PATCH 03/26] 9p: Convert to release_folio

[Matthew Wilcox (Oracle)]

A straightforward conversion as it already works in terms of folios.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/9p/vfs_addr.c | 17 ++++++++---------
 1 file changed, 8 insertions(+), 9 deletions(-)

diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c
index 3a84167f4893..8ce82ff1e40a 100644
--- a/fs/9p/vfs_addr.c
+++ b/fs/9p/vfs_addr.c
@@ -100,29 +100,28 @@ const struct netfs_request_ops v9fs_req_ops = {
 };
 
 /**
- * v9fs_release_page - release the private state associated with a page
- * @page: The page to be released
+ * v9fs_release_folio - release the private state associated with a folio
+ * @folio: The folio to be released
  * @gfp: The caller's allocation restrictions
  *
- * Returns 1 if the page can be released, false otherwise.
+ * Returns true if the page can be released, false otherwise.
  */
 
-static int v9fs_release_page(struct page *page, gfp_t gfp)
+static bool v9fs_release_folio(struct folio *folio, gfp_t gfp)
 {
-	struct folio *folio = page_folio(page);
 	struct inode *inode = folio_inode(folio);
 
 	if (folio_test_private(folio))
-		return 0;
+		return false;
 #ifdef CONFIG_9P_FSCACHE
 	if (folio_test_fscache(folio)) {
 		if (current_is_kswapd() || !(gfp & __GFP_FS))
-			return 0;
+			return false;
 		folio_wait_fscache(folio);
 	}
 #endif
 	fscache_note_page_release(v9fs_inode_cookie(V9FS_I(inode)));
-	return 1;
+	return true;
 }
 
 static void v9fs_invalidate_folio(struct folio *folio, size_t offset,
@@ -342,7 +341,7 @@ const struct address_space_operations v9fs_addr_operations = {
 	.writepage = v9fs_vfs_writepage,
 	.write_begin = v9fs_write_begin,
 	.write_end = v9fs_write_end,
-	.releasepage = v9fs_release_page,
+	.release_folio = v9fs_release_folio,
 	.invalidate_folio = v9fs_invalidate_folio,
 	.launder_folio = v9fs_launder_folio,
 	.direct_IO = v9fs_direct_IO,
-- 
2.34.1

[PATCH 04/26] afs: Convert to release_folio

[Matthew Wilcox (Oracle)]

A straightforward conversion as they already work in terms of folios.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/afs/dir.c      |  7 +++----
 fs/afs/file.c     | 11 +++++------
 fs/afs/internal.h |  2 +-
 3 files changed, 9 insertions(+), 11 deletions(-)

diff --git a/fs/afs/dir.c b/fs/afs/dir.c
index 932e61e28e5d..94aa7356248e 100644
--- a/fs/afs/dir.c
+++ b/fs/afs/dir.c
@@ -41,7 +41,7 @@ static int afs_symlink(struct user_namespace *mnt_userns, struct inode *dir,
 static int afs_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
 		      struct dentry *old_dentry, struct inode *new_dir,
 		      struct dentry *new_dentry, unsigned int flags);
-static int afs_dir_releasepage(struct page *page, gfp_t gfp_flags);
+static bool afs_dir_release_folio(struct folio *folio, gfp_t gfp_flags);
 static void afs_dir_invalidate_folio(struct folio *folio, size_t offset,
 				   size_t length);
 
@@ -75,7 +75,7 @@ const struct inode_operations afs_dir_inode_operations = {
 
 const struct address_space_operations afs_dir_aops = {
 	.dirty_folio	= afs_dir_dirty_folio,
-	.releasepage	= afs_dir_releasepage,
+	.release_folio	= afs_dir_release_folio,
 	.invalidate_folio = afs_dir_invalidate_folio,
 };
 
@@ -2002,9 +2002,8 @@ static int afs_rename(struct user_namespace *mnt_userns, struct inode *old_dir,
  * Release a directory folio and clean up its private state if it's not busy
  * - return true if the folio can now be released, false if not
  */
-static int afs_dir_releasepage(struct page *subpage, gfp_t gfp_flags)
+static bool afs_dir_release_folio(struct folio *folio, gfp_t gfp_flags)
 {
-	struct folio *folio = page_folio(subpage);
 	struct afs_vnode *dvnode = AFS_FS_I(folio_inode(folio));
 
 	_enter("{{%llx:%llu}[%lu]}", dvnode->fid.vid, dvnode->fid.vnode, folio_index(folio));
diff --git a/fs/afs/file.c b/fs/afs/file.c
index 65ef69a1f78e..a8e8832179e4 100644
--- a/fs/afs/file.c
+++ b/fs/afs/file.c
@@ -22,7 +22,7 @@ static int afs_file_mmap(struct file *file, struct vm_area_struct *vma);
 static int afs_symlink_read_folio(struct file *file, struct folio *folio);
 static void afs_invalidate_folio(struct folio *folio, size_t offset,
 			       size_t length);
-static int afs_releasepage(struct page *page, gfp_t gfp_flags);
+static bool afs_release_folio(struct folio *folio, gfp_t gfp_flags);
 
 static ssize_t afs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter);
 static void afs_vm_open(struct vm_area_struct *area);
@@ -54,7 +54,7 @@ const struct address_space_operations afs_file_aops = {
 	.readahead	= netfs_readahead,
 	.dirty_folio	= afs_dirty_folio,
 	.launder_folio	= afs_launder_folio,
-	.releasepage	= afs_releasepage,
+	.release_folio	= afs_release_folio,
 	.invalidate_folio = afs_invalidate_folio,
 	.write_begin	= afs_write_begin,
 	.write_end	= afs_write_end,
@@ -64,7 +64,7 @@ const struct address_space_operations afs_file_aops = {
 
 const struct address_space_operations afs_symlink_aops = {
 	.read_folio	= afs_symlink_read_folio,
-	.releasepage	= afs_releasepage,
+	.release_folio	= afs_release_folio,
 	.invalidate_folio = afs_invalidate_folio,
 };
 
@@ -481,16 +481,15 @@ static void afs_invalidate_folio(struct folio *folio, size_t offset,
  * release a page and clean up its private state if it's not busy
  * - return true if the page can now be released, false if not
  */
-static int afs_releasepage(struct page *page, gfp_t gfp)
+static bool afs_release_folio(struct folio *folio, gfp_t gfp)
 {
-	struct folio *folio = page_folio(page);
 	struct afs_vnode *vnode = AFS_FS_I(folio_inode(folio));
 
 	_enter("{{%llx:%llu}[%lu],%lx},%x",
 	       vnode->fid.vid, vnode->fid.vnode, folio_index(folio), folio->flags,
 	       gfp);
 
-	/* deny if page is being written to the cache and the caller hasn't
+	/* deny if folio is being written to the cache and the caller hasn't
 	 * elected to wait */
 #ifdef CONFIG_AFS_FSCACHE
 	if (folio_test_fscache(folio)) {
diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 7a72e9c60423..a30995901266 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -311,7 +311,7 @@ struct afs_net {
 	atomic_t		n_lookup;	/* Number of lookups done */
 	atomic_t		n_reval;	/* Number of dentries needing revalidation */
 	atomic_t		n_inval;	/* Number of invalidations by the server */
-	atomic_t		n_relpg;	/* Number of invalidations by releasepage */
+	atomic_t		n_relpg;	/* Number of invalidations by release_folio */
 	atomic_t		n_read_dir;	/* Number of directory pages read */
 	atomic_t		n_dir_cr;	/* Number of directory entry creation edits */
 	atomic_t		n_dir_rm;	/* Number of directory entry removal edits */
-- 
2.34.1

[PATCH 05/26] btrfs: Convert to release_folio

[Matthew Wilcox (Oracle)]

I've only converted the outer layers of the btrfs release_folio paths
to use folios; the use of folios should be pushed further down into
btrfs from here.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/btrfs/disk-io.c   | 12 ++++++------
 fs/btrfs/extent_io.c | 14 +++++++-------
 fs/btrfs/file.c      |  2 +-
 fs/btrfs/inode.c     | 24 ++++++++++++------------
 4 files changed, 26 insertions(+), 26 deletions(-)

diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
index 31c3f592e587..c281b044a409 100644
--- a/fs/btrfs/disk-io.c
+++ b/fs/btrfs/disk-io.c
@@ -1005,12 +1005,12 @@ static int btree_writepages(struct address_space *mapping,
 	return btree_write_cache_pages(mapping, wbc);
 }
 
-static int btree_releasepage(struct page *page, gfp_t gfp_flags)
+static bool btree_release_folio(struct folio *folio, gfp_t gfp_flags)
 {
-	if (PageWriteback(page) || PageDirty(page))
-		return 0;
+	if (folio_test_writeback(folio) || folio_test_dirty(folio))
+		return false;
 
-	return try_release_extent_buffer(page);
+	return try_release_extent_buffer(&folio->page);
 }
 
 static void btree_invalidate_folio(struct folio *folio, size_t offset,
@@ -1019,7 +1019,7 @@ static void btree_invalidate_folio(struct folio *folio, size_t offset,
 	struct extent_io_tree *tree;
 	tree = &BTRFS_I(folio->mapping->host)->io_tree;
 	extent_invalidate_folio(tree, folio, offset);
-	btree_releasepage(&folio->page, GFP_NOFS);
+	btree_release_folio(folio, GFP_NOFS);
 	if (folio_get_private(folio)) {
 		btrfs_warn(BTRFS_I(folio->mapping->host)->root->fs_info,
 			   "folio private not zero on folio %llu",
@@ -1080,7 +1080,7 @@ static bool btree_dirty_folio(struct address_space *mapping,
 
 static const struct address_space_operations btree_aops = {
 	.writepages	= btree_writepages,
-	.releasepage	= btree_releasepage,
+	.release_folio	= btree_release_folio,
 	.invalidate_folio = btree_invalidate_folio,
 #ifdef CONFIG_MIGRATION
 	.migratepage	= btree_migratepage,
diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c
index 33c19f51d79b..e7a6e8757859 100644
--- a/fs/btrfs/extent_io.c
+++ b/fs/btrfs/extent_io.c
@@ -5271,7 +5271,7 @@ int extent_invalidate_folio(struct extent_io_tree *tree,
 }
 
 /*
- * a helper for releasepage, this tests for areas of the page that
+ * a helper for release_folio, this tests for areas of the page that
  * are locked or under IO and drops the related state bits if it is safe
  * to drop the page.
  */
@@ -5307,7 +5307,7 @@ static int try_release_extent_state(struct extent_io_tree *tree,
 }
 
 /*
- * a helper for releasepage.  As long as there are no locked extents
+ * a helper for release_folio.  As long as there are no locked extents
  * in the range corresponding to the page, both state records and extent
  * map records are removed
  */
@@ -6001,10 +6001,10 @@ static void check_buffer_tree_ref(struct extent_buffer *eb)
 	 *
 	 * It is only cleared in two cases: freeing the last non-tree
 	 * reference to the extent_buffer when its STALE bit is set or
-	 * calling releasepage when the tree reference is the only reference.
+	 * calling release_folio when the tree reference is the only reference.
 	 *
 	 * In both cases, care is taken to ensure that the extent_buffer's
-	 * pages are not under io. However, releasepage can be concurrently
+	 * pages are not under io. However, release_folio can be concurrently
 	 * called with creating new references, which is prone to race
 	 * conditions between the calls to check_buffer_tree_ref in those
 	 * codepaths and clearing TREE_REF in try_release_extent_buffer.
@@ -6257,7 +6257,7 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 		/*
 		 * We can't unlock the pages just yet since the extent buffer
 		 * hasn't been properly inserted in the radix tree, this
-		 * opens a race with btree_releasepage which can free a page
+		 * opens a race with btree_release_folio which can free a page
 		 * while we are still filling in all pages for the buffer and
 		 * we could crash.
 		 */
@@ -6289,7 +6289,7 @@ struct extent_buffer *alloc_extent_buffer(struct btrfs_fs_info *fs_info,
 
 	/*
 	 * Now it's safe to unlock the pages because any calls to
-	 * btree_releasepage will correctly detect that a page belongs to a
+	 * btree_release_folio will correctly detect that a page belongs to a
 	 * live buffer and won't free them prematurely.
 	 */
 	for (i = 0; i < num_pages; i++)
@@ -6659,7 +6659,7 @@ int read_extent_buffer_pages(struct extent_buffer *eb, int wait, int mirror_num)
 	eb->read_mirror = 0;
 	atomic_set(&eb->io_pages, num_reads);
 	/*
-	 * It is possible for releasepage to clear the TREE_REF bit before we
+	 * It is possible for release_folio to clear the TREE_REF bit before we
 	 * set io_pages. See check_buffer_tree_ref for a more detailed comment.
 	 */
 	check_buffer_tree_ref(eb);
diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c
index 57fba5abb059..c1eadb3f715c 100644
--- a/fs/btrfs/file.c
+++ b/fs/btrfs/file.c
@@ -1323,7 +1323,7 @@ static int prepare_uptodate_page(struct inode *inode,
 
 		/*
 		 * Since btrfs_read_folio() will unlock the folio before it
-		 * returns, there is a window where btrfs_releasepage() can be
+		 * returns, there is a window where btrfs_release_folio() can be
 		 * called to release the page.  Here we check both inode
 		 * mapping and PagePrivate() to make sure the page was not
 		 * released.
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 9a2530b1695d..1e52710babb4 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -8172,7 +8172,7 @@ static void btrfs_readahead(struct readahead_control *rac)
 }
 
 /*
- * For releasepage() and invalidate_folio() we have a race window where
+ * For release_folio() and invalidate_folio() we have a race window where
  * folio_end_writeback() is called but the subpage spinlock is not yet released.
  * If we continue to release/invalidate the page, we could cause use-after-free
  * for subpage spinlock.  So this function is to spin and wait for subpage
@@ -8204,22 +8204,22 @@ static void wait_subpage_spinlock(struct page *page)
 	spin_unlock_irq(&subpage->lock);
 }
 
-static int __btrfs_releasepage(struct page *page, gfp_t gfp_flags)
+static bool __btrfs_release_folio(struct folio *folio, gfp_t gfp_flags)
 {
-	int ret = try_release_extent_mapping(page, gfp_flags);
+	int ret = try_release_extent_mapping(&folio->page, gfp_flags);
 
 	if (ret == 1) {
-		wait_subpage_spinlock(page);
-		clear_page_extent_mapped(page);
+		wait_subpage_spinlock(&folio->page);
+		clear_page_extent_mapped(&folio->page);
 	}
 	return ret;
 }
 
-static int btrfs_releasepage(struct page *page, gfp_t gfp_flags)
+static bool btrfs_release_folio(struct folio *folio, gfp_t gfp_flags)
 {
-	if (PageWriteback(page) || PageDirty(page))
-		return 0;
-	return __btrfs_releasepage(page, gfp_flags);
+	if (folio_test_writeback(folio) || folio_test_dirty(folio))
+		return false;
+	return __btrfs_release_folio(folio, gfp_flags);
 }
 
