x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

When using eBPF to probe CS CALL instructions within a function,
a crash can be triggered.

The eBPF tool probes offset 257 of the __hrtimer_run_queues()
function:

<__hrtimer_run_queues+249>:  nopl   0x0(%rax,%rax,1)
<__hrtimer_run_queues+254>:  mov    %r14,%rdi
<__hrtimer_run_queues+257>:  cs call <__x86_indirect_thunk_r12>
<__hrtimer_run_queues+263>:  mov    %eax,%r12d
<__hrtimer_run_queues+266>:  xchg   %ax,%ax
<__hrtimer_run_queues+268>:  mov    %r13,%rdi

Which triggers this crash:

  BUG: unable to handle page fault for address: 00000000000f41c9
  #PF: supervisor write access in kernel mode
  #PF: error_code(0x0002) - not-present page
  PGD 0 P4D 0
  Oops: 0002 [#1] SMP NOPTI
  CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P
  RIP: 0010:__hrtimer_run_queues+0x106/0x230

Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is
at the 6th byte of the above CS CALL instruction. Since the CS CALL
instruction occupies 6 bytes, the exception occurred in the middle of that
call instruction.

The root cause is that when using eBPF tools to probe in the middle of a
function, a kprobe with INT3 is used as the underlying implementation.

During single-step emulation of the original CALL instruction,
int3_emulate_call() assumes that the probed CALL instruction is 5 bytes
long. However, the actual CS-prefixed CALL instruction occupies 6 bytes,
so it constructs an incorrect exception return address. When the CPU
returns from the kprobe handler, the next instruction to be executed is at
the address of the last byte of that CS CALL instruction. Coincidentally,
starting from that address, the CPU fetches and decodes a completely
different instruction, which ultimately triggers a kernel crash.

Fix the issue by using the actual instruction length obtained from
the instruction decoder when constructing the exception return
address, rather than relying on the hardcoded CALL_INSN_SIZE macro.

[ mingo: Refined the changelog ]

Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step")
Suggested-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 672 insertions, 0 deletions

