kernel_optimize_test/fs/jffs2/super.c
David Howells 454e2398be [PATCH] VFS: Permit filesystem to override root dentry on mount
Extend the get_sb() filesystem operation to take an extra argument that
permits the VFS to pass in the target vfsmount that defines the mountpoint.

The filesystem is then required to manually set the superblock and root dentry
pointers.  For most filesystems, this should be done with simple_set_mnt()
which will set the superblock pointer and then set the root dentry to the
superblock's s_root (as per the old default behaviour).

The get_sb() op now returns an integer as there's now no need to return the
superblock pointer.

This patch permits a superblock to be implicitly shared amongst several mount
points, such as can be done with NFS to avoid potential inode aliasing.  In
such a case, simple_set_mnt() would not be called, and instead the mnt_root
and mnt_sb would be set directly.

The patch also makes the following changes:

 (*) the get_sb_*() convenience functions in the core kernel now take a vfsmount
     pointer argument and return an integer, so most filesystems have to change
     very little.

 (*) If one of the convenience function is not used, then get_sb() should
     normally call simple_set_mnt() to instantiate the vfsmount. This will
     always return 0, and so can be tail-called from get_sb().

 (*) generic_shutdown_super() now calls shrink_dcache_sb() to clean up the
     dcache upon superblock destruction rather than shrink_dcache_anon().

     This is required because the superblock may now have multiple trees that
     aren't actually bound to s_root, but that still need to be cleaned up. The
     currently called functions assume that the whole tree is rooted at s_root,
     and that anonymous dentries are not the roots of trees which results in
     dentries being left unculled.

     However, with the way NFS superblock sharing are currently set to be
     implemented, these assumptions are violated: the root of the filesystem is
     simply a dummy dentry and inode (the real inode for '/' may well be
     inaccessible), and all the vfsmounts are rooted on anonymous[*] dentries
     with child trees.

     [*] Anonymous until discovered from another tree.

 (*) The documentation has been adjusted, including the additional bit of
     changing ext2_* into foo_* in the documentation.

[akpm@osdl.org: convert ipath_fs, do other stuff]
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Al Viro <viro@zeniv.linux.org.uk>
Cc: Nathan Scott <nathans@sgi.com>
Cc: Roland Dreier <rolandd@cisco.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-23 07:42:45 -07:00