 #ifdef CONFIG_MIGRATION
@@ -8290,7 +8290,7 @@ static void btrfs_invalidate_folio(struct folio *folio, size_t offset,
 	 * still safe to wait for ordered extent to finish.
 	 */
 	if (!(offset == 0 && length == folio_size(folio))) {
-		btrfs_releasepage(&folio->page, GFP_NOFS);
+		btrfs_release_folio(folio, GFP_NOFS);
 		return;
 	}
 
@@ -8414,7 +8414,7 @@ static void btrfs_invalidate_folio(struct folio *folio, size_t offset,
 	ASSERT(!folio_test_ordered(folio));
 	btrfs_page_clear_checked(fs_info, &folio->page, folio_pos(folio), folio_size(folio));
 	if (!inode_evicting)
-		__btrfs_releasepage(&folio->page, GFP_NOFS);
+		__btrfs_release_folio(folio, GFP_NOFS);
 	clear_page_extent_mapped(&folio->page);
 }
 
@@ -11364,7 +11364,7 @@ static const struct address_space_operations btrfs_aops = {
 	.readahead	= btrfs_readahead,
 	.direct_IO	= noop_direct_IO,
 	.invalidate_folio = btrfs_invalidate_folio,
-	.releasepage	= btrfs_releasepage,
+	.release_folio	= btrfs_release_folio,
 #ifdef CONFIG_MIGRATION
 	.migratepage	= btrfs_migratepage,
 #endif
-- 
2.34.1

[PATCH 06/26] ceph: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio throughout ceph_release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ceph/addr.c | 24 ++++++++++++------------
 1 file changed, 12 insertions(+), 12 deletions(-)

diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index e040b92bb17c..737d13931d52 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -162,24 +162,24 @@ static void ceph_invalidate_folio(struct folio *folio, size_t offset,
 	folio_wait_fscache(folio);
 }
 
-static int ceph_releasepage(struct page *page, gfp_t gfp)
+static bool ceph_release_folio(struct folio *folio, gfp_t gfp)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 
-	dout("%llx:%llx releasepage %p idx %lu (%sdirty)\n",
-	     ceph_vinop(inode), page,
-	     page->index, PageDirty(page) ? "" : "not ");
+	dout("%llx:%llx release_folio idx %lu (%sdirty)\n",
+	     ceph_vinop(inode),
+	     folio->index, folio_test_dirty(folio) ? "" : "not ");
 
-	if (PagePrivate(page))
-		return 0;
+	if (folio_test_private(folio))
+		return false;
 
-	if (PageFsCache(page)) {
+	if (folio_test_fscache(folio)) {
 		if (current_is_kswapd() || !(gfp & __GFP_FS))
-			return 0;
-		wait_on_page_fscache(page);
+			return false;
+		folio_wait_fscache(folio);
 	}
 	ceph_fscache_note_page_release(inode);
-	return 1;
+	return true;
 }
 
 static void ceph_netfs_expand_readahead(struct netfs_io_request *rreq)
@@ -1380,7 +1380,7 @@ const struct address_space_operations ceph_aops = {
 	.write_end = ceph_write_end,
 	.dirty_folio = ceph_dirty_folio,
 	.invalidate_folio = ceph_invalidate_folio,
-	.releasepage = ceph_releasepage,
+	.release_folio = ceph_release_folio,
 	.direct_IO = noop_direct_IO,
 };
 
-- 
2.34.1

[PATCH 07/26] cifs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio throughout cifs_release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/cifs/file.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/fs/cifs/file.c b/fs/cifs/file.c
index bc6d88e2e672..06003bb9cbe9 100644
--- a/fs/cifs/file.c
+++ b/fs/cifs/file.c
@@ -4758,16 +4758,16 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping,
 	return rc;
 }
 
-static int cifs_release_page(struct page *page, gfp_t gfp)
+static bool cifs_release_folio(struct folio *folio, gfp_t gfp)
 {
-	if (PagePrivate(page))
+	if (folio_test_private(folio))
 		return 0;
-	if (PageFsCache(page)) {
+	if (folio_test_fscache(folio)) {
 		if (current_is_kswapd() || !(gfp & __GFP_FS))
 			return false;
-		wait_on_page_fscache(page);
+		folio_wait_fscache(folio);
 	}
-	fscache_note_page_release(cifs_inode_cookie(page->mapping->host));
+	fscache_note_page_release(cifs_inode_cookie(folio->mapping->host));
 	return true;
 }
 
@@ -4973,7 +4973,7 @@ const struct address_space_operations cifs_addr_ops = {
 	.write_begin = cifs_write_begin,
 	.write_end = cifs_write_end,
 	.dirty_folio = cifs_dirty_folio,
-	.releasepage = cifs_release_page,
+	.release_folio = cifs_release_folio,
 	.direct_IO = cifs_direct_io,
 	.invalidate_folio = cifs_invalidate_folio,
 	.launder_folio = cifs_launder_folio,
@@ -4998,7 +4998,7 @@ const struct address_space_operations cifs_addr_ops_smallbuf = {
 	.write_begin = cifs_write_begin,
 	.write_end = cifs_write_end,
 	.dirty_folio = cifs_dirty_folio,
-	.releasepage = cifs_release_page,
+	.release_folio = cifs_release_folio,
 	.invalidate_folio = cifs_invalidate_folio,
 	.launder_folio = cifs_launder_folio,
 };
-- 
2.34.1

[PATCH 08/26] erofs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio in erofs_managed_cache_release_folio(), but use of folios
should be pushed into erofs_try_to_free_cached_page().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/erofs/super.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/fs/erofs/super.c b/fs/erofs/super.c
index 0c4b41130c2f..0e3862a72bfe 100644
--- a/fs/erofs/super.c
+++ b/fs/erofs/super.c
@@ -518,16 +518,16 @@ static int erofs_fc_parse_param(struct fs_context *fc,
 #ifdef CONFIG_EROFS_FS_ZIP
 static const struct address_space_operations managed_cache_aops;
 
-static int erofs_managed_cache_releasepage(struct page *page, gfp_t gfp_mask)
+static bool erofs_managed_cache_release_folio(struct folio *folio, gfp_t gfp)
 {
-	int ret = 1;	/* 0 - busy */
-	struct address_space *const mapping = page->mapping;
+	bool ret = true;
+	struct address_space *const mapping = folio->mapping;
 
-	DBG_BUGON(!PageLocked(page));
+	DBG_BUGON(!folio_test_locked(folio));
 	DBG_BUGON(mapping->a_ops != &managed_cache_aops);
 
-	if (PagePrivate(page))
-		ret = erofs_try_to_free_cached_page(page);
+	if (folio_test_private(folio))
+		ret = erofs_try_to_free_cached_page(&folio->page);
 
 	return ret;
 }
@@ -548,12 +548,12 @@ static void erofs_managed_cache_invalidate_folio(struct folio *folio,
 	DBG_BUGON(stop > folio_size(folio) || stop < length);
 
 	if (offset == 0 && stop == folio_size(folio))
-		while (!erofs_managed_cache_releasepage(&folio->page, GFP_NOFS))
+		while (!erofs_managed_cache_release_folio(folio, GFP_NOFS))
 			cond_resched();
 }
 
 static const struct address_space_operations managed_cache_aops = {
-	.releasepage = erofs_managed_cache_releasepage,
+	.release_folio = erofs_managed_cache_release_folio,
 	.invalidate_folio = erofs_managed_cache_invalidate_folio,
 };
 
-- 
2.34.1

[PATCH 09/26] ext4: Convert to release_folio

[Matthew Wilcox (Oracle)]

The use of folios should be pushed deeper into ext4 from here.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ext4/inode.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index c6b8cb4949f1..52c46ac5bc8a 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3243,19 +3243,19 @@ static void ext4_journalled_invalidate_folio(struct folio *folio,
 	WARN_ON(__ext4_journalled_invalidate_folio(folio, offset, length) < 0);
 }
 
-static int ext4_releasepage(struct page *page, gfp_t wait)
+static bool ext4_release_folio(struct folio *folio, gfp_t wait)
 {
-	journal_t *journal = EXT4_JOURNAL(page->mapping->host);
+	journal_t *journal = EXT4_JOURNAL(folio->mapping->host);
 
-	trace_ext4_releasepage(page);
+	trace_ext4_releasepage(&folio->page);
 
 	/* Page has dirty journalled data -> cannot release */
-	if (PageChecked(page))
-		return 0;
+	if (folio_test_checked(folio))
+		return false;
 	if (journal)
-		return jbd2_journal_try_to_free_buffers(journal, page);
+		return jbd2_journal_try_to_free_buffers(journal, &folio->page);
 	else
-		return try_to_free_buffers(page);
+		return try_to_free_buffers(&folio->page);
 }
 
 static bool ext4_inode_datasync_dirty(struct inode *inode)
@@ -3618,7 +3618,7 @@ static const struct address_space_operations ext4_aops = {
 	.dirty_folio		= ext4_dirty_folio,
 	.bmap			= ext4_bmap,
 	.invalidate_folio	= ext4_invalidate_folio,
-	.releasepage		= ext4_releasepage,
+	.release_folio		= ext4_release_folio,
 	.direct_IO		= noop_direct_IO,
 	.migratepage		= buffer_migrate_page,
 	.is_partially_uptodate  = block_is_partially_uptodate,
@@ -3636,7 +3636,7 @@ static const struct address_space_operations ext4_journalled_aops = {
 	.dirty_folio		= ext4_journalled_dirty_folio,
 	.bmap			= ext4_bmap,
 	.invalidate_folio	= ext4_journalled_invalidate_folio,
-	.releasepage		= ext4_releasepage,
+	.release_folio		= ext4_release_folio,
 	.direct_IO		= noop_direct_IO,
 	.is_partially_uptodate  = block_is_partially_uptodate,
 	.error_remove_page	= generic_error_remove_page,
@@ -3653,7 +3653,7 @@ static const struct address_space_operations ext4_da_aops = {
 	.dirty_folio		= ext4_dirty_folio,
 	.bmap			= ext4_bmap,
 	.invalidate_folio	= ext4_invalidate_folio,
-	.releasepage		= ext4_releasepage,
+	.release_folio		= ext4_release_folio,
 	.direct_IO		= noop_direct_IO,
 	.migratepage		= buffer_migrate_page,
 	.is_partially_uptodate  = block_is_partially_uptodate,
-- 
2.34.1

Re: [PATCH 09/26] ext4: Convert to release_folio

[Theodore Ts'o]

On Sun, May 08, 2022 at 09:32:30PM +0100, Matthew Wilcox (Oracle) wrote:
> The use of folios should be pushed deeper into ext4 from here.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Acked-by: Theodore Ts'o <tytso@mit.edu>

						- Ted

[PATCH 10/26] f2fs: Convert to release_folio

[Matthew Wilcox (Oracle)]

While converting f2fs_release_page() to f2fs_release_folio(), cache the
sb_info so we don't need to retrieve it twice, and remove the redundant
call to set_page_private().  The use of folios should be pushed further
into f2fs from here.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/f2fs/checkpoint.c |  2 +-
 fs/f2fs/compress.c   |  2 +-
 fs/f2fs/data.c       | 32 +++++++++++++++++---------------
 fs/f2fs/f2fs.h       |  2 +-
 fs/f2fs/node.c       |  2 +-
 5 files changed, 21 insertions(+), 19 deletions(-)

diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c
index 909085a78f9c..456c1e89386a 100644
--- a/fs/f2fs/checkpoint.c
+++ b/fs/f2fs/checkpoint.c
@@ -468,7 +468,7 @@ const struct address_space_operations f2fs_meta_aops = {
 	.writepages	= f2fs_write_meta_pages,
 	.dirty_folio	= f2fs_dirty_meta_folio,
 	.invalidate_folio = f2fs_invalidate_folio,
-	.releasepage	= f2fs_release_page,
+	.release_folio	= f2fs_release_folio,
 #ifdef CONFIG_MIGRATION
 	.migratepage    = f2fs_migrate_page,
 #endif
diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c
index 12a56f9e1572..24824cd96f36 100644
--- a/fs/f2fs/compress.c
+++ b/fs/f2fs/compress.c
@@ -1746,7 +1746,7 @@ unsigned int f2fs_cluster_blocks_are_contiguous(struct dnode_of_data *dn)
 }
 
 const struct address_space_operations f2fs_compress_aops = {
-	.releasepage = f2fs_release_page,
+	.release_folio = f2fs_release_folio,
 	.invalidate_folio = f2fs_invalidate_folio,
 };
 
diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c
index f894267f0722..8f38c26bb16c 100644
--- a/fs/f2fs/data.c
+++ b/fs/f2fs/data.c
@@ -3528,28 +3528,30 @@ void f2fs_invalidate_folio(struct folio *folio, size_t offset, size_t length)
 	folio_detach_private(folio);
 }
 
-int f2fs_release_page(struct page *page, gfp_t wait)
+bool f2fs_release_folio(struct folio *folio, gfp_t wait)
 {
-	/* If this is dirty page, keep PagePrivate */
-	if (PageDirty(page))
-		return 0;
+	struct f2fs_sb_info *sbi;
+
+	/* If this is dirty folio, keep private data */
+	if (folio_test_dirty(folio))
+		return false;
 
 	/* This is atomic written page, keep Private */
-	if (page_private_atomic(page))
-		return 0;
+	if (page_private_atomic(&folio->page))
+		return false;
 
-	if (test_opt(F2FS_P_SB(page), COMPRESS_CACHE)) {
-		struct inode *inode = page->mapping->host;
+	sbi = F2FS_M_SB(folio->mapping);
+	if (test_opt(sbi, COMPRESS_CACHE)) {
+		struct inode *inode = folio->mapping->host;
 
-		if (inode->i_ino == F2FS_COMPRESS_INO(F2FS_I_SB(inode)))
-			clear_page_private_data(page);
+		if (inode->i_ino == F2FS_COMPRESS_INO(sbi))
+			clear_page_private_data(&folio->page);
 	}
 
-	clear_page_private_gcing(page);
+	clear_page_private_gcing(&folio->page);
 
