Merge patch series "convert the majority of file systems to mmap_prepare"

Lorenzo Stoakes <[email protected]> says:

In commit c84bf6d ("mm: introduce new .mmap_prepare() file
callback"), a new hook for mmap was introduced - f_op->mmap_prepare().

This is preferred to the existing f_op->mmap() hook as it does require a
VMA to be established yet, thus allowing the mmap logic to invoke this hook
far, far earlier, prior to inserting a VMA into the virtual address space,
or performing any other heavy handed operations.

This allows for much simpler unwinding on error, and for there to be a
single attempt at merging a VMA rather than having to possibly reattempt a
merge based on potentially altered VMA state.

Far more importantly, it prevents inappropriate manipulation of
incompletely initialised VMA state, which is something that has been the
cause of bugs and complexity in the past.

The intent is to gradually deprecate f_op->mmap, and in that vein this
series coverts the majority of file systems to using f_op->mmap_prepare.

Prerequisite steps are taken - firstly ensuring all checks for mmap
capabilities use the file_has_valid_mmap_hooks() helper rather than
directly checking for f_op->mmap (which is now not a valid check) and
secondly updating daxdev_mapping_supported() to not require a VMA parameter
to allow ext4 and xfs to be converted.

Commit bb666b7 ("mm: add mmap_prepare() compatibility layer for nested
file systems") handles the nasty edge-case of nested file systems like
overlayfs, which introduces a compatibility shim to allow
f_op->mmap_prepare() to be invoked from an f_op->mmap() callback.

This allows for nested filesystems to continue to function correctly with
all file systems regardless of which callback is used. Once we finally
convert all file systems, this shim can be removed.

As a result, ecryptfs, fuse, and overlayfs remain unaltered so they can
nest all other file systems.

We additionally do not update resctl - as this requires an update to
remap_pfn_range() (or an alternative to it) which we defer to a later
series, equally we do not update cramfs which needs a mixed mapping
insertion with the same issue, nor do we update procfs, hugetlbfs, syfs or
kernfs all of which require VMAs for internal state and hooks. We shall
return to all of these later.

* patches from https://lore.kernel.org/[email protected]:
  fs: replace mmap hook with .mmap_prepare for simple mappings
  fs: convert most other generic_file_*mmap() users to .mmap_prepare()
  fs: convert simple use of generic_file_*_mmap() to .mmap_prepare()
  mm/filemap: introduce generic_file_*_mmap_prepare() helpers
  fs/xfs: transition from deprecated .mmap hook to .mmap_prepare
  fs/ext4: transition from deprecated .mmap hook to .mmap_prepare
  fs/dax: make it possible to check dev dax support without a VMA
  fs: consistently use file_has_valid_mmap_hooks() helper
  mm/nommu: use file_has_valid_mmap_hooks() helper
  mm: rename call_mmap/mmap_prepare to vfs_mmap/mmap_prepare

Link: https://lore.kernel.org/[email protected]
Signed-off-by: Christian Brauner <[email protected]>

Author: brauner

block/fops.c

@@ -911,14 +911,14 @@ static long blkdev_fallocate(struct file *file, int mode, loff_t start,
 	return error;
 }
 
-static int blkdev_mmap(struct file *file, struct vm_area_struct *vma)
+static int blkdev_mmap_prepare(struct vm_area_desc *desc)
 {
-	struct inode *bd_inode = bdev_file_inode(file);
+	struct file *file = desc->file;
 
-	if (bdev_read_only(I_BDEV(bd_inode)))
-		return generic_file_readonly_mmap(file, vma);
+	if (bdev_read_only(I_BDEV(bdev_file_inode(file))))
+		return generic_file_readonly_mmap_prepare(desc);
 
-	return generic_file_mmap(file, vma);
+	return generic_file_mmap_prepare(desc);
 }
 
 const struct file_operations def_blk_fops = {
@@ -928,7 +928,7 @@ const struct file_operations def_blk_fops = {
 	.read_iter	= blkdev_read_iter,
 	.write_iter	= blkdev_write_iter,
 	.iopoll		= iocb_bio_iopoll,
-	.mmap		= blkdev_mmap,
+	.mmap_prepare	= blkdev_mmap_prepare,
 	.fsync		= blkdev_fsync,
 	.unlocked_ioctl	= blkdev_ioctl,
 #ifdef CONFIG_COMPAT

drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c

@@ -105,7 +105,7 @@ static int i915_gem_dmabuf_mmap(struct dma_buf *dma_buf, struct vm_area_struct *
 	if (!obj->base.filp)
 		return -ENODEV;
 
