fs: introduce f_op->mmap_capabilities for nommu mmap support

Since "BDI: Provide backing device capability information [try #3]" the
backing_dev_info structure also provides flags for the kind of mmap
operation available in a nommu environment, which is entirely unrelated
to it's original purpose.

Introduce a new nommu-only file operation to provide this information to
the nommu mmap code instead.  Splitting this from the backing_dev_info
structure allows to remove lots of backing_dev_info instance that aren't
otherwise needed, and entirely gets rid of the concept of providing a
backing_dev_info for a character device.  It also removes the need for
the mtd_inodefs filesystem.

Signed-off-by: Christoph Hellwig <[email protected]>
Reviewed-by: Tejun Heo <[email protected]>
Acked-by: Brian Norris <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>

Documentation/nommu-mmap.txt

@@ -43,12 +43,12 @@ and it's also much more restricted in the latter case:
            even if this was created by another process.
 
          - If possible, the file mapping will be directly on the backing device
-           if the backing device has the BDI_CAP_MAP_DIRECT capability and
+           if the backing device has the NOMMU_MAP_DIRECT capability and
            appropriate mapping protection capabilities. Ramfs, romfs, cramfs
            and mtd might all permit this.
 
 	 - If the backing device device can't or won't permit direct sharing,
-           but does have the BDI_CAP_MAP_COPY capability, then a copy of the
+           but does have the NOMMU_MAP_COPY capability, then a copy of the
            appropriate bit of the file will be read into a contiguous bit of
            memory and any extraneous space beyond the EOF will be cleared
 
@@ -220,7 +220,7 @@ directly (can't be copied).
 
 The file->f_op->mmap() operation will be called to actually inaugurate the
 mapping. It can be rejected at that point. Returning the ENOSYS error will
-cause the mapping to be copied instead if BDI_CAP_MAP_COPY is specified.
+cause the mapping to be copied instead if NOMMU_MAP_COPY is specified.
 
 The vm_ops->close() routine will be invoked when the last mapping on a chardev
 is removed. An existing mapping will be shared, partially or not, if possible
@@ -232,7 +232,7 @@ want to handle it, despite the fact it's got an operation. For instance, it
 might try directing the call to a secondary driver which turns out not to
 implement it. Such is the case for the framebuffer driver which attempts to
 direct the call to the device-specific driver. Under such circumstances, the
-mapping request will be rejected if BDI_CAP_MAP_COPY is not specified, and a
+mapping request will be rejected if NOMMU_MAP_COPY is not specified, and a
 copy mapped otherwise.
 
 IMPORTANT NOTE:

block/blk-core.c

@@ -607,7 +607,7 @@ struct request_queue *blk_alloc_queue_node(gfp_t gfp_mask, int node_id)
 	q->backing_dev_info.ra_pages =
 			(VM_MAX_READAHEAD * 1024) / PAGE_CACHE_SIZE;
 	q->backing_dev_info.state = 0;
-	q->backing_dev_info.capabilities = BDI_CAP_MAP_COPY;
+	q->backing_dev_info.capabilities = 0;
 	q->backing_dev_info.name = "block";
 	q->node = node_id;
 

drivers/char/mem.c

@@ -287,13 +287,24 @@ static unsigned long get_unmapped_area_mem(struct file *file,
 	return pgoff << PAGE_SHIFT;
 }
 