-	detach_page_private(page);
-	set_page_private(page, 0);
-	return 1;
+	folio_detach_private(folio);
+	return true;
 }
 
 static bool f2fs_dirty_data_folio(struct address_space *mapping,
@@ -3944,7 +3946,7 @@ const struct address_space_operations f2fs_dblock_aops = {
 	.write_end	= f2fs_write_end,
 	.dirty_folio	= f2fs_dirty_data_folio,
 	.invalidate_folio = f2fs_invalidate_folio,
-	.releasepage	= f2fs_release_page,
+	.release_folio	= f2fs_release_folio,
 	.direct_IO	= noop_direct_IO,
 	.bmap		= f2fs_bmap,
 	.swap_activate  = f2fs_swap_activate,
diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h
index 18df53ef3d7e..73ebac078884 100644
--- a/fs/f2fs/f2fs.h
+++ b/fs/f2fs/f2fs.h
@@ -3768,7 +3768,7 @@ int f2fs_write_single_data_page(struct page *page, int *submitted,
 				int compr_blocks, bool allow_balance);
 void f2fs_write_failed(struct inode *inode, loff_t to);
 void f2fs_invalidate_folio(struct folio *folio, size_t offset, size_t length);
-int f2fs_release_page(struct page *page, gfp_t wait);
+bool f2fs_release_folio(struct folio *folio, gfp_t wait);
 #ifdef CONFIG_MIGRATION
 int f2fs_migrate_page(struct address_space *mapping, struct page *newpage,
 			struct page *page, enum migrate_mode mode);
diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c
index c45d341dcf6e..8ccff18560ff 100644
--- a/fs/f2fs/node.c
+++ b/fs/f2fs/node.c
@@ -2165,7 +2165,7 @@ const struct address_space_operations f2fs_node_aops = {
 	.writepages	= f2fs_write_node_pages,
 	.dirty_folio	= f2fs_dirty_node_folio,
 	.invalidate_folio = f2fs_invalidate_folio,
-	.releasepage	= f2fs_release_page,
+	.release_folio	= f2fs_release_folio,
 #ifdef CONFIG_MIGRATION
 	.migratepage	= f2fs_migrate_page,
 #endif
-- 
2.34.1

[PATCH 11/26] gfs2: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio throughout gfs2_release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/gfs2/aops.c    | 42 ++++++++++++++++++++++--------------------
 fs/gfs2/inode.h   |  2 +-
 fs/gfs2/meta_io.c |  4 ++--
 3 files changed, 25 insertions(+), 23 deletions(-)

diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 3d6c5c5eb4f1..95a674d70c04 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -691,38 +691,40 @@ static void gfs2_invalidate_folio(struct folio *folio, size_t offset,
 }
 
 /**
- * gfs2_releasepage - free the metadata associated with a page
- * @page: the page that's being released
+ * gfs2_release_folio - free the metadata associated with a folio
+ * @folio: the folio that's being released
  * @gfp_mask: passed from Linux VFS, ignored by us
  *
- * Calls try_to_free_buffers() to free the buffers and put the page if the
+ * Calls try_to_free_buffers() to free the buffers and put the folio if the
  * buffers can be released.
  *
- * Returns: 1 if the page was put or else 0
+ * Returns: true if the folio was put or else false
  */
 
-int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
+bool gfs2_release_folio(struct folio *folio, gfp_t gfp_mask)
 {
-	struct address_space *mapping = page->mapping;
+	struct address_space *mapping = folio->mapping;
 	struct gfs2_sbd *sdp = gfs2_mapping2sbd(mapping);
 	struct buffer_head *bh, *head;
 	struct gfs2_bufdata *bd;
 
-	if (!page_has_buffers(page))
-		return 0;
+	head = folio_buffers(folio);
+	if (!head)
+		return false;
 
 	/*
-	 * From xfs_vm_releasepage: mm accommodates an old ext3 case where
-	 * clean pages might not have had the dirty bit cleared.  Thus, it can
-	 * send actual dirty pages to ->releasepage() via shrink_active_list().
+	 * mm accommodates an old ext3 case where clean folios might
+	 * not have had the dirty bit cleared.	Thus, it can send actual
+	 * dirty folios to ->release_folio() via shrink_active_list().
 	 *
-	 * As a workaround, we skip pages that contain dirty buffers below.
-	 * Once ->releasepage isn't called on dirty pages anymore, we can warn
-	 * on dirty buffers like we used to here again.
+	 * As a workaround, we skip folios that contain dirty buffers
+	 * below.  Once ->release_folio isn't called on dirty folios
+	 * anymore, we can warn on dirty buffers like we used to here
+	 * again.
 	 */
 
 	gfs2_log_lock(sdp);
-	head = bh = page_buffers(page);
+	bh = head;
 	do {
 		if (atomic_read(&bh->b_count))
 			goto cannot_release;
@@ -732,9 +734,9 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 		if (buffer_dirty(bh) || WARN_ON(buffer_pinned(bh)))
 			goto cannot_release;
 		bh = bh->b_this_page;
-	} while(bh != head);
+	} while (bh != head);
 
-	head = bh = page_buffers(page);
+	bh = head;
 	do {
 		bd = bh->b_private;
 		if (bd) {
@@ -755,11 +757,11 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask)
 	} while (bh != head);
 	gfs2_log_unlock(sdp);
 
-	return try_to_free_buffers(page);
+	return try_to_free_buffers(&folio->page);
 
 cannot_release:
 	gfs2_log_unlock(sdp);
-	return 0;
+	return false;
 }
 
 static const struct address_space_operations gfs2_aops = {
@@ -785,7 +787,7 @@ static const struct address_space_operations gfs2_jdata_aops = {
 	.dirty_folio = jdata_dirty_folio,
 	.bmap = gfs2_bmap,
 	.invalidate_folio = gfs2_invalidate_folio,
-	.releasepage = gfs2_releasepage,
+	.release_folio = gfs2_release_folio,
 	.is_partially_uptodate = block_is_partially_uptodate,
 	.error_remove_page = generic_error_remove_page,
 };
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 7b2c1f390db7..0264d514dda7 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -12,7 +12,7 @@
 #include <linux/mm.h>
 #include "util.h"
 
-extern int gfs2_releasepage(struct page *page, gfp_t gfp_mask);
+bool gfs2_release_folio(struct folio *folio, gfp_t gfp_mask);
 extern int gfs2_internal_read(struct gfs2_inode *ip,
 			      char *buf, loff_t *pos, unsigned size);
 extern void gfs2_set_aops(struct inode *inode);
diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c
index d8bd1d48bd78..868dcc71b581 100644
--- a/fs/gfs2/meta_io.c
+++ b/fs/gfs2/meta_io.c
@@ -92,14 +92,14 @@ const struct address_space_operations gfs2_meta_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
 	.writepage = gfs2_aspace_writepage,
-	.releasepage = gfs2_releasepage,
+	.release_folio = gfs2_release_folio,
 };
 
 const struct address_space_operations gfs2_rgrp_aops = {
 	.dirty_folio	= block_dirty_folio,
 	.invalidate_folio = block_invalidate_folio,
 	.writepage = gfs2_aspace_writepage,
-	.releasepage = gfs2_releasepage,
+	.release_folio = gfs2_release_folio,
 };
 
 /**
-- 
2.34.1

Re: [PATCH 11/26] gfs2: Convert to release_folio

[Bob Peterson]

On 5/8/22 3:32 PM, Matthew Wilcox (Oracle) wrote:
> Use a folio throughout gfs2_release_folio().
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> ---
The gfs2 portion looks good to me.

Reviewed-by: Bob Peterson <rpeterso@redhat.com>

Regards,

Bob Peterson

[PATCH 12/26] hfs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio throughout hfs_release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/hfs/inode.c | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index ba3ff9cd7cfc..86fd50e5fccb 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -69,14 +69,15 @@ static sector_t hfs_bmap(struct address_space *mapping, sector_t block)
 	return generic_block_bmap(mapping, block, hfs_get_block);
 }
 
-static int hfs_releasepage(struct page *page, gfp_t mask)
+static bool hfs_release_folio(struct folio *folio, gfp_t mask)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct super_block *sb = inode->i_sb;
 	struct hfs_btree *tree;
 	struct hfs_bnode *node;
 	u32 nidx;
-	int i, res = 1;
+	int i;
+	bool res = true;
 
 	switch (inode->i_ino) {
 	case HFS_EXT_CNID:
@@ -87,27 +88,27 @@ static int hfs_releasepage(struct page *page, gfp_t mask)
 		break;
 	default:
 		BUG();
-		return 0;
+		return false;
 	}
 
 	if (!tree)
-		return 0;
+		return false;
 
 	if (tree->node_size >= PAGE_SIZE) {
-		nidx = page->index >> (tree->node_size_shift - PAGE_SHIFT);
+		nidx = folio->index >> (tree->node_size_shift - PAGE_SHIFT);
 		spin_lock(&tree->hash_lock);
 		node = hfs_bnode_findhash(tree, nidx);
 		if (!node)
 			;
 		else if (atomic_read(&node->refcnt))
-			res = 0;
+			res = false;
 		if (res && node) {
 			hfs_bnode_unhash(node);
 			hfs_bnode_free(node);
 		}
 		spin_unlock(&tree->hash_lock);
 	} else {
-		nidx = page->index << (PAGE_SHIFT - tree->node_size_shift);
+		nidx = folio->index << (PAGE_SHIFT - tree->node_size_shift);
 		i = 1 << (PAGE_SHIFT - tree->node_size_shift);
 		spin_lock(&tree->hash_lock);
 		do {
@@ -115,7 +116,7 @@ static int hfs_releasepage(struct page *page, gfp_t mask)
 			if (!node)
 				continue;
 			if (atomic_read(&node->refcnt)) {
-				res = 0;
+				res = false;
 				break;
 			}
 			hfs_bnode_unhash(node);
@@ -123,7 +124,7 @@ static int hfs_releasepage(struct page *page, gfp_t mask)
 		} while (--i && nidx < tree->node_count);
 		spin_unlock(&tree->hash_lock);
 	}
-	return res ? try_to_free_buffers(page) : 0;
+	return res ? try_to_free_buffers(&folio->page) : false;
 }
 
 static ssize_t hfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
@@ -165,7 +166,7 @@ const struct address_space_operations hfs_btree_aops = {
 	.write_begin	= hfs_write_begin,
 	.write_end	= generic_write_end,
 	.bmap		= hfs_bmap,
-	.releasepage	= hfs_releasepage,
+	.release_folio	= hfs_release_folio,
 };
 
 const struct address_space_operations hfs_aops = {
-- 
2.34.1

[PATCH 13/26] hfsplus: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use a folio throughout hfsplus_release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/hfsplus/inode.c | 23 ++++++++++++-----------
 1 file changed, 12 insertions(+), 11 deletions(-)

diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index 982b34eefec7..f723e0e91d51 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -63,14 +63,15 @@ static sector_t hfsplus_bmap(struct address_space *mapping, sector_t block)
 	return generic_block_bmap(mapping, block, hfsplus_get_block);
 }
 
-static int hfsplus_releasepage(struct page *page, gfp_t mask)
+static bool hfsplus_release_folio(struct folio *folio, gfp_t mask)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct super_block *sb = inode->i_sb;
 	struct hfs_btree *tree;
 	struct hfs_bnode *node;
 	u32 nidx;
-	int i, res = 1;
+	int i;
+	bool res = true;
 
 	switch (inode->i_ino) {
 	case HFSPLUS_EXT_CNID:
@@ -84,26 +85,26 @@ static int hfsplus_releasepage(struct page *page, gfp_t mask)
 		break;
 	default:
 		BUG();
-		return 0;
+		return false;
 	}
 	if (!tree)
-		return 0;
+		return false;
 	if (tree->node_size >= PAGE_SIZE) {
-		nidx = page->index >>
+		nidx = folio->index >>
 			(tree->node_size_shift - PAGE_SHIFT);
 		spin_lock(&tree->hash_lock);
 		node = hfs_bnode_findhash(tree, nidx);
 		if (!node)
 			;
 		else if (atomic_read(&node->refcnt))
-			res = 0;
+			res = false;
 		if (res && node) {
 			hfs_bnode_unhash(node);
 			hfs_bnode_free(node);
 		}
 		spin_unlock(&tree->hash_lock);
 	} else {
-		nidx = page->index <<
+		nidx = folio->index <<
 			(PAGE_SHIFT - tree->node_size_shift);
 		i = 1 << (PAGE_SHIFT - tree->node_size_shift);
 		spin_lock(&tree->hash_lock);
@@ -112,7 +113,7 @@ static int hfsplus_releasepage(struct page *page, gfp_t mask)
 			if (!node)
 				continue;
 			if (atomic_read(&node->refcnt)) {
-				res = 0;
+				res = false;
 				break;
 			}
 			hfs_bnode_unhash(node);
@@ -120,7 +121,7 @@ static int hfsplus_releasepage(struct page *page, gfp_t mask)
 		} while (--i && nidx < tree->node_count);
 		spin_unlock(&tree->hash_lock);
 	}
-	return res ? try_to_free_buffers(page) : 0;
+	return res ? try_to_free_buffers(&folio->page) : false;
 }
 
 static ssize_t hfsplus_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
@@ -162,7 +163,7 @@ const struct address_space_operations hfsplus_btree_aops = {
 	.write_begin	= hfsplus_write_begin,
 	.write_end	= generic_write_end,
 	.bmap		= hfsplus_bmap,
-	.releasepage	= hfsplus_releasepage,
+	.release_folio	= hfsplus_release_folio,
 };
 
 const struct address_space_operations hfsplus_aops = {
-- 
2.34.1

[PATCH 14/26] jfs: Convert to release_folio

[Matthew Wilcox (Oracle)]

The use of folios should be pushed further down into jfs from here.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/jfs/jfs_metapage.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/fs/jfs/jfs_metapage.c b/fs/jfs/jfs_metapage.c
index 2fc78405b3f2..387652ae14c2 100644
--- a/fs/jfs/jfs_metapage.c
+++ b/fs/jfs/jfs_metapage.c
@@ -524,29 +524,29 @@ static int metapage_read_folio(struct file *fp, struct folio *folio)
 	return -EIO;
 }
 
-static int metapage_releasepage(struct page *page, gfp_t gfp_mask)
+static bool metapage_release_folio(struct folio *folio, gfp_t gfp_mask)
 {
 	struct metapage *mp;
-	int ret = 1;
+	bool ret = true;
 	int offset;
 
 	for (offset = 0; offset < PAGE_SIZE; offset += PSIZE) {
-		mp = page_to_mp(page, offset);
+		mp = page_to_mp(&folio->page, offset);
 
 		if (!mp)
 			continue;
 
-		jfs_info("metapage_releasepage: mp = 0x%p", mp);
+		jfs_info("metapage_release_folio: mp = 0x%p", mp);
 		if (mp->count || mp->nohomeok ||
 		    test_bit(META_dirty, &mp->flag)) {
 			jfs_info("count = %ld, nohomeok = %d", mp->count,
 				 mp->nohomeok);
-			ret = 0;
+			ret = false;
 			continue;
 		}
 		if (mp->lsn)
 			remove_from_logsync(mp);
-		remove_metapage(page, mp);
+		remove_metapage(&folio->page, mp);
 		INCREMENT(mpStat.pagefree);
 		free_metapage(mp);
 	}
@@ -560,13 +560,13 @@ static void metapage_invalidate_folio(struct folio *folio, size_t offset,
 