403 lines
10 KiB
C

/*
* JFFS2 -- Journalling Flash File System, Version 2.
*
* Copyright (C) 2001-2003 Red Hat, Inc.
*
* Created by David Woodhouse <dwmw2@infradead.org>
*
* For licensing information, see the file 'LICENCE' in this directory.
*
* $Id: super.c,v 1.110 2005/11/07 11:14:42 gleixner Exp $
*
*/
#include <linux/config.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <linux/init.h>
#include <linux/list.h>
#include <linux/fs.h>
#include <linux/mount.h>
#include <linux/jffs2.h>
#include <linux/pagemap.h>
#include <linux/mtd/mtd.h>
#include <linux/ctype.h>
#include <linux/namei.h>
#include "compr.h"
#include "nodelist.h"
static void jffs2_put_super(struct super_block *);
static kmem_cache_t *jffs2_inode_cachep;
static struct inode *jffs2_alloc_inode(struct super_block *sb)
{
struct jffs2_inode_info *ei;
ei = (struct jffs2_inode_info *)kmem_cache_alloc(jffs2_inode_cachep, SLAB_KERNEL);
if (!ei)
return NULL;
return &ei->vfs_inode;
}
static void jffs2_destroy_inode(struct inode *inode)
{
kmem_cache_free(jffs2_inode_cachep, JFFS2_INODE_INFO(inode));
}
static void jffs2_i_init_once(void * foo, kmem_cache_t * cachep, unsigned long flags)
{
struct jffs2_inode_info *ei = (struct jffs2_inode_info *) foo;
if ((flags & (SLAB_CTOR_VERIFY|SLAB_CTOR_CONSTRUCTOR)) ==
SLAB_CTOR_CONSTRUCTOR) {
init_MUTEX(&ei->sem);
inode_init_once(&ei->vfs_inode);
}
}
static int jffs2_sync_fs(struct super_block *sb, int wait)
{
struct jffs2_sb_info *c = JFFS2_SB_INFO(sb);
down(&c->alloc_sem);
jffs2_flush_wbuf_pad(c);
up(&c->alloc_sem);
return 0;
}
static struct super_operations jffs2_super_operations =
{
.alloc_inode = jffs2_alloc_inode,
.destroy_inode =jffs2_destroy_inode,
.read_inode = jffs2_read_inode,
.put_super = jffs2_put_super,
.write_super = jffs2_write_super,
.statfs = jffs2_statfs,
.remount_fs = jffs2_remount_fs,
.clear_inode = jffs2_clear_inode,
.dirty_inode = jffs2_dirty_inode,
.sync_fs = jffs2_sync_fs,
};
static int jffs2_sb_compare(struct super_block *sb, void *data)
{
struct jffs2_sb_info *p = data;
struct jffs2_sb_info *c = JFFS2_SB_INFO(sb);
/* The superblocks are considered to be equivalent if the underlying MTD
device is the same one */
if (c->mtd == p->mtd) {
D1(printk(KERN_DEBUG "jffs2_sb_compare: match on device %d (\"%s\")\n", p->mtd->index, p->mtd->name));
return 1;
} else {
D1(printk(KERN_DEBUG "jffs2_sb_compare: No match, device %d (\"%s\"), device %d (\"%s\")\n",
c->mtd->index, c->mtd->name, p->mtd->index, p->mtd->name));
return 0;
}
}
static int jffs2_sb_set(struct super_block *sb, void *data)
{
struct jffs2_sb_info *p = data;
/* For persistence of NFS exports etc. we use the same s_dev
each time we mount the device, don't just use an anonymous
device */
sb->s_fs_info = p;
p->os_priv = sb;
sb->s_dev = MKDEV(MTD_BLOCK_MAJOR, p->mtd->index);
return 0;
}
static int jffs2_get_sb_mtd(struct file_system_type *fs_type,
int flags, const char *dev_name,
void *data, struct mtd_info *mtd,
struct vfsmount *mnt)
{
struct super_block *sb;
struct jffs2_sb_info *c;
int ret;
c = kmalloc(sizeof(*c), GFP_KERNEL);
if (!c)
return -ENOMEM;
memset(c, 0, sizeof(*c));
c->mtd = mtd;
sb = sget(fs_type, jffs2_sb_compare, jffs2_sb_set, c);
if (IS_ERR(sb))
goto out_error;
if (sb->s_root) {
/* New mountpoint for JFFS2 which is already mounted */