+/* permit direct mmap, for read, write or exec */
+static unsigned memory_mmap_capabilities(struct file *file)
+{
+	return NOMMU_MAP_DIRECT |
+		NOMMU_MAP_READ | NOMMU_MAP_WRITE | NOMMU_MAP_EXEC;
+}
+
+static unsigned zero_mmap_capabilities(struct file *file)
+{
+	return NOMMU_MAP_COPY;
+}
+
 /* can't do an in-place private mapping if there's no MMU */
 static inline int private_mapping_ok(struct vm_area_struct *vma)
 {
 	return vma->vm_flags & VM_MAYSHARE;
 }
 #else
-#define get_unmapped_area_mem	NULL
 
 static inline int private_mapping_ok(struct vm_area_struct *vma)
 {
@@ -721,7 +732,10 @@ static const struct file_operations mem_fops = {
 	.write		= write_mem,
 	.mmap		= mmap_mem,
 	.open		= open_mem,
+#ifndef CONFIG_MMU
 	.get_unmapped_area = get_unmapped_area_mem,
+	.mmap_capabilities = memory_mmap_capabilities,
+#endif
 };
 
 #ifdef CONFIG_DEVKMEM
@@ -731,7 +745,10 @@ static const struct file_operations kmem_fops = {
 	.write		= write_kmem,
 	.mmap		= mmap_kmem,
 	.open		= open_kmem,
+#ifndef CONFIG_MMU
 	.get_unmapped_area = get_unmapped_area_mem,
+	.mmap_capabilities = memory_mmap_capabilities,
+#endif
 };
 #endif
 
@@ -760,16 +777,9 @@ static const struct file_operations zero_fops = {
 	.read_iter	= read_iter_zero,
 	.aio_write	= aio_write_zero,
 	.mmap		= mmap_zero,
-};
-
-/*
- * capabilities for /dev/zero
- * - permits private mappings, "copies" are taken of the source of zeros
- * - no writeback happens
- */
-static struct backing_dev_info zero_bdi = {
-	.name		= "char/mem",
-	.capabilities	= BDI_CAP_MAP_COPY | BDI_CAP_NO_ACCT_AND_WRITEBACK,
+#ifndef CONFIG_MMU
+	.mmap_capabilities = zero_mmap_capabilities,
+#endif
 };
 
 static const struct file_operations full_fops = {
@@ -783,22 +793,22 @@ static const struct memdev {
 	const char *name;
 	umode_t mode;
 	const struct file_operations *fops;
-	struct backing_dev_info *dev_info;
+	fmode_t fmode;
 } devlist[] = {
-	 [1] = { "mem", 0, &mem_fops, &directly_mappable_cdev_bdi },
+	 [1] = { "mem", 0, &mem_fops, FMODE_UNSIGNED_OFFSET },
 #ifdef CONFIG_DEVKMEM
-	 [2] = { "kmem", 0, &kmem_fops, &directly_mappable_cdev_bdi },
+	 [2] = { "kmem", 0, &kmem_fops, FMODE_UNSIGNED_OFFSET },
 #endif
-	 [3] = { "null", 0666, &null_fops, NULL },
+	 [3] = { "null", 0666, &null_fops, 0 },
 #ifdef CONFIG_DEVPORT
-	 [4] = { "port", 0, &port_fops, NULL },
+	 [4] = { "port", 0, &port_fops, 0 },
 #endif
-	 [5] = { "zero", 0666, &zero_fops, &zero_bdi },
-	 [7] = { "full", 0666, &full_fops, NULL },
-	 [8] = { "random", 0666, &random_fops, NULL },
-	 [9] = { "urandom", 0666, &urandom_fops, NULL },
+	 [5] = { "zero", 0666, &zero_fops, 0 },
+	 [7] = { "full", 0666, &full_fops, 0 },
+	 [8] = { "random", 0666, &random_fops, 0 },
+	 [9] = { "urandom", 0666, &urandom_fops, 0 },
 #ifdef CONFIG_PRINTK
-	[11] = { "kmsg", 0644, &kmsg_fops, NULL },
+	[11] = { "kmsg", 0644, &kmsg_fops, 0 },
 #endif
 };
 