 	BUG_ON(folio_test_writeback(folio));
 
-	metapage_releasepage(&folio->page, 0);
+	metapage_release_folio(folio, 0);
 }
 
 const struct address_space_operations jfs_metapage_aops = {
 	.read_folio	= metapage_read_folio,
 	.writepage	= metapage_writepage,
-	.releasepage	= metapage_releasepage,
+	.release_folio	= metapage_release_folio,
 	.invalidate_folio = metapage_invalidate_folio,
 	.dirty_folio	= filemap_dirty_folio,
 };
-- 
2.34.1

[PATCH 15/26] nfs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use folios throughout the release_folio paths.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/nfs/file.c    | 22 +++++++++++-----------
 fs/nfs/fscache.h | 14 +++++++-------
 2 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 4f6d1f90b87f..d764b3ce7905 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -415,19 +415,19 @@ static void nfs_invalidate_folio(struct folio *folio, size_t offset,
 }
 
 /*
- * Attempt to release the private state associated with a page
- * - Called if either PG_private or PG_fscache is set on the page
- * - Caller holds page lock
- * - Return true (may release page) or false (may not)
+ * Attempt to release the private state associated with a folio
+ * - Called if either private or fscache flags are set on the folio
+ * - Caller holds folio lock
+ * - Return true (may release folio) or false (may not)
  */
-static int nfs_release_page(struct page *page, gfp_t gfp)
+static bool nfs_release_folio(struct folio *folio, gfp_t gfp)
 {
-	dfprintk(PAGECACHE, "NFS: release_page(%p)\n", page);
+	dfprintk(PAGECACHE, "NFS: release_folio(%p)\n", folio);
 
-	/* If PagePrivate() is set, then the page is not freeable */
-	if (PagePrivate(page))
-		return 0;
-	return nfs_fscache_release_page(page, gfp);
+	/* If the private flag is set, then the folio is not freeable */
+	if (folio_test_private(folio))
+		return false;
+	return nfs_fscache_release_folio(folio, gfp);
 }
 
 static void nfs_check_dirty_writeback(struct folio *folio,
@@ -522,7 +522,7 @@ const struct address_space_operations nfs_file_aops = {
 	.write_begin = nfs_write_begin,
 	.write_end = nfs_write_end,
 	.invalidate_folio = nfs_invalidate_folio,
-	.releasepage = nfs_release_page,
+	.release_folio = nfs_release_folio,
 	.direct_IO = nfs_direct_IO,
 #ifdef CONFIG_MIGRATION
 	.migratepage = nfs_migrate_page,
diff --git a/fs/nfs/fscache.h b/fs/nfs/fscache.h
index 4e980cc04779..2a37af880978 100644
--- a/fs/nfs/fscache.h
+++ b/fs/nfs/fscache.h
@@ -48,14 +48,14 @@ extern void nfs_fscache_release_file(struct inode *, struct file *);
 extern int __nfs_fscache_read_page(struct inode *, struct page *);
 extern void __nfs_fscache_write_page(struct inode *, struct page *);
 
-static inline int nfs_fscache_release_page(struct page *page, gfp_t gfp)
+static inline bool nfs_fscache_release_folio(struct folio *folio, gfp_t gfp)
 {
-	if (PageFsCache(page)) {
+	if (folio_test_fscache(folio)) {
 		if (current_is_kswapd() || !(gfp & __GFP_FS))
 			return false;
-		wait_on_page_fscache(page);
-		fscache_note_page_release(nfs_i_fscache(page->mapping->host));
-		nfs_inc_fscache_stats(page->mapping->host,
+		folio_wait_fscache(folio);
+		fscache_note_page_release(nfs_i_fscache(folio->mapping->host));
+		nfs_inc_fscache_stats(folio->mapping->host,
 				      NFSIOS_FSCACHE_PAGES_UNCACHED);
 	}
 	return true;
@@ -129,9 +129,9 @@ static inline void nfs_fscache_open_file(struct inode *inode,
 					 struct file *filp) {}
 static inline void nfs_fscache_release_file(struct inode *inode, struct file *file) {}
 
-static inline int nfs_fscache_release_page(struct page *page, gfp_t gfp)
+static inline bool nfs_fscache_release_folio(struct folio *folio, gfp_t gfp)
 {
-	return 1; /* True: may release page */
+	return true; /* may release folio */
 }
 static inline int nfs_fscache_read_page(struct inode *inode, struct page *page)
 {
-- 
2.34.1

[PATCH 16/26] nilfs2: Remove comment about releasepage

[Matthew Wilcox (Oracle)]

If we need a release_folio, we can add it back.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/nilfs2/inode.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c
index 26b8065401b0..538ca5473b0d 100644
--- a/fs/nilfs2/inode.c
+++ b/fs/nilfs2/inode.c
@@ -304,7 +304,6 @@ const struct address_space_operations nilfs_aops = {
 	.readahead		= nilfs_readahead,
 	.write_begin		= nilfs_write_begin,
 	.write_end		= nilfs_write_end,
-	/* .releasepage		= nilfs_releasepage, */
 	.invalidate_folio	= block_invalidate_folio,
 	.direct_IO		= nilfs_direct_IO,
 	.is_partially_uptodate  = block_is_partially_uptodate,
-- 
2.34.1

[PATCH 17/26] ocfs2: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use folios throughout the release_folio path.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ocfs2/aops.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 6b1679db9636..7d7b86ca078f 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -498,11 +498,11 @@ static sector_t ocfs2_bmap(struct address_space *mapping, sector_t block)
 	return status;
 }
 
-static int ocfs2_releasepage(struct page *page, gfp_t wait)
+static bool ocfs2_release_folio(struct folio *folio, gfp_t wait)
 {
-	if (!page_has_buffers(page))
-		return 0;
-	return try_to_free_buffers(page);
+	if (!folio_buffers(folio))
+		return false;
+	return try_to_free_buffers(&folio->page);
 }
 
 static void ocfs2_figure_cluster_boundaries(struct ocfs2_super *osb,
@@ -2463,7 +2463,7 @@ const struct address_space_operations ocfs2_aops = {
 	.bmap			= ocfs2_bmap,
 	.direct_IO		= ocfs2_direct_IO,
 	.invalidate_folio	= block_invalidate_folio,
-	.releasepage		= ocfs2_releasepage,
+	.release_folio		= ocfs2_release_folio,
 	.migratepage		= buffer_migrate_page,
 	.is_partially_uptodate	= block_is_partially_uptodate,
 	.error_remove_page	= generic_error_remove_page,
-- 
2.34.1

[PATCH 18/26] orangefs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use folios throughout the release_folio path.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/orangefs/inode.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 241ac21f527b..3509618e7b37 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -485,9 +485,9 @@ static void orangefs_invalidate_folio(struct folio *folio,
 	orangefs_launder_folio(folio);
 }
 
-static int orangefs_releasepage(struct page *page, gfp_t foo)
+static bool orangefs_release_folio(struct folio *folio, gfp_t foo)
 {
-	return !PagePrivate(page);
+	return !folio_test_private(folio);
 }
 
 static void orangefs_freepage(struct page *page)
@@ -636,7 +636,7 @@ static const struct address_space_operations orangefs_address_operations = {
 	.write_begin = orangefs_write_begin,
 	.write_end = orangefs_write_end,
 	.invalidate_folio = orangefs_invalidate_folio,
-	.releasepage = orangefs_releasepage,
+	.release_folio = orangefs_release_folio,
 	.freepage = orangefs_freepage,
 	.launder_folio = orangefs_launder_folio,
 	.direct_IO = orangefs_direct_IO,
-- 
2.34.1

[PATCH 19/26] reiserfs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use folios throughout the release_folio path.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/reiserfs/inode.c | 20 ++++++++++----------
 1 file changed, 10 insertions(+), 10 deletions(-)

diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 33a9555f77b9..9cf2e1420a74 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -3202,39 +3202,39 @@ static bool reiserfs_dirty_folio(struct address_space *mapping,
 }
 
 /*
- * Returns 1 if the page's buffers were dropped.  The page is locked.
+ * Returns true if the folio's buffers were dropped.  The folio is locked.
  *
  * Takes j_dirty_buffers_lock to protect the b_assoc_buffers list_heads
- * in the buffers at page_buffers(page).
+ * in the buffers at folio_buffers(folio).
  *
  * even in -o notail mode, we can't be sure an old mount without -o notail
  * didn't create files with tails.
  */
-static int reiserfs_releasepage(struct page *page, gfp_t unused_gfp_flags)
+static bool reiserfs_release_folio(struct folio *folio, gfp_t unused_gfp_flags)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct reiserfs_journal *j = SB_JOURNAL(inode->i_sb);
 	struct buffer_head *head;
 	struct buffer_head *bh;
-	int ret = 1;
+	bool ret = true;
 
-	WARN_ON(PageChecked(page));
+	WARN_ON(folio_test_checked(folio));
 	spin_lock(&j->j_dirty_buffers_lock);
-	head = page_buffers(page);
+	head = folio_buffers(folio);
 	bh = head;
 	do {
 		if (bh->b_private) {
 			if (!buffer_dirty(bh) && !buffer_locked(bh)) {
 				reiserfs_free_jh(bh);
 			} else {
-				ret = 0;
+				ret = false;
 				break;
 			}
 		}
 		bh = bh->b_this_page;
 	} while (bh != head);
 	if (ret)
-		ret = try_to_free_buffers(page);
+		ret = try_to_free_buffers(&folio->page);
 	spin_unlock(&j->j_dirty_buffers_lock);
 	return ret;
 }
@@ -3423,7 +3423,7 @@ const struct address_space_operations reiserfs_address_space_operations = {
 	.writepage = reiserfs_writepage,
 	.read_folio = reiserfs_read_folio,
 	.readahead = reiserfs_readahead,
-	.releasepage = reiserfs_releasepage,
+	.release_folio = reiserfs_release_folio,
 	.invalidate_folio = reiserfs_invalidate_folio,
 	.write_begin = reiserfs_write_begin,
 	.write_end = reiserfs_write_end,
-- 
2.34.1

[PATCH 20/26] ubifs: Convert to release_folio

[Matthew Wilcox (Oracle)]

Use folios throughout the release_folio path.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ubifs/file.c | 18 +++++++++---------
 1 file changed, 9 insertions(+), 9 deletions(-)

diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c
index 7cbf2edf8907..04ced154960f 100644
--- a/fs/ubifs/file.c
+++ b/fs/ubifs/file.c
@@ -1484,22 +1484,22 @@ static int ubifs_migrate_page(struct address_space *mapping,
 }
 #endif
 
-static int ubifs_releasepage(struct page *page, gfp_t unused_gfp_flags)
+static bool ubifs_release_folio(struct folio *folio, gfp_t unused_gfp_flags)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct ubifs_info *c = inode->i_sb->s_fs_info;
 
 	/*
 	 * An attempt to release a dirty page without budgeting for it - should
 	 * not happen.
 	 */
-	if (PageWriteback(page))
-		return 0;
-	ubifs_assert(c, PagePrivate(page));
+	if (folio_test_writeback(folio))
+		return false;
+	ubifs_assert(c, folio_test_private(folio));
 	ubifs_assert(c, 0);
-	detach_page_private(page);
-	ClearPageChecked(page);
-	return 1;
+	folio_detach_private(folio);
+	folio_clear_checked(folio);
+	return true;
 }
 
 /*
@@ -1652,7 +1652,7 @@ const struct address_space_operations ubifs_file_address_operations = {
 #ifdef CONFIG_MIGRATION
 	.migratepage	= ubifs_migrate_page,
 #endif
-	.releasepage    = ubifs_releasepage,
+	.release_folio    = ubifs_release_folio,
 };
 
 const struct inode_operations ubifs_file_inode_operations = {
-- 
2.34.1

[PATCH 21/26] fs: Remove last vestiges of releasepage

[Matthew Wilcox (Oracle)]

All users are now converted to release_folio

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 include/linux/fs.h         | 1 -
 include/linux/page-flags.h | 2 +-
 mm/filemap.c               | 2 --
 3 files changed, 1 insertion(+), 4 deletions(-)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index ad768f13f485..1cee64d9724b 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -356,7 +356,6 @@ struct address_space_operations {
 	sector_t (*bmap)(struct address_space *, sector_t);
 	void (*invalidate_folio) (struct folio *, size_t offset, size_t len);
 	bool (*release_folio)(struct folio *, gfp_t);
-	int (*releasepage) (struct page *, gfp_t);
 	void (*freepage)(struct page *);
 	ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 	/*
diff --git a/include/linux/page-flags.h b/include/linux/page-flags.h
index 9d8eeaa67d05..af10149a6c31 100644
--- a/include/linux/page-flags.h
+++ b/include/linux/page-flags.h
@@ -516,7 +516,7 @@ PAGEFLAG(SwapBacked, swapbacked, PF_NO_TAIL)
 /*
  * Private page markings that may be used by the filesystem that owns the page
  * for its own purposes.
- * - PG_private and PG_private_2 cause releasepage() and co to be invoked
+ * - PG_private and PG_private_2 cause release_folio() and co to be invoked
  */
 PAGEFLAG(Private, private, PF_ANY)
 PAGEFLAG(Private2, private_2, PF_ANY) TESTSCFLAG(Private2, private_2, PF_ANY)
diff --git a/mm/filemap.c b/mm/filemap.c
index 78e4a7dc3a56..ee892853a214 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3957,8 +3957,6 @@ bool filemap_release_folio(struct folio *folio, gfp_t gfp)
 
 	if (mapping && mapping->a_ops->release_folio)
 		return mapping->a_ops->release_folio(folio, gfp);
-	if (mapping && mapping->a_ops->releasepage)
-		return mapping->a_ops->releasepage(&folio->page, gfp);
 	return try_to_free_buffers(&folio->page);
 }
 EXPORT_SYMBOL(filemap_release_folio);
-- 
2.34.1

[PATCH 22/26] reiserfs: Convert release_buffer_page() to use a folio

[Matthew Wilcox (Oracle)]

Saves 671 bytes from an allmodconfig build (!)