-	ret = call_mmap(obj->base.filp, vma);
+	ret = vfs_mmap(obj->base.filp, vma);
 	if (ret)
 		return ret;
 

fs/9p/vfs_file.c

@@ -454,22 +454,23 @@ int v9fs_file_fsync_dotl(struct file *filp, loff_t start, loff_t end,
 }
 
 static int
-v9fs_file_mmap(struct file *filp, struct vm_area_struct *vma)
+v9fs_file_mmap_prepare(struct vm_area_desc *desc)
 {
 	int retval;
+	struct file *filp = desc->file;
 	struct inode *inode = file_inode(filp);
 	struct v9fs_session_info *v9ses = v9fs_inode2v9ses(inode);
 
 	p9_debug(P9_DEBUG_MMAP, "filp :%p\n", filp);
 
 	if (!(v9ses->cache & CACHE_WRITEBACK)) {
 		p9_debug(P9_DEBUG_CACHE, "(read-only mmap mode)");
-		return generic_file_readonly_mmap(filp, vma);
+		return generic_file_readonly_mmap_prepare(desc);
 	}
 
-	retval = generic_file_mmap(filp, vma);
+	retval = generic_file_mmap_prepare(desc);
 	if (!retval)
-		vma->vm_ops = &v9fs_mmap_file_vm_ops;
+		desc->vm_ops = &v9fs_mmap_file_vm_ops;
 
 	return retval;
 }
@@ -516,7 +517,7 @@ const struct file_operations v9fs_file_operations = {
 	.open = v9fs_file_open,
 	.release = v9fs_dir_release,
 	.lock = v9fs_file_lock,
-	.mmap = generic_file_readonly_mmap,
+	.mmap_prepare = generic_file_readonly_mmap_prepare,
 	.splice_read = v9fs_file_splice_read,
 	.splice_write = iter_file_splice_write,
 	.fsync = v9fs_file_fsync,
@@ -531,7 +532,7 @@ const struct file_operations v9fs_file_operations_dotl = {
 	.release = v9fs_dir_release,
 	.lock = v9fs_file_lock_dotl,
 	.flock = v9fs_file_flock_dotl,
-	.mmap = v9fs_file_mmap,
+	.mmap_prepare = v9fs_file_mmap_prepare,
 	.splice_read = v9fs_file_splice_read,
 	.splice_write = iter_file_splice_write,
 	.fsync = v9fs_file_fsync_dotl,

fs/adfs/file.c

@@ -25,7 +25,7 @@
 const struct file_operations adfs_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read_iter	= generic_file_read_iter,
-	.mmap		= generic_file_mmap,
+	.mmap_prepare	= generic_file_mmap_prepare,
 	.fsync		= generic_file_fsync,
 	.write_iter	= generic_file_write_iter,
 	.splice_read	= filemap_splice_read,

fs/affs/file.c

@@ -999,7 +999,7 @@ const struct file_operations affs_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read_iter	= generic_file_read_iter,
 	.write_iter	= generic_file_write_iter,
-	.mmap		= generic_file_mmap,
+	.mmap_prepare	= generic_file_mmap_prepare,
 	.open		= affs_file_open,
 	.release	= affs_file_release,
 	.fsync		= affs_file_fsync,

fs/afs/file.c

@@ -19,7 +19,7 @@
 #include <trace/events/netfs.h>
 #include "internal.h"
 