@@ -816,12 +826,7 @@ static int memory_open(struct inode *inode, struct file *filp)
 		return -ENXIO;
 
 	filp->f_op = dev->fops;
-	if (dev->dev_info)
-		filp->f_mapping->backing_dev_info = dev->dev_info;
-
-	/* Is /dev/mem or /dev/kmem ? */
-	if (dev->dev_info == &directly_mappable_cdev_bdi)
-		filp->f_mode |= FMODE_UNSIGNED_OFFSET;
+	filp->f_mode |= dev->fmode;
 
 	if (dev->fops->open)
 		return dev->fops->open(inode, filp);
@@ -846,11 +851,6 @@ static struct class *mem_class;
 static int __init chr_dev_init(void)
 {
 	int minor;
-	int err;
-
-	err = bdi_init(&zero_bdi);
-	if (err)
-		return err;
 
 	if (register_chrdev(MEM_MAJOR, "mem", &memory_fops))
 		printk("unable to get major %d for memory devs\n", MEM_MAJOR);

drivers/mtd/mtdchar.c

@@ -49,7 +49,6 @@ static DEFINE_MUTEX(mtd_mutex);
  */
 struct mtd_file_info {
 	struct mtd_info *mtd;
-	struct inode *ino;
 	enum mtd_file_modes mode;
 };
 
@@ -59,29 +58,20 @@ static loff_t mtdchar_lseek(struct file *file, loff_t offset, int orig)
 	return fixed_size_llseek(file, offset, orig, mfi->mtd->size);
 }
 
-static int count;
-static struct vfsmount *mnt;
-static struct file_system_type mtd_inodefs_type;
-
 static int mtdchar_open(struct inode *inode, struct file *file)
 {
 	int minor = iminor(inode);
 	int devnum = minor >> 1;
 	int ret = 0;
 	struct mtd_info *mtd;
 	struct mtd_file_info *mfi;
-	struct inode *mtd_ino;
 
 	pr_debug("MTD_open\n");
 
 	/* You can't open the RO devices RW */
 	if ((file->f_mode & FMODE_WRITE) && (minor & 1))
 		return -EACCES;
 
-	ret = simple_pin_fs(&mtd_inodefs_type, &mnt, &count);
-	if (ret)
-		return ret;
-
 	mutex_lock(&mtd_mutex);
 	mtd = get_mtd_device(NULL, devnum);
 
@@ -95,43 +85,26 @@ static int mtdchar_open(struct inode *inode, struct file *file)
 		goto out1;
 	}
 
-	mtd_ino = iget_locked(mnt->mnt_sb, devnum);
-	if (!mtd_ino) {
-		ret = -ENOMEM;
-		goto out1;
-	}
-	if (mtd_ino->i_state & I_NEW) {
-		mtd_ino->i_private = mtd;
-		mtd_ino->i_mode = S_IFCHR;
-		mtd_ino->i_data.backing_dev_info = mtd->backing_dev_info;
-		unlock_new_inode(mtd_ino);
-	}
-	file->f_mapping = mtd_ino->i_mapping;
-
 	/* You can't open it RW if it's not a writeable device */
 	if ((file->f_mode & FMODE_WRITE) && !(mtd->flags & MTD_WRITEABLE)) {
 		ret = -EACCES;
-		goto out2;
+		goto out1;
 	}
 
 	mfi = kzalloc(sizeof(*mfi), GFP_KERNEL);
 	if (!mfi) {
 		ret = -ENOMEM;
-		goto out2;
+		goto out1;
 	}
-	mfi->ino = mtd_ino;
 	mfi->mtd = mtd;
 	file->private_data = mfi;
 	mutex_unlock(&mtd_mutex);
 	return 0;
 
-out2:
-	iput(mtd_ino);
 out1:
 	put_mtd_device(mtd);
 out:
 	mutex_unlock(&mtd_mutex);
-	simple_release_fs(&mnt, &count);
 	return ret;
 } /* mtdchar_open */
 
@@ -148,12 +121,9 @@ static int mtdchar_close(struct inode *inode, struct file *file)
 	if ((file->f_mode & FMODE_WRITE))
 		mtd_sync(mtd);
 