Function                                     old     new   delta
release_buffer_page                         1617     946    -671
Total: Before=67656, After=66985, chg -0.99%

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/reiserfs/journal.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)

diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c
index b5b6f6201bed..99ba495b0f28 100644
--- a/fs/reiserfs/journal.c
+++ b/fs/reiserfs/journal.c
@@ -601,14 +601,14 @@ static int journal_list_still_alive(struct super_block *s,
  */
 static void release_buffer_page(struct buffer_head *bh)
 {
-	struct page *page = bh->b_page;
-	if (!page->mapping && trylock_page(page)) {
-		get_page(page);
+	struct folio *folio = page_folio(bh->b_page);
+	if (!folio->mapping && folio_trylock(folio)) {
+		folio_get(folio);
 		put_bh(bh);
-		if (!page->mapping)
-			try_to_free_buffers(page);
-		unlock_page(page);
-		put_page(page);
+		if (!folio->mapping)
+			try_to_free_buffers(&folio->page);
+		folio_unlock(folio);
+		folio_put(folio);
 	} else {
 		put_bh(bh);
 	}
-- 
2.34.1

[PATCH 23/26] jbd2: Convert jbd2_journal_try_to_free_buffers to take a folio

[Matthew Wilcox (Oracle)]

Also convert it to return a bool since it's called from release_folio().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/ext4/inode.c       |  2 +-
 fs/jbd2/transaction.c | 12 ++++++------
 include/linux/jbd2.h  |  2 +-
 3 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 52c46ac5bc8a..943937cb5302 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3253,7 +3253,7 @@ static bool ext4_release_folio(struct folio *folio, gfp_t wait)
 	if (folio_test_checked(folio))
 		return false;
 	if (journal)
-		return jbd2_journal_try_to_free_buffers(journal, &folio->page);
+		return jbd2_journal_try_to_free_buffers(journal, folio);
 	else
 		return try_to_free_buffers(&folio->page);
 }
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index fcb9175016a5..ee33d277d51e 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -2143,17 +2143,17 @@ __journal_try_to_free_buffer(journal_t *journal, struct buffer_head *bh)
  * cannot happen because we never reallocate freed data as metadata
  * while the data is part of a transaction.  Yes?
  *
- * Return 0 on failure, 1 on success
+ * Return false on failure, true on success
  */
-int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page)
+bool jbd2_journal_try_to_free_buffers(journal_t *journal, struct folio *folio)
 {
 	struct buffer_head *head;
 	struct buffer_head *bh;
-	int ret = 0;
+	bool ret = false;
 
-	J_ASSERT(PageLocked(page));
+	J_ASSERT(folio_test_locked(folio));
 
-	head = page_buffers(page);
+	head = folio_buffers(folio);
 	bh = head;
 	do {
 		struct journal_head *jh;
@@ -2175,7 +2175,7 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page)
 			goto busy;
 	} while ((bh = bh->b_this_page) != head);
 
-	ret = try_to_free_buffers(page);
+	ret = try_to_free_buffers(&folio->page);
 busy:
 	return ret;
 }
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index de9536680b2b..e79d6e0b14e8 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -1529,7 +1529,7 @@ extern int	 jbd2_journal_dirty_metadata (handle_t *, struct buffer_head *);
 extern int	 jbd2_journal_forget (handle_t *, struct buffer_head *);
 int jbd2_journal_invalidate_folio(journal_t *, struct folio *,
 					size_t offset, size_t length);
-extern int	 jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page);
+bool jbd2_journal_try_to_free_buffers(journal_t *journal, struct folio *folio);
 extern int	 jbd2_journal_stop(handle_t *);
 extern int	 jbd2_journal_flush(journal_t *journal, unsigned int flags);
 extern void	 jbd2_journal_lock_updates (journal_t *);
-- 
2.34.1

Re: [PATCH 23/26] jbd2: Convert jbd2_journal_try_to_free_buffers to take a folio

[Theodore Ts'o]

On Sun, May 08, 2022 at 09:32:44PM +0100, Matthew Wilcox (Oracle) wrote:
> Also convert it to return a bool since it's called from release_folio().
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Acked-by: Theodore Ts'o <tytso@mit.edu>

[PATCH 24/26] jbd2: Convert release_buffer_page() to use a folio

[Matthew Wilcox (Oracle)]

Saves a few calls to compound_head().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/jbd2/commit.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index ac7f067b7bdd..2f37108da0ec 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -62,6 +62,7 @@ static void journal_end_buffer_io_sync(struct buffer_head *bh, int uptodate)
  */
 static void release_buffer_page(struct buffer_head *bh)
 {
+	struct folio *folio;
 	struct page *page;
 
 	if (buffer_dirty(bh))
@@ -71,18 +72,19 @@ static void release_buffer_page(struct buffer_head *bh)
 	page = bh->b_page;
 	if (!page)
 		goto nope;
-	if (page->mapping)
+	folio = page_folio(page);
+	if (folio->mapping)
 		goto nope;
 
 	/* OK, it's a truncated page */
-	if (!trylock_page(page))
+	if (!folio_trylock(folio))
 		goto nope;
 
-	get_page(page);
+	folio_get(folio);
 	__brelse(bh);
-	try_to_free_buffers(page);
-	unlock_page(page);
-	put_page(page);
+	try_to_free_buffers(&folio->page);
+	folio_unlock(folio);
+	folio_put(folio);
 	return;
 
 nope:
-- 
2.34.1

Re: [PATCH 24/26] jbd2: Convert release_buffer_page() to use a folio

[Theodore Ts'o]

On Sun, May 08, 2022 at 09:32:45PM +0100, Matthew Wilcox (Oracle) wrote:
> Saves a few calls to compound_head().
>
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>

Acked-by: Theodore Ts'o <tytso@mit.edu>

... although I have one question:

> -	get_page(page);
> +	folio_get(folio);
>  	__brelse(bh);
> -	try_to_free_buffers(page);
> -	unlock_page(page);
> -	put_page(page);
> +	try_to_free_buffers(&folio->page);

The patches in this series seem to assume that the folio contains only
a single page.  Or at least, a *lot* more detailed auditing to make
sure the right thing would happen if a page happens to be a member of
a folio with more than a single page.

Should we be adding some kind of WARN_ON to check this particular
assumption?

					- Ted

Re: [PATCH 24/26] jbd2: Convert release_buffer_page() to use a folio

[Matthew Wilcox]

On Mon, May 09, 2022 at 09:23:10AM -0400, Theodore Ts'o wrote:
> On Sun, May 08, 2022 at 09:32:45PM +0100, Matthew Wilcox (Oracle) wrote:
> > Saves a few calls to compound_head().
> >
> > Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> 
> Acked-by: Theodore Ts'o <tytso@mit.edu>
> 
> ... although I have one question:
> 
> > -	get_page(page);
> > +	folio_get(folio);
> >  	__brelse(bh);
> > -	try_to_free_buffers(page);
> > -	unlock_page(page);
> > -	put_page(page);
> > +	try_to_free_buffers(&folio->page);
> 
> The patches in this series seem to assume that the folio contains only
> a single page.  Or at least, a *lot* more detailed auditing to make
> sure the right thing would happen if a page happens to be a member of
> a folio with more than a single page.
> 
> Should we be adding some kind of WARN_ON to check this particular
> assumption?

You're right, a lot of places assume a single-page folio.  Since
filesystems need to explicitly enable large folios (by calling
mapping_set_large_folios()), I haven't been adding assertions to
aops entry points.  I have been adding them to some common utility
functions in fs/buffer.c eg block_read_full_folio has:
        VM_BUG_ON_FOLIO(folio_test_large(folio), folio);

Other functions in fs/buffer.c look like they should handle large folios
OK, eg buffer_check_dirty_writeback() and block_dirty_folio() and then
others have existing assertions which are going to trigger if you try
to use them with large folios, like __block_write_begin_int() has
        BUG_ON(from > PAGE_SIZE);
and I haven't bothered to add assertions to either of those classes of
function.

JBD2 somewhat straddles a line between being considered "generic utility
functions" and "part of a filesystem".  I could go either way on adding
assertions to document that this hasn't been audited for large folios.
Although seeing '&folio->page' is always a sign that you should probably
audit the function.

By the end of this series, it seems to me that jbd2's
release_buffer_page() is actually large-folio-safe.  There's nothing
in it that relies on the size of the folio, it calls functions that
are large-folio-safe and the only actual use of 'page' in it is to
get bh->b_page ... which should probably be in a union with a new
->b_folio so we don't need to muck around with calling page_folio().

[PATCH 25/26] fs: Change try_to_free_buffers() to take a folio

[Matthew Wilcox (Oracle)]

All but two of the callers already have a folio; pass a folio into
try_to_free_buffers().  This removes the last user of cancel_dirty_page()
so remove that wrapper function too.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/buffer.c                 | 42 ++++++++++++++++++-------------------
 fs/ext4/inode.c             |  2 +-
 fs/gfs2/aops.c              |  2 +-
 fs/hfs/inode.c              |  2 +-
 fs/hfsplus/inode.c          |  2 +-
 fs/jbd2/commit.c            |  2 +-
 fs/jbd2/transaction.c       |  4 ++--
 fs/mpage.c                  |  2 +-
 fs/ocfs2/aops.c             |  2 +-
 fs/reiserfs/inode.c         |  2 +-
 fs/reiserfs/journal.c       |  2 +-
 include/linux/buffer_head.h |  4 ++--
 include/linux/pagemap.h     |  4 ----
 mm/filemap.c                |  2 +-
 mm/migrate.c                |  2 +-
 mm/vmscan.c                 |  2 +-
 16 files changed, 37 insertions(+), 41 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index 786ef5b98c80..701af0035802 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -955,7 +955,7 @@ grow_dev_page(struct block_device *bdev, sector_t block,
 						size);
 			goto done;
 		}
-		if (!try_to_free_buffers(page))
+		if (!try_to_free_buffers(page_folio(page)))
 			goto failed;
 	}
 
@@ -3155,20 +3155,20 @@ int sync_dirty_buffer(struct buffer_head *bh)
 EXPORT_SYMBOL(sync_dirty_buffer);
 
 /*
- * try_to_free_buffers() checks if all the buffers on this particular page
+ * try_to_free_buffers() checks if all the buffers on this particular folio
  * are unused, and releases them if so.
  *
  * Exclusion against try_to_free_buffers may be obtained by either
- * locking the page or by holding its mapping's private_lock.
+ * locking the folio or by holding its mapping's private_lock.
  *
- * If the page is dirty but all the buffers are clean then we need to
- * be sure to mark the page clean as well.  This is because the page
+ * If the folio is dirty but all the buffers are clean then we need to
+ * be sure to mark the folio clean as well.  This is because the folio
  * may be against a block device, and a later reattachment of buffers
- * to a dirty page will set *all* buffers dirty.  Which would corrupt
+ * to a dirty folio will set *all* buffers dirty.  Which would corrupt
  * filesystem data on the same device.
  *
- * The same applies to regular filesystem pages: if all the buffers are
- * clean then we set the page clean and proceed.  To do that, we require
+ * The same applies to regular filesystem folios: if all the buffers are
+ * clean then we set the folio clean and proceed.  To do that, we require
  * total exclusion from block_dirty_folio().  That is obtained with
  * private_lock.
  *
@@ -3207,40 +3207,40 @@ drop_buffers(struct page *page, struct buffer_head **buffers_to_free)
 	return 0;
 }
 
-int try_to_free_buffers(struct page *page)
+bool try_to_free_buffers(struct folio *folio)
 {
-	struct address_space * const mapping = page->mapping;
+	struct address_space * const mapping = folio->mapping;
 	struct buffer_head *buffers_to_free = NULL;
-	int ret = 0;
+	bool ret = 0;
 
-	BUG_ON(!PageLocked(page));
-	if (PageWriteback(page))
-		return 0;
+	BUG_ON(!folio_test_locked(folio));
+	if (folio_test_writeback(folio))
+		return false;
 
 	if (mapping == NULL) {		/* can this still happen? */
-		ret = drop_buffers(page, &buffers_to_free);
+		ret = drop_buffers(&folio->page, &buffers_to_free);
 		goto out;
 	}
 
 	spin_lock(&mapping->private_lock);
-	ret = drop_buffers(page, &buffers_to_free);
+	ret = drop_buffers(&folio->page, &buffers_to_free);
 
 	/*
 	 * If the filesystem writes its buffers by hand (eg ext3)
-	 * then we can have clean buffers against a dirty page.  We
-	 * clean the page here; otherwise the VM will never notice
+	 * then we can have clean buffers against a dirty folio.  We
+	 * clean the folio here; otherwise the VM will never notice
 	 * that the filesystem did any IO at all.
 	 *
 	 * Also, during truncate, discard_buffer will have marked all
-	 * the page's buffers clean.  We discover that here and clean
-	 * the page also.
+	 * the folio's buffers clean.  We discover that here and clean
+	 * the folio also.
 	 *
 	 * private_lock must be held over this entire operation in order
 	 * to synchronise against block_dirty_folio and prevent the
 	 * dirty bit from being lost.
 	 */
 	if (ret)
-		cancel_dirty_page(page);
+		folio_cancel_dirty(folio);
 	spin_unlock(&mapping->private_lock);
 out:
 	if (buffers_to_free) {
diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
index 943937cb5302..987ea77e672d 100644
--- a/fs/ext4/inode.c
+++ b/fs/ext4/inode.c
@@ -3255,7 +3255,7 @@ static bool ext4_release_folio(struct folio *folio, gfp_t wait)
 	if (journal)
 		return jbd2_journal_try_to_free_buffers(journal, folio);
 	else
-		return try_to_free_buffers(&folio->page);
+		return try_to_free_buffers(folio);
 }
 
 static bool ext4_inode_datasync_dirty(struct inode *inode)
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 95a674d70c04..106e90a36583 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -757,7 +757,7 @@ bool gfs2_release_folio(struct folio *folio, gfp_t gfp_mask)
 	} while (bh != head);
 	gfs2_log_unlock(sdp);
 
-	return try_to_free_buffers(&folio->page);
+	return try_to_free_buffers(folio);
 
 cannot_release:
 	gfs2_log_unlock(sdp);
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 86fd50e5fccb..c4526f16355d 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -124,7 +124,7 @@ static bool hfs_release_folio(struct folio *folio, gfp_t mask)
 		} while (--i && nidx < tree->node_count);
 		spin_unlock(&tree->hash_lock);
 	}
-	return res ? try_to_free_buffers(&folio->page) : false;
+	return res ? try_to_free_buffers(folio) : false;
 }
 
 static ssize_t hfs_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
diff --git a/fs/hfsplus/inode.c b/fs/hfsplus/inode.c
index f723e0e91d51..aeab83ed1c9c 100644
--- a/fs/hfsplus/inode.c
+++ b/fs/hfsplus/inode.c
@@ -121,7 +121,7 @@ static bool hfsplus_release_folio(struct folio *folio, gfp_t mask)
 		} while (--i && nidx < tree->node_count);
 		spin_unlock(&tree->hash_lock);
 	}
-	return res ? try_to_free_buffers(&folio->page) : false;
+	return res ? try_to_free_buffers(folio) : false;
 }
 
 static ssize_t hfsplus_direct_IO(struct kiocb *iocb, struct iov_iter *iter)
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 2f37108da0ec..eb315e81f1a6 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -82,7 +82,7 @@ static void release_buffer_page(struct buffer_head *bh)
 
 	folio_get(folio);
 	__brelse(bh);
-	try_to_free_buffers(&folio->page);
+	try_to_free_buffers(folio);
 	folio_unlock(folio);
 	folio_put(folio);
 	return;
diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index ee33d277d51e..e49bb0938376 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -2175,7 +2175,7 @@ bool jbd2_journal_try_to_free_buffers(journal_t *journal, struct folio *folio)
 			goto busy;
 	} while ((bh = bh->b_this_page) != head);
 