D1(printk(KERN_DEBUG "jffs2_get_sb_mtd(): Device %d (\"%s\") is already mounted\n",
mtd->index, mtd->name));
ret = simple_set_mnt(mnt, sb);
goto out_put;
}
D1(printk(KERN_DEBUG "jffs2_get_sb_mtd(): New superblock for device %d (\"%s\")\n",
mtd->index, mtd->name));
/* Initialize JFFS2 superblock locks, the further initialization will be
* done later */
init_MUTEX(&c->alloc_sem);
init_MUTEX(&c->erase_free_sem);
init_waitqueue_head(&c->erase_wait);
init_waitqueue_head(&c->inocache_wq);
spin_lock_init(&c->erase_completion_lock);
spin_lock_init(&c->inocache_lock);
sb->s_op = &jffs2_super_operations;
sb->s_flags = flags | MS_NOATIME;
sb->s_xattr = jffs2_xattr_handlers;
#ifdef CONFIG_JFFS2_FS_POSIX_ACL
sb->s_flags |= MS_POSIXACL;
#endif
ret = jffs2_do_fill_super(sb, data, flags & MS_SILENT ? 1 : 0);
if (ret) {
/* Failure case... */
up_write(&sb->s_umount);
deactivate_super(sb);
return ret;
}
sb->s_flags |= MS_ACTIVE;
return simple_set_mnt(mnt, sb);
out_error:
ret = PTR_ERR(sb);
out_put:
kfree(c);
put_mtd_device(mtd);
return ret;
}
static int jffs2_get_sb_mtdnr(struct file_system_type *fs_type,
int flags, const char *dev_name,
void *data, int mtdnr,
struct vfsmount *mnt)
{
struct mtd_info *mtd;
mtd = get_mtd_device(NULL, mtdnr);
if (!mtd) {
D1(printk(KERN_DEBUG "jffs2: MTD device #%u doesn't appear to exist\n", mtdnr));
return -EINVAL;
}
return jffs2_get_sb_mtd(fs_type, flags, dev_name, data, mtd, mnt);
}
static int jffs2_get_sb(struct file_system_type *fs_type,
int flags, const char *dev_name,
void *data, struct vfsmount *mnt)
{
int err;
struct nameidata nd;
int mtdnr;
if (!dev_name)
return -EINVAL;
D1(printk(KERN_DEBUG "jffs2_get_sb(): dev_name \"%s\"\n", dev_name));
/* The preferred way of mounting in future; especially when
CONFIG_BLK_DEV is implemented - we specify the underlying
MTD device by number or by name, so that we don't require
block device support to be present in the kernel. */
/* FIXME: How to do the root fs this way? */
if (dev_name[0] == 'm' && dev_name[1] == 't' && dev_name[2] == 'd') {
/* Probably mounting without the blkdev crap */
if (dev_name[3] == ':') {
struct mtd_info *mtd;
/* Mount by MTD device name */
D1(printk(KERN_DEBUG "jffs2_get_sb(): mtd:%%s, name \"%s\"\n", dev_name+4));
for (mtdnr = 0; mtdnr < MAX_MTD_DEVICES; mtdnr++) {
mtd = get_mtd_device(NULL, mtdnr);
if (mtd) {
if (!strcmp(mtd->name, dev_name+4))
return jffs2_get_sb_mtd(fs_type, flags, dev_name, data, mtd, mnt);
put_mtd_device(mtd);
}
}
printk(KERN_NOTICE "jffs2_get_sb(): MTD device with name \"%s\" not found.\n", dev_name+4);
} else if (isdigit(dev_name[3])) {
/* Mount by MTD device number name */
char *endptr;
mtdnr = simple_strtoul(dev_name+3, &endptr, 0);
if (!*endptr) {
/* It was a valid number */
D1(printk(KERN_DEBUG "jffs2_get_sb(): mtd%%d, mtdnr %d\n", mtdnr));
return jffs2_get_sb_mtdnr(fs_type, flags, dev_name, data, mtdnr, mnt);
}
}
}
/* Try the old way - the hack where we allowed users to mount
/dev/mtdblock$(n) but didn't actually _use_ the blkdev */
err = path_lookup(dev_name, LOOKUP_FOLLOW, &nd);
D1(printk(KERN_DEBUG "jffs2_get_sb(): path_lookup() returned %d, inode %p\n",
err, nd.dentry->d_inode));
if (err)
return err;
err = -EINVAL;
if (!S_ISBLK(nd.dentry->d_inode->i_mode))
goto out;
if (nd.mnt->mnt_flags & MNT_NODEV) {
err = -EACCES;
goto out;
}