-	iput(mfi->ino);
-
 	put_mtd_device(mtd);
 	file->private_data = NULL;
 	kfree(mfi);
-	simple_release_fs(&mnt, &count);
 
 	return 0;
 } /* mtdchar_close */
@@ -1117,6 +1087,13 @@ static unsigned long mtdchar_get_unmapped_area(struct file *file,
 	ret = mtd_get_unmapped_area(mtd, len, offset, flags);
 	return ret == -EOPNOTSUPP ? -ENODEV : ret;
 }
+
+static unsigned mtdchar_mmap_capabilities(struct file *file)
+{
+	struct mtd_file_info *mfi = file->private_data;
+
+	return mtd_mmap_capabilities(mfi->mtd);
+}
 #endif
 
 /*
@@ -1160,27 +1137,10 @@ static const struct file_operations mtd_fops = {
 	.mmap		= mtdchar_mmap,
 #ifndef CONFIG_MMU
 	.get_unmapped_area = mtdchar_get_unmapped_area,
+	.mmap_capabilities = mtdchar_mmap_capabilities,
 #endif
 };
 
-static const struct super_operations mtd_ops = {
-	.drop_inode = generic_delete_inode,
-	.statfs = simple_statfs,
-};
-
-static struct dentry *mtd_inodefs_mount(struct file_system_type *fs_type,
-				int flags, const char *dev_name, void *data)
-{
-	return mount_pseudo(fs_type, "mtd_inode:", &mtd_ops, NULL, MTD_INODE_FS_MAGIC);
-}
-
-static struct file_system_type mtd_inodefs_type = {
-       .name = "mtd_inodefs",
-       .mount = mtd_inodefs_mount,
-       .kill_sb = kill_anon_super,
-};
-MODULE_ALIAS_FS("mtd_inodefs");
-
 int __init init_mtdchar(void)
 {
 	int ret;
@@ -1193,23 +1153,11 @@ int __init init_mtdchar(void)
 		return ret;
 	}
 
-	ret = register_filesystem(&mtd_inodefs_type);
-	if (ret) {
-		pr_err("Can't register mtd_inodefs filesystem, error %d\n",
-		       ret);
-		goto err_unregister_chdev;
-	}
-
-	return ret;
-
-err_unregister_chdev:
-	__unregister_chrdev(MTD_CHAR_MAJOR, 0, 1 << MINORBITS, "mtd");
 	return ret;
 }
 
 void __exit cleanup_mtdchar(void)
 {
-	unregister_filesystem(&mtd_inodefs_type);
 	__unregister_chrdev(MTD_CHAR_MAJOR, 0, 1 << MINORBITS, "mtd");
 }
 

drivers/mtd/mtdconcat.c

@@ -732,8 +732,6 @@ struct mtd_info *mtd_concat_create(struct mtd_info *subdev[],	/* subdevices to c
 
 	concat->mtd.ecc_stats.badblocks = subdev[0]->ecc_stats.badblocks;
 
-	concat->mtd.backing_dev_info = subdev[0]->backing_dev_info;
-
 	concat->subdev[0] = subdev[0];
 
 	for (i = 1; i < num_devs; i++) {
@@ -761,14 +759,6 @@ struct mtd_info *mtd_concat_create(struct mtd_info *subdev[],	/* subdevices to c
 				    subdev[i]->flags & MTD_WRITEABLE;
 		}
 
-		/* only permit direct mapping if the BDIs are all the same
-		 * - copy-mapping is still permitted
-		 */
-		if (concat->mtd.backing_dev_info !=
-		    subdev[i]->backing_dev_info)
-			concat->mtd.backing_dev_info =
-				&default_backing_dev_info;
-
 		concat->mtd.size += subdev[i]->size;
 		concat->mtd.ecc_stats.badblocks +=
 			subdev[i]->ecc_stats.badblocks;