-	ret = try_to_free_buffers(&folio->page);
+	ret = try_to_free_buffers(folio);
 busy:
 	return ret;
 }
@@ -2482,7 +2482,7 @@ int jbd2_journal_invalidate_folio(journal_t *journal, struct folio *folio,
 	} while (bh != head);
 
 	if (!partial_page) {
-		if (may_free && try_to_free_buffers(&folio->page))
+		if (may_free && try_to_free_buffers(folio))
 			J_ASSERT(!folio_buffers(folio));
 	}
 	return 0;
diff --git a/fs/mpage.c b/fs/mpage.c
index 6df9c3aa5728..0d25f44f5707 100644
--- a/fs/mpage.c
+++ b/fs/mpage.c
@@ -431,7 +431,7 @@ static void clean_buffers(struct page *page, unsigned first_unmapped)
 	 * disk before we reach the platter.
 	 */
 	if (buffer_heads_over_limit && PageUptodate(page))
-		try_to_free_buffers(page);
+		try_to_free_buffers(page_folio(page));
 }
 
 /*
diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c
index 7d7b86ca078f..35d40a67204c 100644
--- a/fs/ocfs2/aops.c
+++ b/fs/ocfs2/aops.c
@@ -502,7 +502,7 @@ static bool ocfs2_release_folio(struct folio *folio, gfp_t wait)
 {
 	if (!folio_buffers(folio))
 		return false;
-	return try_to_free_buffers(&folio->page);
+	return try_to_free_buffers(folio);
 }
 
 static void ocfs2_figure_cluster_boundaries(struct ocfs2_super *osb,
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 9cf2e1420a74..0cffe054b78e 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -3234,7 +3234,7 @@ static bool reiserfs_release_folio(struct folio *folio, gfp_t unused_gfp_flags)
 		bh = bh->b_this_page;
 	} while (bh != head);
 	if (ret)
-		ret = try_to_free_buffers(&folio->page);
+		ret = try_to_free_buffers(folio);
 	spin_unlock(&j->j_dirty_buffers_lock);
 	return ret;
 }
diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c
index 99ba495b0f28..d8cc9a366124 100644
--- a/fs/reiserfs/journal.c
+++ b/fs/reiserfs/journal.c
@@ -606,7 +606,7 @@ static void release_buffer_page(struct buffer_head *bh)
 		folio_get(folio);
 		put_bh(bh);
 		if (!folio->mapping)
-			try_to_free_buffers(&folio->page);
+			try_to_free_buffers(folio);
 		folio_unlock(folio);
 		folio_put(folio);
 	} else {
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 31d82fd9abe8..c9d1463bb20f 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -158,7 +158,7 @@ void mark_buffer_write_io_error(struct buffer_head *bh);
 void touch_buffer(struct buffer_head *bh);
 void set_bh_page(struct buffer_head *bh,
 		struct page *page, unsigned long offset);
-int try_to_free_buffers(struct page *);
+bool try_to_free_buffers(struct folio *);
 struct buffer_head *alloc_page_buffers(struct page *page, unsigned long size,
 		bool retry);
 void create_empty_buffers(struct page *, unsigned long,
@@ -402,7 +402,7 @@ bool block_dirty_folio(struct address_space *mapping, struct folio *folio);
 #else /* CONFIG_BLOCK */
 
 static inline void buffer_init(void) {}
-static inline int try_to_free_buffers(struct page *page) { return 1; }
+static inline bool try_to_free_buffers(struct folio *folio) { return true; }
 static inline int inode_has_buffers(struct inode *inode) { return 0; }
 static inline void invalidate_inode_buffers(struct inode *inode) {}
 static inline int remove_inode_buffers(struct inode *inode) { return 1; }
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index 831b28dab01a..82dfb279e0c4 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -1067,10 +1067,6 @@ static inline void folio_cancel_dirty(struct folio *folio)
 	if (folio_test_dirty(folio))
 		__folio_cancel_dirty(folio);
 }
-static inline void cancel_dirty_page(struct page *page)
-{
-	folio_cancel_dirty(page_folio(page));
-}
 bool folio_clear_dirty_for_io(struct folio *folio);
 bool clear_page_dirty_for_io(struct page *page);
 void folio_invalidate(struct folio *folio, size_t offset, size_t length);
diff --git a/mm/filemap.c b/mm/filemap.c
index ee892853a214..d335a154a0d9 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3957,6 +3957,6 @@ bool filemap_release_folio(struct folio *folio, gfp_t gfp)
 
 	if (mapping && mapping->a_ops->release_folio)
 		return mapping->a_ops->release_folio(folio, gfp);
-	return try_to_free_buffers(&folio->page);
+	return try_to_free_buffers(folio);
 }
 EXPORT_SYMBOL(filemap_release_folio);
diff --git a/mm/migrate.c b/mm/migrate.c
index 6c31ee1e1c9b..21d82636c291 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1013,7 +1013,7 @@ static int __unmap_and_move(struct page *page, struct page *newpage,
 	if (!page->mapping) {
 		VM_BUG_ON_PAGE(PageAnon(page), page);
 		if (page_has_private(page)) {
-			try_to_free_buffers(page);
+			try_to_free_buffers(folio);
 			goto out_unlock_both;
 		}
 	} else if (page_mapped(page)) {
diff --git a/mm/vmscan.c b/mm/vmscan.c
index 27851232e00c..f3f7ce2c4068 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1181,7 +1181,7 @@ static pageout_t pageout(struct folio *folio, struct address_space *mapping)
 		 * folio->mapping == NULL while being dirty with clean buffers.
 		 */
 		if (folio_test_private(folio)) {
-			if (try_to_free_buffers(&folio->page)) {
+			if (try_to_free_buffers(folio)) {
 				folio_clear_dirty(folio);
 				pr_info("%s: orphaned folio\n", __func__);
 				return PAGE_CLEAN;
-- 
2.34.1

[PATCH 26/26] fs: Convert drop_buffers() to use a folio

[Matthew Wilcox (Oracle)]

All callers now have a folio.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/buffer.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/fs/buffer.c b/fs/buffer.c
index 701af0035802..898c7f301b1b 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -3180,10 +3180,10 @@ static inline int buffer_busy(struct buffer_head *bh)
 		(bh->b_state & ((1 << BH_Dirty) | (1 << BH_Lock)));
 }
 
-static int
-drop_buffers(struct page *page, struct buffer_head **buffers_to_free)
+static bool
+drop_buffers(struct folio *folio, struct buffer_head **buffers_to_free)
 {
-	struct buffer_head *head = page_buffers(page);
+	struct buffer_head *head = folio_buffers(folio);
 	struct buffer_head *bh;
 
 	bh = head;
@@ -3201,10 +3201,10 @@ drop_buffers(struct page *page, struct buffer_head **buffers_to_free)
 		bh = next;
 	} while (bh != head);
 	*buffers_to_free = head;
-	detach_page_private(page);
-	return 1;
+	folio_detach_private(folio);
+	return true;
 failed:
-	return 0;
+	return false;
 }
 
 bool try_to_free_buffers(struct folio *folio)
@@ -3218,12 +3218,12 @@ bool try_to_free_buffers(struct folio *folio)
 		return false;
 
 	if (mapping == NULL) {		/* can this still happen? */
-		ret = drop_buffers(&folio->page, &buffers_to_free);
+		ret = drop_buffers(folio, &buffers_to_free);
 		goto out;
 	}
 
 	spin_lock(&mapping->private_lock);
-	ret = drop_buffers(&folio->page, &buffers_to_free);
+	ret = drop_buffers(folio, &buffers_to_free);
 
 	/*
 	 * If the filesystem writes its buffers by hand (eg ext3)
-- 
2.34.1

Re: [PATCH 00/26] Convert aops->releasepage to aops->release_folio

[Jeff Layton]

On Sun, 2022-05-08 at 21:32 +0100, Matthew Wilcox (Oracle) wrote:
> Convert all filesystems.  The last five patches are further cleanups
> that are enabled by this change.
> 
> Matthew Wilcox (Oracle) (26):
>   fs: Add aops->release_folio
>   iomap: Convert to release_folio
>   9p: Convert to release_folio
>   afs: Convert to release_folio
>   btrfs: Convert to release_folio
>   ceph: Convert to release_folio
>   cifs: Convert to release_folio
>   erofs: Convert to release_folio
>   ext4: Convert to release_folio
>   f2fs: Convert to release_folio
>   gfs2: Convert to release_folio
>   hfs: Convert to release_folio
>   hfsplus: Convert to release_folio
>   jfs: Convert to release_folio
>   nfs: Convert to release_folio
>   nilfs2: Remove comment about releasepage
>   ocfs2: Convert to release_folio
>   orangefs: Convert to release_folio
>   reiserfs: Convert to release_folio
>   ubifs: Convert to release_folio
>   fs: Remove last vestiges of releasepage
>   reiserfs: Convert release_buffer_page() to use a folio
>   jbd2: Convert jbd2_journal_try_to_free_buffers to take a folio
>   jbd2: Convert release_buffer_page() to use a folio
>   fs: Change try_to_free_buffers() to take a folio
>   fs: Convert drop_buffers() to use a folio
> 
>  .../filesystems/caching/netfs-api.rst         |  4 +-
>  Documentation/filesystems/locking.rst         | 14 ++---
>  Documentation/filesystems/vfs.rst             | 45 ++++++++--------
>  fs/9p/vfs_addr.c                              | 17 +++---
>  fs/afs/dir.c                                  |  7 ++-
>  fs/afs/file.c                                 | 11 ++--
>  fs/afs/internal.h                             |  2 +-
>  fs/btrfs/disk-io.c                            | 12 ++---
>  fs/btrfs/extent_io.c                          | 14 ++---
>  fs/btrfs/file.c                               |  2 +-
>  fs/btrfs/inode.c                              | 24 ++++-----
>  fs/buffer.c                                   | 54 +++++++++----------
>  fs/ceph/addr.c                                | 24 ++++-----
>  fs/cifs/file.c                                | 14 ++---
>  fs/erofs/super.c                              | 16 +++---
>  fs/ext4/inode.c                               | 20 +++----
>  fs/f2fs/checkpoint.c                          |  2 +-
>  fs/f2fs/compress.c                            |  2 +-
>  fs/f2fs/data.c                                | 32 +++++------
>  fs/f2fs/f2fs.h                                |  2 +-
>  fs/f2fs/node.c                                |  2 +-
>  fs/gfs2/aops.c                                | 44 +++++++--------
>  fs/gfs2/inode.h                               |  2 +-
>  fs/gfs2/meta_io.c                             |  4 +-
>  fs/hfs/inode.c                                | 23 ++++----
>  fs/hfsplus/inode.c                            | 23 ++++----
>  fs/iomap/buffered-io.c                        | 22 ++++----
>  fs/iomap/trace.h                              |  2 +-
>  fs/jbd2/commit.c                              | 14 ++---
>  fs/jbd2/transaction.c                         | 14 ++---
>  fs/jfs/jfs_metapage.c                         | 16 +++---
>  fs/mpage.c                                    |  2 +-
>  fs/nfs/file.c                                 | 22 ++++----
>  fs/nfs/fscache.h                              | 14 ++---
>  fs/nilfs2/inode.c                             |  1 -
>  fs/ocfs2/aops.c                               | 10 ++--
>  fs/orangefs/inode.c                           |  6 +--
>  fs/reiserfs/inode.c                           | 20 +++----
>  fs/reiserfs/journal.c                         | 14 ++---
>  fs/ubifs/file.c                               | 18 +++----
>  fs/xfs/xfs_aops.c                             |  2 +-
>  fs/zonefs/super.c                             |  2 +-
>  include/linux/buffer_head.h                   |  4 +-
>  include/linux/fs.h                            |  2 +-
>  include/linux/iomap.h                         |  2 +-
>  include/linux/jbd2.h                          |  2 +-
>  include/linux/page-flags.h                    |  2 +-
>  include/linux/pagemap.h                       |  4 --
>  mm/filemap.c                                  |  6 +--
>  mm/migrate.c                                  |  2 +-
>  mm/vmscan.c                                   |  2 +-
>  51 files changed, 309 insertions(+), 312 deletions(-)
> 

Looks like a mechanical change overall, and I like the conversion to
bool return instead of int. You can add this to the pile:

Reviewed-by: Jeff Layton <jlayton@kernel.org>

[PATCH 0/4] Unify filler_t and aops->read_folio

[Matthew Wilcox (Oracle)]

Make these two function signatures the same.  This should enable more
cleanup in the future, but for now it helps ensure a bit of type safety
and cleans up the filesystems a bit.

Matthew Wilcox (Oracle) (4):
  jffs2: Pass the file pointer to jffs2_do_readpage_unlock()
  nfs: Pass the file pointer to nfs_symlink_filler()
  fs: Change the type of filler_t
  mm/filemap: Hoist filler_t decision to the top of
    do_read_cache_folio()

 fs/gfs2/aops.c          | 29 ++++++++++---------------
 fs/jffs2/file.c         |  9 ++++----
 fs/jffs2/gc.c           |  2 +-
 fs/jffs2/os-linux.h     |  2 +-
 fs/nfs/symlink.c        | 16 +++++++-------
 include/linux/pagemap.h |  6 +++---
 mm/filemap.c            | 47 +++++++++++++++++++----------------------
 7 files changed, 50 insertions(+), 61 deletions(-)

-- 
2.34.1

[PATCH 1/4] jffs2: Pass the file pointer to jffs2_do_readpage_unlock()

[Matthew Wilcox (Oracle)]

In preparation for unifying the read_cache_page() and read_folio()
implementations, make jffs2_do_readpage_unlock() get the inode
from the page instead of passing it in from read_cache_page().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/jffs2/file.c | 4 ++--
 fs/jffs2/gc.c   | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index f8616683fbee..492fb2da0403 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -112,7 +112,7 @@ static int jffs2_do_readpage_nolock (struct inode *inode, struct page *pg)
 
 int jffs2_do_readpage_unlock(void *data, struct page *pg)
 {
-	int ret = jffs2_do_readpage_nolock(data, pg);
+	int ret = jffs2_do_readpage_nolock(pg->mapping->host, pg);
 	unlock_page(pg);
 	return ret;
 }
@@ -124,7 +124,7 @@ static int jffs2_read_folio(struct file *file, struct folio *folio)
 	int ret;
 
 	mutex_lock(&f->sem);
-	ret = jffs2_do_readpage_unlock(folio->mapping->host, &folio->page);
+	ret = jffs2_do_readpage_unlock(file, &folio->page);
 	mutex_unlock(&f->sem);
 	return ret;
 }
diff --git a/fs/jffs2/gc.c b/fs/jffs2/gc.c
index 373b3b7c9f44..a53bac7569b6 100644
--- a/fs/jffs2/gc.c
+++ b/fs/jffs2/gc.c
@@ -1327,7 +1327,7 @@ static int jffs2_garbage_collect_dnode(struct jffs2_sb_info *c, struct jffs2_era
 	 * trying to write out, read_cache_page() will not deadlock. */
 	mutex_unlock(&f->sem);
 	page = read_cache_page(inode->i_mapping, start >> PAGE_SHIFT,
-			       jffs2_do_readpage_unlock, inode);
+			       jffs2_do_readpage_unlock, NULL);
 	if (IS_ERR(page)) {
 		pr_warn("read_cache_page() returned error: %ld\n",
 			PTR_ERR(page));
-- 
2.34.1

[PATCH 2/4] nfs: Pass the file pointer to nfs_symlink_filler()

[Matthew Wilcox (Oracle)]

In preparation for unifying the read_cache_page() and read_folio()
implementations, make nfs_symlink_filler() get the inode
from the page instead of passing it in from read_cache_page().