if (imajor(nd.dentry->d_inode) != MTD_BLOCK_MAJOR) {
if (!(flags & MS_SILENT))
printk(KERN_NOTICE "Attempt to mount non-MTD device \"%s\" as JFFS2\n",
dev_name);
goto out;
}
mtdnr = iminor(nd.dentry->d_inode);
path_release(&nd);
return jffs2_get_sb_mtdnr(fs_type, flags, dev_name, data, mtdnr, mnt);
out:
path_release(&nd);
return err;
}
static void jffs2_put_super (struct super_block *sb)
{
struct jffs2_sb_info *c = JFFS2_SB_INFO(sb);
D2(printk(KERN_DEBUG "jffs2: jffs2_put_super()\n"));
down(&c->alloc_sem);
jffs2_flush_wbuf_pad(c);
up(&c->alloc_sem);
jffs2_sum_exit(c);
jffs2_free_ino_caches(c);
jffs2_free_raw_node_refs(c);
if (jffs2_blocks_use_vmalloc(c))
vfree(c->blocks);
else
kfree(c->blocks);
jffs2_flash_cleanup(c);
kfree(c->inocache_list);
jffs2_clear_xattr_subsystem(c);
if (c->mtd->sync)
c->mtd->sync(c->mtd);
D1(printk(KERN_DEBUG "jffs2_put_super returning\n"));
}
static void jffs2_kill_sb(struct super_block *sb)
{
struct jffs2_sb_info *c = JFFS2_SB_INFO(sb);
if (!(sb->s_flags & MS_RDONLY))
jffs2_stop_garbage_collect_thread(c);
generic_shutdown_super(sb);
put_mtd_device(c->mtd);
kfree(c);
}
static struct file_system_type jffs2_fs_type = {
.owner = THIS_MODULE,
.name = "jffs2",
.get_sb = jffs2_get_sb,
.kill_sb = jffs2_kill_sb,
};
static int __init init_jffs2_fs(void)
{
int ret;
/* Paranoia checks for on-medium structures. If we ask GCC
to pack them with __attribute__((packed)) then it _also_
assumes that they're not aligned -- so it emits crappy
code on some architectures. Ideally we want an attribute
which means just 'no padding', without the alignment
thing. But GCC doesn't have that -- we have to just
hope the structs are the right sizes, instead. */
BUG_ON(sizeof(struct jffs2_unknown_node) != 12);
BUG_ON(sizeof(struct jffs2_raw_dirent) != 40);
BUG_ON(sizeof(struct jffs2_raw_inode) != 68);
BUG_ON(sizeof(struct jffs2_raw_summary) != 32);
printk(KERN_INFO "JFFS2 version 2.2."
#ifdef CONFIG_JFFS2_FS_WRITEBUFFER
" (NAND)"
#endif
#ifdef CONFIG_JFFS2_SUMMARY
" (SUMMARY) "
#endif
" (C) 2001-2006 Red Hat, Inc.\n");
jffs2_inode_cachep = kmem_cache_create("jffs2_i",
sizeof(struct jffs2_inode_info),
0, (SLAB_RECLAIM_ACCOUNT|
SLAB_MEM_SPREAD),
jffs2_i_init_once, NULL);
if (!jffs2_inode_cachep) {
printk(KERN_ERR "JFFS2 error: Failed to initialise inode cache\n");
return -ENOMEM;
}
ret = jffs2_compressors_init();
if (ret) {
printk(KERN_ERR "JFFS2 error: Failed to initialise compressors\n");
goto out;
}
ret = jffs2_create_slab_caches();
if (ret) {
printk(KERN_ERR "JFFS2 error: Failed to initialise slab caches\n");
goto out_compressors;
}
ret = register_filesystem(&jffs2_fs_type);
if (ret) {
printk(KERN_ERR "JFFS2 error: Failed to register filesystem\n");
goto out_slab;
}
return 0;
out_slab:
jffs2_destroy_slab_caches();
out_compressors:
jffs2_compressors_exit();
out:
kmem_cache_destroy(jffs2_inode_cachep);
return ret;
}
static void __exit exit_jffs2_fs(void)
{
unregister_filesystem(&jffs2_fs_type);
jffs2_destroy_slab_caches();
jffs2_compressors_exit();
kmem_cache_destroy(jffs2_inode_cachep);
}
module_init(init_jffs2_fs);
module_exit(exit_jffs2_fs);
MODULE_DESCRIPTION("The Journalling Flash File System, v2");
MODULE_AUTHOR("Red Hat, Inc.");
MODULE_LICENSE("GPL"); // Actually dual-licensed, but it doesn't matter for
// the sake of this tag. It's Free Software.