-static int afs_file_mmap(struct file *file, struct vm_area_struct *vma);
+static int afs_file_mmap_prepare(struct vm_area_desc *desc);
 
 static ssize_t afs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter);
 static ssize_t afs_file_splice_read(struct file *in, loff_t *ppos,
@@ -35,7 +35,7 @@ const struct file_operations afs_file_operations = {
 	.llseek		= generic_file_llseek,
 	.read_iter	= afs_file_read_iter,
 	.write_iter	= netfs_file_write_iter,
-	.mmap		= afs_file_mmap,
+	.mmap_prepare	= afs_file_mmap_prepare,
 	.splice_read	= afs_file_splice_read,
 	.splice_write	= iter_file_splice_write,
 	.fsync		= afs_fsync,
@@ -492,16 +492,16 @@ static void afs_drop_open_mmap(struct afs_vnode *vnode)
 /*
  * Handle setting up a memory mapping on an AFS file.
  */
-static int afs_file_mmap(struct file *file, struct vm_area_struct *vma)
+static int afs_file_mmap_prepare(struct vm_area_desc *desc)
 {
-	struct afs_vnode *vnode = AFS_FS_I(file_inode(file));
+	struct afs_vnode *vnode = AFS_FS_I(file_inode(desc->file));
 	int ret;
 
 	afs_add_open_mmap(vnode);
 
-	ret = generic_file_mmap(file, vma);
+	ret = generic_file_mmap_prepare(desc);
 	if (ret == 0)
-		vma->vm_ops = &afs_vm_ops;
+		desc->vm_ops = &afs_vm_ops;
 	else
 		afs_drop_open_mmap(vnode);
 	return ret;

fs/aio.c

@@ -392,15 +392,15 @@ static const struct vm_operations_struct aio_ring_vm_ops = {
 #endif
 };
 
-static int aio_ring_mmap(struct file *file, struct vm_area_struct *vma)
+static int aio_ring_mmap_prepare(struct vm_area_desc *desc)
 {
-	vm_flags_set(vma, VM_DONTEXPAND);
-	vma->vm_ops = &aio_ring_vm_ops;
+	desc->vm_flags |= VM_DONTEXPAND;
+	desc->vm_ops = &aio_ring_vm_ops;
 	return 0;
 }
 
 static const struct file_operations aio_ring_fops = {
-	.mmap = aio_ring_mmap,
+	.mmap_prepare = aio_ring_mmap_prepare,
 };
 
 #if IS_ENABLED(CONFIG_MIGRATION)

fs/backing-file.c

@@ -333,13 +333,13 @@ int backing_file_mmap(struct file *file, struct vm_area_struct *vma,
 	if (WARN_ON_ONCE(!(file->f_mode & FMODE_BACKING)))
 		return -EIO;
 
-	if (!file->f_op->mmap)
+	if (!can_mmap_file(file))
 		return -ENODEV;
 
 	vma_set_file(vma, file);
 
 	old_cred = override_creds(ctx->cred);
-	ret = call_mmap(vma->vm_file, vma);
+	ret = vfs_mmap(vma->vm_file, vma);
 	revert_creds(old_cred);
 
 	if (ctx->accessed)

fs/bcachefs/fs.c

@@ -1553,11 +1553,11 @@ static const struct vm_operations_struct bch_vm_ops = {
 	.page_mkwrite   = bch2_page_mkwrite,
 };
 
-static int bch2_mmap(struct file *file, struct vm_area_struct *vma)
+static int bch2_mmap_prepare(struct vm_area_desc *desc)
 {
-	file_accessed(file);
+	file_accessed(desc->file);
 
-	vma->vm_ops = &bch_vm_ops;
+	desc->vm_ops = &bch_vm_ops;
 	return 0;
 }
 
@@ -1740,7 +1740,7 @@ static const struct file_operations bch_file_operations = {
 	.llseek		= bch2_llseek,
 	.read_iter	= bch2_read_iter,
 	.write_iter	= bch2_write_iter,
-	.mmap		= bch2_mmap,
+	.mmap_prepare	= bch2_mmap_prepare,
 	.get_unmapped_area = thp_get_unmapped_area,
 	.fsync		= bch2_fsync,
 	.splice_read	= filemap_splice_read,

fs/bfs/file.c

@@ -27,7 +27,7 @@ const struct file_operations bfs_file_operations = {
 	.llseek 	= generic_file_llseek,
 	.read_iter	= generic_file_read_iter,
 	.write_iter	= generic_file_write_iter,
-	.mmap		= generic_file_mmap,
+	.mmap_prepare	= generic_file_mmap_prepare,
 	.splice_read	= filemap_splice_read,
 };