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/nfs/symlink.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/symlink.c b/fs/nfs/symlink.c
index 25ba299fdac2..8b53538bcc75 100644
--- a/fs/nfs/symlink.c
+++ b/fs/nfs/symlink.c
@@ -28,7 +28,7 @@
 
 static int nfs_symlink_filler(void *data, struct page *page)
 {
-	struct inode *inode = data;
+	struct inode *inode = page->mapping->host;
 	int error;
 
 	error = NFS_PROTO(inode)->readlink(inode, page, 0, PAGE_SIZE);
@@ -67,7 +67,7 @@ static const char *nfs_get_link(struct dentry *dentry,
 		if (err)
 			return err;
 		page = read_cache_page(&inode->i_data, 0, nfs_symlink_filler,
-				inode);
+				NULL);
 		if (IS_ERR(page))
 			return ERR_CAST(page);
 	}
-- 
2.34.1

[PATCH 3/4] fs: Change the type of filler_t

[Matthew Wilcox (Oracle)]

By making filler_t the same as read_folio, we can use the same function
for both in gfs2.  We can push the use of folios down one more level
in jffs2 and nfs.  We also increase type safety for future users of the
various read_cache_page() family of functions by forcing the parameter
to be a pointer to struct file (or NULL).

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Andreas Gruenbacher <agruenba@redhat.com>
---
 fs/gfs2/aops.c          | 29 +++++++++++-----------------
 fs/jffs2/file.c         |  9 ++++-----
 fs/jffs2/gc.c           |  2 +-
 fs/jffs2/os-linux.h     |  2 +-
 fs/nfs/symlink.c        | 14 +++++++-------
 include/linux/pagemap.h |  6 +++---
 mm/filemap.c            | 42 ++++++++++++++++++++---------------------
 7 files changed, 47 insertions(+), 57 deletions(-)

diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index 340bf5d0e835..1016631bcbdc 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -464,21 +464,24 @@ static int stuffed_readpage(struct gfs2_inode *ip, struct page *page)
 	return 0;
 }
 
-
-static int __gfs2_readpage(void *file, struct page *page)
+/**
+ * gfs2_read_folio - read a folio from a file
+ * @file: The file to read
+ * @folio: The folio in the file
+ */
+static int gfs2_read_folio(struct file *file, struct folio *folio)
 {
-	struct folio *folio = page_folio(page);
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
 	int error;
 
 	if (!gfs2_is_jdata(ip) ||
-	    (i_blocksize(inode) == PAGE_SIZE && !page_has_buffers(page))) {
+	    (i_blocksize(inode) == PAGE_SIZE && !folio_buffers(folio))) {
 		error = iomap_read_folio(folio, &gfs2_iomap_ops);
 	} else if (gfs2_is_stuffed(ip)) {
-		error = stuffed_readpage(ip, page);
-		unlock_page(page);
+		error = stuffed_readpage(ip, &folio->page);
+		folio_unlock(folio);
 	} else {
 		error = mpage_read_folio(folio, gfs2_block_map);
 	}
@@ -489,16 +492,6 @@ static int __gfs2_readpage(void *file, struct page *page)
 	return error;
 }
 
-/**
- * gfs2_read_folio - read a folio from a file
- * @file: The file to read
- * @folio: The folio in the file
- */
-static int gfs2_read_folio(struct file *file, struct folio *folio)
-{
-	return __gfs2_readpage(file, &folio->page);
-}
-
 /**
  * gfs2_internal_read - read an internal file
  * @ip: The gfs2 inode
@@ -523,7 +516,7 @@ int gfs2_internal_read(struct gfs2_inode *ip, char *buf, loff_t *pos,
 		amt = size - copied;
 		if (offset + size > PAGE_SIZE)
 			amt = PAGE_SIZE - offset;
-		page = read_cache_page(mapping, index, __gfs2_readpage, NULL);
+		page = read_cache_page(mapping, index, gfs2_read_folio, NULL);
 		if (IS_ERR(page))
 			return PTR_ERR(page);
 		p = kmap_atomic(page);
diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c
index 492fb2da0403..ba86acbe12d3 100644
--- a/fs/jffs2/file.c
+++ b/fs/jffs2/file.c
@@ -110,21 +110,20 @@ static int jffs2_do_readpage_nolock (struct inode *inode, struct page *pg)
 	return ret;
 }
 
-int jffs2_do_readpage_unlock(void *data, struct page *pg)
+int __jffs2_read_folio(struct file *file, struct folio *folio)
 {
-	int ret = jffs2_do_readpage_nolock(pg->mapping->host, pg);
-	unlock_page(pg);
+	int ret = jffs2_do_readpage_nolock(folio->mapping->host, &folio->page);
+	folio_unlock(folio);
 	return ret;
 }
 
-
 static int jffs2_read_folio(struct file *file, struct folio *folio)
 {
 	struct jffs2_inode_info *f = JFFS2_INODE_INFO(folio->mapping->host);
 	int ret;
 
 	mutex_lock(&f->sem);
-	ret = jffs2_do_readpage_unlock(file, &folio->page);
+	ret = __jffs2_read_folio(file, folio);
 	mutex_unlock(&f->sem);
 	return ret;
 }
diff --git a/fs/jffs2/gc.c b/fs/jffs2/gc.c
index a53bac7569b6..5c6602f3c189 100644
--- a/fs/jffs2/gc.c
+++ b/fs/jffs2/gc.c
@@ -1327,7 +1327,7 @@ static int jffs2_garbage_collect_dnode(struct jffs2_sb_info *c, struct jffs2_era
 	 * trying to write out, read_cache_page() will not deadlock. */
 	mutex_unlock(&f->sem);
 	page = read_cache_page(inode->i_mapping, start >> PAGE_SHIFT,
-			       jffs2_do_readpage_unlock, NULL);
+			       __jffs2_read_folio, NULL);
 	if (IS_ERR(page)) {
 		pr_warn("read_cache_page() returned error: %ld\n",
 			PTR_ERR(page));
diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h
index 173eccac691d..921d782583d6 100644
--- a/fs/jffs2/os-linux.h
+++ b/fs/jffs2/os-linux.h
@@ -155,7 +155,7 @@ extern const struct file_operations jffs2_file_operations;
 extern const struct inode_operations jffs2_file_inode_operations;
 extern const struct address_space_operations jffs2_file_address_operations;
 int jffs2_fsync(struct file *, loff_t, loff_t, int);
-int jffs2_do_readpage_unlock(void *data, struct page *pg);
+int __jffs2_read_folio(struct file *file, struct folio *folio);
 
 /* ioctl.c */
 long jffs2_ioctl(struct file *, unsigned int, unsigned long);
diff --git a/fs/nfs/symlink.c b/fs/nfs/symlink.c
index 8b53538bcc75..0e27a2e4e68b 100644
--- a/fs/nfs/symlink.c
+++ b/fs/nfs/symlink.c
@@ -26,21 +26,21 @@
  * and straight-forward than readdir caching.
  */
 
-static int nfs_symlink_filler(void *data, struct page *page)
+static int nfs_symlink_filler(struct file *file, struct folio *folio)
 {
-	struct inode *inode = page->mapping->host;
+	struct inode *inode = folio->mapping->host;
 	int error;
 
-	error = NFS_PROTO(inode)->readlink(inode, page, 0, PAGE_SIZE);
+	error = NFS_PROTO(inode)->readlink(inode, &folio->page, 0, PAGE_SIZE);
 	if (error < 0)
 		goto error;
-	SetPageUptodate(page);
-	unlock_page(page);
+	folio_mark_uptodate(folio);
+	folio_unlock(folio);
 	return 0;
 
 error:
-	SetPageError(page);
-	unlock_page(page);
+	folio_set_error(folio);
+	folio_unlock(folio);
 	return -EIO;
 }
 
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index b70192f56454..831b28dab01a 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -492,7 +492,7 @@ static inline gfp_t readahead_gfp_mask(struct address_space *x)
 	return mapping_gfp_mask(x) | __GFP_NORETRY | __GFP_NOWARN;
 }
 
-typedef int filler_t(void *, struct page *);
+typedef int filler_t(struct file *, struct folio *);
 
 pgoff_t page_cache_next_miss(struct address_space *mapping,
 			     pgoff_t index, unsigned long max_scan);
@@ -747,9 +747,9 @@ static inline struct page *grab_cache_page(struct address_space *mapping,
 }
 
 struct folio *read_cache_folio(struct address_space *, pgoff_t index,
-		filler_t *filler, void *data);
+		filler_t *filler, struct file *file);
 struct page *read_cache_page(struct address_space *, pgoff_t index,
-		filler_t *filler, void *data);
+		filler_t *filler, struct file *file);
 extern struct page * read_cache_page_gfp(struct address_space *mapping,
 				pgoff_t index, gfp_t gfp_mask);
 
diff --git a/mm/filemap.c b/mm/filemap.c
index 6ede3ab76a5a..81a0ed08a82c 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3483,7 +3483,7 @@ EXPORT_SYMBOL(generic_file_mmap);
 EXPORT_SYMBOL(generic_file_readonly_mmap);
 
 static struct folio *do_read_cache_folio(struct address_space *mapping,
-		pgoff_t index, filler_t filler, void *data, gfp_t gfp)
+		pgoff_t index, filler_t filler, struct file *file, gfp_t gfp)
 {
 	struct folio *folio;
 	int err;
@@ -3504,11 +3504,9 @@ static struct folio *do_read_cache_folio(struct address_space *mapping,
 
 filler:
 		if (filler)
-			err = filler(data, &folio->page);
-		else if (mapping->a_ops->read_folio)
-			err = mapping->a_ops->read_folio(data, folio);
+			err = filler(file, folio);
 		else
-			err = mapping->a_ops->readpage(data, &folio->page);
+			err = mapping->a_ops->read_folio(file, folio);
 
 		if (err < 0) {
 			folio_put(folio);
@@ -3559,44 +3557,44 @@ static struct folio *do_read_cache_folio(struct address_space *mapping,
 }
 
 /**
- * read_cache_folio - read into page cache, fill it if needed
- * @mapping:	the page's address_space
- * @index:	the page index
- * @filler:	function to perform the read
- * @data:	first arg to filler(data, page) function, often left as NULL
- *
- * Read into the page cache. If a page already exists, and PageUptodate() is
- * not set, try to fill the page and wait for it to become unlocked.
+ * read_cache_folio - Read into page cache, fill it if needed.
+ * @mapping: The address_space to read from.
+ * @index: The index to read.
+ * @filler: Function to perform the read, or NULL to use aops->read_folio().
+ * @file: Passed to filler function, may be NULL if not required.
  *
- * If the page does not get brought uptodate, return -EIO.
+ * Read one page into the page cache.  If it succeeds, the folio returned
+ * will contain @index, but it may not be the first page of the folio.
  *
- * The function expects mapping->invalidate_lock to be already held.
+ * If the filler function returns an error, it will be returned to the
+ * caller.
  *
- * Return: up to date page on success, ERR_PTR() on failure.
+ * Context: May sleep.  Expects mapping->invalidate_lock to be held.
+ * Return: An uptodate folio on success, ERR_PTR() on failure.
  */
 struct folio *read_cache_folio(struct address_space *mapping, pgoff_t index,
-		filler_t filler, void *data)
+		filler_t filler, struct file *file)
 {
-	return do_read_cache_folio(mapping, index, filler, data,
+	return do_read_cache_folio(mapping, index, filler, file,
 			mapping_gfp_mask(mapping));
 }
 EXPORT_SYMBOL(read_cache_folio);
 
 static struct page *do_read_cache_page(struct address_space *mapping,
-		pgoff_t index, filler_t *filler, void *data, gfp_t gfp)
+		pgoff_t index, filler_t *filler, struct file *file, gfp_t gfp)
 {
 	struct folio *folio;
 
-	folio = do_read_cache_folio(mapping, index, filler, data, gfp);
+	folio = do_read_cache_folio(mapping, index, filler, file, gfp);
 	if (IS_ERR(folio))
 		return &folio->page;
 	return folio_file_page(folio, index);
 }
 
 struct page *read_cache_page(struct address_space *mapping,
-				pgoff_t index, filler_t *filler, void *data)
+			pgoff_t index, filler_t *filler, struct file *file)
 {
-	return do_read_cache_page(mapping, index, filler, data,
+	return do_read_cache_page(mapping, index, filler, file,
 			mapping_gfp_mask(mapping));
 }
 EXPORT_SYMBOL(read_cache_page);
-- 
2.34.1

[PATCH 4/4] mm/filemap: Hoist filler_t decision to the top of do_read_cache_folio()

[Matthew Wilcox (Oracle)]

Now that filler_t and aops->read_folio() have the same type, we can decide
which one to use at the top of the function, and cache ->read_folio in
the filler parameter.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 mm/filemap.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/mm/filemap.c b/mm/filemap.c
index 81a0ed08a82c..9b7fa47feb5e 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -3487,6 +3487,9 @@ static struct folio *do_read_cache_folio(struct address_space *mapping,
 {
 	struct folio *folio;
 	int err;
+
+	if (!filler)
+		filler = mapping->a_ops->read_folio;
 repeat:
 	folio = filemap_get_folio(mapping, index);
 	if (!folio) {
@@ -3503,11 +3506,7 @@ static struct folio *do_read_cache_folio(struct address_space *mapping,
 		}
 
 filler:
-		if (filler)
-			err = filler(file, folio);
-		else
-			err = mapping->a_ops->read_folio(file, folio);
-
+		err = filler(file, folio);
 		if (err < 0) {
 			folio_put(folio);
 			return ERR_PTR(err);
-- 
2.34.1

[PATCH 0/5] Convert aops->freepage to aops->free_folio

[Matthew Wilcox (Oracle)]

Not a lot of filesystems implement ->freepage, thankfully, so this
is a slightly smaller patchset than the last few.

Matthew Wilcox (Oracle) (5):
  fs: Add free_folio address space operation
  orangefs: Convert to free_folio
  nfs: Convert to free_folio
  secretmem: Convert to free_folio
  fs: Remove aops->freepage

 Documentation/filesystems/locking.rst | 10 +++++-----
 Documentation/filesystems/vfs.rst     |  6 +++---
 fs/nfs/dir.c                          |  9 +++++++--
 fs/orangefs/inode.c                   |  6 +++---
 include/linux/fs.h                    |  2 +-
 mm/filemap.c                          | 16 ++++++++--------
 mm/secretmem.c                        |  8 ++++----
 mm/vmscan.c                           |  8 ++++----
 8 files changed, 35 insertions(+), 30 deletions(-)

-- 
2.34.1

[PATCH 1/5] fs: Add free_folio address space operation

[Matthew Wilcox (Oracle)]

Include documentation and convert the callers to use ->free_folio as
well as ->freepage.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 Documentation/filesystems/locking.rst | 10 +++++-----
 Documentation/filesystems/vfs.rst     |  6 +++---
 include/linux/fs.h                    |  1 +
 mm/filemap.c                          |  9 ++++++++-
 mm/vmscan.c                           |  6 +++++-
 5 files changed, 22 insertions(+), 10 deletions(-)

diff --git a/Documentation/filesystems/locking.rst b/Documentation/filesystems/locking.rst
index 2a295bb72dbc..55b029b2d420 100644
--- a/Documentation/filesystems/locking.rst
+++ b/Documentation/filesystems/locking.rst
@@ -250,7 +250,7 @@ prototypes::
 	sector_t (*bmap)(struct address_space *, sector_t);
 	void (*invalidate_folio) (struct folio *, size_t start, size_t len);
 	int (*release_folio)(struct folio *, gfp_t);
-	void (*freepage)(struct page *);
+	void (*free_folio)(struct folio *);
 	int (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 	bool (*isolate_page) (struct page *, isolate_mode_t);
 	int (*migratepage)(struct address_space *, struct page *, struct page *);
@@ -262,10 +262,10 @@ prototypes::
 	int (*swap_deactivate)(struct file *);
 
 locking rules:
-	All except dirty_folio and freepage may block
+	All except dirty_folio and free_folio may block
 
 ======================	======================== =========	===============
-ops			PageLocked(page)	 i_rwsem	invalidate_lock
+ops			folio locked		 i_rwsem	invalidate_lock
 ======================	======================== =========	===============
 writepage:		yes, unlocks (see below)
 read_folio:		yes, unlocks				shared
@@ -277,7 +277,7 @@ write_end:		yes, unlocks		 exclusive
 bmap:
 invalidate_folio:	yes					exclusive
 release_folio:		yes
-freepage:		yes
+free_folio:		yes
 direct_IO:
 isolate_page:		yes
 migratepage:		yes (both)
@@ -377,7 +377,7 @@ buffers from the folio in preparation for freeing it.  It returns false to
 indicate that the buffers are (or may be) freeable.  If ->release_folio is
 NULL, the kernel assumes that the fs has no private interest in the buffers.
 
-->freepage() is called when the kernel is done dropping the page
+->free_folio() is called when the kernel has dropped the folio
 from the page cache.
 
 ->launder_folio() may be called prior to releasing a folio if
diff --git a/Documentation/filesystems/vfs.rst b/Documentation/filesystems/vfs.rst
index 679887b5c8fc..12a011d2cbc6 100644
--- a/Documentation/filesystems/vfs.rst
+++ b/Documentation/filesystems/vfs.rst
@@ -735,7 +735,7 @@ cache in your filesystem.  The following members are defined:
 		sector_t (*bmap)(struct address_space *, sector_t);
 		void (*invalidate_folio) (struct folio *, size_t start, size_t len);
 		bool (*release_folio)(struct folio *, gfp_t);
-		void (*freepage)(struct page *);
+		void (*free_folio)(struct folio *);
 		ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 		/* isolate a page for migration */
 		bool (*isolate_page) (struct page *, isolate_mode_t);
@@ -891,8 +891,8 @@ cache in your filesystem.  The following members are defined:
 	its release_folio will need to ensure this.  Possibly it can
 	clear the uptodate flag if it cannot free private data yet.
 
-``freepage``
-	freepage is called once the page is no longer visible in the
+``free_folio``
+	free_folio is called once the folio is no longer visible in the
 	page cache in order to allow the cleanup of any private data.
 	Since it may be called by the memory reclaimer, it should not
 	assume that the original address_space mapping still exists, and
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 1cee64d9724b..915844e6293e 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -356,6 +356,7 @@ struct address_space_operations {
 	sector_t (*bmap)(struct address_space *, sector_t);
 	void (*invalidate_folio) (struct folio *, size_t offset, size_t len);
 	bool (*release_folio)(struct folio *, gfp_t);
+	void (*free_folio)(struct folio *folio);
 	void (*freepage)(struct page *);
 	ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 	/*
diff --git a/mm/filemap.c b/mm/filemap.c
index d335a154a0d9..adcdef56890f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -226,8 +226,12 @@ void __filemap_remove_folio(struct folio *folio, void *shadow)
 void filemap_free_folio(struct address_space *mapping, struct folio *folio)
 {
 	void (*freepage)(struct page *);
+	void (*free_folio)(struct folio *);
 	int refs = 1;
 
+	free_folio = mapping->a_ops->free_folio;
+	if (free_folio)
+		free_folio(folio);
 	freepage = mapping->a_ops->freepage;
 	if (freepage)
 		freepage(&folio->page);
@@ -807,6 +811,7 @@ void replace_page_cache_page(struct page *old, struct page *new)
 	struct folio *fold = page_folio(old);
 	struct folio *fnew = page_folio(new);
 	struct address_space *mapping = old->mapping;
+	void (*free_folio)(struct folio *) = mapping->a_ops->free_folio;
 	void (*freepage)(struct page *) = mapping->a_ops->freepage;
 	pgoff_t offset = old->index;
 	XA_STATE(xas, &mapping->i_pages, offset);
@@ -835,9 +840,11 @@ void replace_page_cache_page(struct page *old, struct page *new)
 	if (PageSwapBacked(new))
 		__inc_lruvec_page_state(new, NR_SHMEM);
 	xas_unlock_irq(&xas);
+	if (free_folio)
+		free_folio(fold);
 	if (freepage)
 		freepage(old);
-	put_page(old);
+	folio_put(fold);
 }
 EXPORT_SYMBOL_GPL(replace_page_cache_page);
 
diff --git a/mm/vmscan.c b/mm/vmscan.c
index f3f7ce2c4068..d8a031128ad0 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1282,8 +1282,10 @@ static int __remove_mapping(struct address_space *mapping, struct folio *folio,
 		xa_unlock_irq(&mapping->i_pages);
 		put_swap_page(&folio->page, swap);
 	} else {
+		void (*free_folio)(struct folio *);
 		void (*freepage)(struct page *);
 
+		free_folio = mapping->a_ops->free_folio;
 		freepage = mapping->a_ops->freepage;
 		/*
 		 * Remember a shadow entry for reclaimed file cache in
@@ -1310,7 +1312,9 @@ static int __remove_mapping(struct address_space *mapping, struct folio *folio,
 			inode_add_lru(mapping->host);
 		spin_unlock(&mapping->host->i_lock);
 
-		if (freepage != NULL)
+		if (free_folio)
+			free_folio(folio);
+		if (freepage)
 			freepage(&folio->page);
 	}
 
-- 
2.34.1

[PATCH 2/5] orangefs: Convert to free_folio

[Matthew Wilcox (Oracle)]

I suspect this isn't actually needed and that releasepage will have
done the job, but convert it for now and we can delete it later.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/orangefs/inode.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c
index 3509618e7b37..5ce27dde3c79 100644
--- a/fs/orangefs/inode.c
+++ b/fs/orangefs/inode.c
@@ -490,9 +490,9 @@ static bool orangefs_release_folio(struct folio *folio, gfp_t foo)
 	return !folio_test_private(folio);
 }
 
-static void orangefs_freepage(struct page *page)
+static void orangefs_free_folio(struct folio *folio)
 {
-	kfree(detach_page_private(page));
+	kfree(folio_detach_private(folio));
 }
 
 static int orangefs_launder_folio(struct folio *folio)
@@ -637,7 +637,7 @@ static const struct address_space_operations orangefs_address_operations = {
 	.write_end = orangefs_write_end,
 	.invalidate_folio = orangefs_invalidate_folio,
 	.release_folio = orangefs_release_folio,
-	.freepage = orangefs_freepage,
+	.free_folio = orangefs_free_folio,
 	.launder_folio = orangefs_launder_folio,
 	.direct_IO = orangefs_direct_IO,
 };
-- 
2.34.1

[PATCH 3/5] nfs: Convert to free_folio

[Matthew Wilcox (Oracle)]

Add a wrapper that converts back from the folio to the page.  This
entire file needs to be converted to use folios, but that's a
task for a different set of patches.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 fs/nfs/dir.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index c6b263b5faf1..a8ecdd527662 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -55,7 +55,7 @@ static int nfs_closedir(struct inode *, struct file *);
 static int nfs_readdir(struct file *, struct dir_context *);
 static int nfs_fsync_dir(struct file *, loff_t, loff_t, int);
 static loff_t nfs_llseek_dir(struct file *, loff_t, int);
-static void nfs_readdir_clear_array(struct page*);
+static void nfs_readdir_free_folio(struct folio *);
 
 const struct file_operations nfs_dir_operations = {
 	.llseek		= nfs_llseek_dir,
@@ -67,7 +67,7 @@ const struct file_operations nfs_dir_operations = {
 };
 
 const struct address_space_operations nfs_dir_aops = {
-	.freepage = nfs_readdir_clear_array,
+	.free_folio = nfs_readdir_free_folio,
 };
 
 #define NFS_INIT_DTSIZE PAGE_SIZE
@@ -228,6 +228,11 @@ static void nfs_readdir_clear_array(struct page *page)
 	kunmap_atomic(array);
 }
 
+static void nfs_readdir_free_folio(struct folio *folio)
+{
+	nfs_readdir_clear_array(&folio->page);
+}
+
 static void nfs_readdir_page_reinit_array(struct page *page, u64 last_cookie,
 					  u64 change_attr)
 {
-- 
2.34.1

[PATCH 4/5] secretmem: Convert to free_folio

[Matthew Wilcox (Oracle)]

Prepare for any size of folio, even though secretmem only uses order-0
folios for now.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 mm/secretmem.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/mm/secretmem.c b/mm/secretmem.c
index 3b3cf2892b6a..206ed6b40c1d 100644
--- a/mm/secretmem.c
+++ b/mm/secretmem.c
@@ -145,15 +145,15 @@ static int secretmem_migratepage(struct address_space *mapping,
 	return -EBUSY;
 }
 
-static void secretmem_freepage(struct page *page)
+static void secretmem_free_folio(struct folio *folio)
 {
-	set_direct_map_default_noflush(page);
-	clear_highpage(page);
+	set_direct_map_default_noflush(&folio->page);
+	folio_zero_segment(folio, 0, folio_size(folio));
 }
 
 const struct address_space_operations secretmem_aops = {
 	.dirty_folio	= noop_dirty_folio,
-	.freepage	= secretmem_freepage,
+	.free_folio	= secretmem_free_folio,
 	.migratepage	= secretmem_migratepage,
 	.isolate_page	= secretmem_isolate_page,
 };
-- 
2.34.1

[PATCH 5/5] fs: Remove aops->freepage

[Matthew Wilcox (Oracle)]

All implementations now use free_folio so we can delete the callers
and the method.

Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
---
 include/linux/fs.h | 1 -
 mm/filemap.c       | 7 -------
 mm/vmscan.c        | 4 ----
 3 files changed, 12 deletions(-)

diff --git a/include/linux/fs.h b/include/linux/fs.h
index 915844e6293e..6f305f1097a5 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -357,7 +357,6 @@ struct address_space_operations {
 	void (*invalidate_folio) (struct folio *, size_t offset, size_t len);
 	bool (*release_folio)(struct folio *, gfp_t);
 	void (*free_folio)(struct folio *folio);
-	void (*freepage)(struct page *);
 	ssize_t (*direct_IO)(struct kiocb *, struct iov_iter *iter);
 	/*
 	 * migrate the contents of a page to the specified target. If
diff --git a/mm/filemap.c b/mm/filemap.c
index adcdef56890f..fa0ca674450f 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -225,16 +225,12 @@ void __filemap_remove_folio(struct folio *folio, void *shadow)
 
 void filemap_free_folio(struct address_space *mapping, struct folio *folio)
 {
-	void (*freepage)(struct page *);
 	void (*free_folio)(struct folio *);
 	int refs = 1;
 
 	free_folio = mapping->a_ops->free_folio;
 	if (free_folio)
 		free_folio(folio);
-	freepage = mapping->a_ops->freepage;
-	if (freepage)
-		freepage(&folio->page);
 
 	if (folio_test_large(folio) && !folio_test_hugetlb(folio))
 		refs = folio_nr_pages(folio);
@@ -812,7 +808,6 @@ void replace_page_cache_page(struct page *old, struct page *new)
 	struct folio *fnew = page_folio(new);
 	struct address_space *mapping = old->mapping;
 	void (*free_folio)(struct folio *) = mapping->a_ops->free_folio;
-	void (*freepage)(struct page *) = mapping->a_ops->freepage;
 	pgoff_t offset = old->index;
 	XA_STATE(xas, &mapping->i_pages, offset);
 
@@ -842,8 +837,6 @@ void replace_page_cache_page(struct page *old, struct page *new)
 	xas_unlock_irq(&xas);
 	if (free_folio)
 		free_folio(fold);
-	if (freepage)
-		freepage(old);
 	folio_put(fold);
 }
 EXPORT_SYMBOL_GPL(replace_page_cache_page);
diff --git a/mm/vmscan.c b/mm/vmscan.c
index d8a031128ad0..edc89f26b738 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1283,10 +1283,8 @@ static int __remove_mapping(struct address_space *mapping, struct folio *folio,
 		put_swap_page(&folio->page, swap);
 	} else {
 		void (*free_folio)(struct folio *);
-		void (*freepage)(struct page *);
 
 		free_folio = mapping->a_ops->free_folio;
-		freepage = mapping->a_ops->freepage;
 		/*
 		 * Remember a shadow entry for reclaimed file cache in
 		 * order to detect refaults, thus thrashing, later on.
@@ -1314,8 +1312,6 @@ static int __remove_mapping(struct address_space *mapping, struct folio *folio,
 
 		if (free_folio)
 			free_folio(folio);
-		if (freepage)
-			freepage(&folio->page);
 	}
 
 	return 1;
-- 
2.34.1