diff --git a/fs/romfs/super.c b/fs/romfs/super.c
new file mode 100644
index 000000000..7f783d617
--- /dev/null
+++ b/fs/romfs/super.c
@@ -0,0 +1,672 @@
+/* Block- or MTD-based romfs
+ *
+ * Copyright © 2007 Red Hat, Inc. All Rights Reserved.
+ * Written by David Howells (dhowells@redhat.com)
+ *
+ * Derived from: ROMFS file system, Linux implementation
+ *
+ * Copyright © 1997-1999  Janos Farkas <chexum@shadow.banki.hu>
+ *
+ * Using parts of the minix filesystem
+ * Copyright © 1991, 1992  Linus Torvalds
+ *
+ * and parts of the affs filesystem additionally
+ * Copyright © 1993  Ray Burr
+ * Copyright © 1996  Hans-Joachim Widmaier
+ *
+ * Changes
+ *					Changed for 2.1.19 modules
+ *	Jan 1997			Initial release
+ *	Jun 1997			2.1.43+ changes
+ *					Proper page locking in read_folio
+ *					Changed to work with 2.1.45+ fs
+ *	Jul 1997			Fixed follow_link
+ *			2.1.47
+ *					lookup shouldn't return -ENOENT
+ *					from Horst von Brand:
+ *					  fail on wrong checksum
+ *					  double unlock_super was possible
+ *					  correct namelen for statfs
+ *					spotted by Bill Hawes:
+ *					  readlink shouldn't iput()
+ *	Jun 1998	2.1.106		from Avery Pennarun: glibc scandir()
+ *					  exposed a problem in readdir
+ *			2.1.107		code-freeze spellchecker run
+ *	Aug 1998			2.1.118+ VFS changes
+ *	Sep 1998	2.1.122		another VFS change (follow_link)
+ *	Apr 1999	2.2.7		no more EBADF checking in
+ *					  lookup/readdir, use ERR_PTR
+ *	Jun 1999	2.3.6		d_alloc_root use changed
+ *			2.3.9		clean up usage of ENOENT/negative
+ *					  dentries in lookup
+ *					clean up page flags setting
+ *					  (error, uptodate, locking) in
+ *					  in read_folio
+ *					use init_special_inode for
+ *					  fifos/sockets (and streamline) in
+ *					  read_inode, fix _ops table order
+ *	Aug 1999	2.3.16		__initfunc() => __init change
+ *	Oct 1999	2.3.24		page->owner hack obsoleted
+ *	Nov 1999	2.3.27		2.3.25+ page->offset => index change
+ *
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public Licence
+ * as published by the Free Software Foundation; either version
+ * 2 of the Licence, or (at your option) any later version.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/module.h>
+#include <linux/string.h>
+#include <linux/fs.h>
+#include <linux/time.h>
+#include <linux/slab.h>
+#include <linux/init.h>
+#include <linux/blkdev.h>
+#include <linux/fs_context.h>
+#include <linux/mount.h>
+#include <linux/namei.h>
+#include <linux/statfs.h>
+#include <linux/mtd/super.h>
+#include <linux/ctype.h>
+#include <linux/highmem.h>
+#include <linux/pagemap.h>
+#include <linux/uaccess.h>
+#include <linux/major.h>
+#include "internal.h"
+
+static struct kmem_cache *romfs_inode_cachep;
+
+static const umode_t romfs_modemap[8] = {
+	0,			/* hard link */
+	S_IFDIR  | 0644,	/* directory */
+	S_IFREG  | 0644,	/* regular file */
+	S_IFLNK  | 0777,	/* symlink */
+	S_IFBLK  | 0600,	/* blockdev */
+	S_IFCHR  | 0600,	/* chardev */
+	S_IFSOCK | 0644,	/* socket */
+	S_IFIFO  | 0644		/* FIFO */
+};
+
+static const unsigned char romfs_dtype_table[] = {
+	DT_UNKNOWN, DT_DIR, DT_REG, DT_LNK, DT_BLK, DT_CHR, DT_SOCK, DT_FIFO
+};
+
+static struct inode *romfs_iget(struct super_block *sb, unsigned long pos);
+
+/*
+ * read a page worth of data from the image
+ */
+static int romfs_read_folio(struct file *file, struct folio *folio)
+{
+	struct inode *inode = folio->mapping->host;
+	loff_t offset, size;
+	unsigned long fillsize, pos;
+	void *buf;
+	int ret;
+
+	buf = kmap_local_folio(folio, 0);
+
+	offset = folio_pos(folio);
+	size = i_size_read(inode);
+	fillsize = 0;
+	ret = 0;
+	if (offset < size) {
+		size -= offset;
+		fillsize = size > PAGE_SIZE ? PAGE_SIZE : size;
+
+		pos = ROMFS_I(inode)->i_dataoffset + offset;
+
+		ret = romfs_dev_read(inode->i_sb, pos, buf, fillsize);
+		if (ret < 0) {
+			fillsize = 0;
+			ret = -EIO;
+		}
+	}
+
+	buf = folio_zero_tail(folio, fillsize, buf + fillsize);
+	kunmap_local(buf);
+	folio_end_read(folio, ret == 0);
+	return ret;
+}
+
+static const struct address_space_operations romfs_aops = {
+	.read_folio	= romfs_read_folio
+};
+
+/*
+ * read the entries from a directory
+ */
+static int romfs_readdir(struct file *file, struct dir_context *ctx)
+{
+	struct inode *i = file_inode(file);
+	struct romfs_inode ri;
+	unsigned long offset, maxoff;
+	int j, ino, nextfh;
+	char fsname[ROMFS_MAXFN];	/* XXX dynamic? */
+	int ret;
+
+	maxoff = romfs_maxsize(i->i_sb);
+
+	offset = ctx->pos;
+	if (!offset) {
+		offset = i->i_ino & ROMFH_MASK;
+		ret = romfs_dev_read(i->i_sb, offset, &ri, ROMFH_SIZE);
+		if (ret < 0)
+			goto out;
+		offset = be32_to_cpu(ri.spec) & ROMFH_MASK;
+	}
+
+	/* Not really failsafe, but we are read-only... */
+	for (;;) {
+		if (!offset || offset >= maxoff) {
+			offset = maxoff;
+			ctx->pos = offset;
+			goto out;
+		}
+		ctx->pos = offset;
+
+		/* Fetch inode info */
+		ret = romfs_dev_read(i->i_sb, offset, &ri, ROMFH_SIZE);
+		if (ret < 0)
+			goto out;
+
+		j = romfs_dev_strnlen(i->i_sb, offset + ROMFH_SIZE,
+				      sizeof(fsname) - 1);
+		if (j < 0)
+			goto out;
+
+		ret = romfs_dev_read(i->i_sb, offset + ROMFH_SIZE, fsname, j);
+		if (ret < 0)
+			goto out;
+		fsname[j] = '\0';
+
+		ino = offset;
+		nextfh = be32_to_cpu(ri.next);
+		if ((nextfh & ROMFH_TYPE) == ROMFH_HRD)
+			ino = be32_to_cpu(ri.spec);
+		if (!dir_emit(ctx, fsname, j, ino,
+			    romfs_dtype_table[nextfh & ROMFH_TYPE]))
+			goto out;
+
+		offset = nextfh & ROMFH_MASK;
+	}
+out:
+	return 0;
+}
+
+/*
+ * look up an entry in a directory
+ */
+static struct dentry *romfs_lookup(struct inode *dir, struct dentry *dentry,
+				   unsigned int flags)
+{
+	unsigned long offset, maxoff;
+	struct inode *inode = NULL;
+	struct romfs_inode ri;
+	const char *name;		/* got from dentry */
+	int len, ret;
+
+	offset = dir->i_ino & ROMFH_MASK;
+	ret = romfs_dev_read(dir->i_sb, offset, &ri, ROMFH_SIZE);
+	if (ret < 0)
+		goto error;
+
+	/* search all the file entries in the list starting from the one
+	 * pointed to by the directory's special data */
+	maxoff = romfs_maxsize(dir->i_sb);
+	offset = be32_to_cpu(ri.spec) & ROMFH_MASK;
+
+	name = dentry->d_name.name;
+	len = dentry->d_name.len;
+
+	for (;;) {
+		if (!offset || offset >= maxoff)
+			break;
+
+		ret = romfs_dev_read(dir->i_sb, offset, &ri, sizeof(ri));
+		if (ret < 0)
+			goto error;
+
+		/* try to match the first 16 bytes of name */
+		ret = romfs_dev_strcmp(dir->i_sb, offset + ROMFH_SIZE, name,
+				       len);
+		if (ret < 0)
+			goto error;
+		if (ret == 1) {
+			/* Hard link handling */
+			if ((be32_to_cpu(ri.next) & ROMFH_TYPE) == ROMFH_HRD)
+				offset = be32_to_cpu(ri.spec) & ROMFH_MASK;
+			inode = romfs_iget(dir->i_sb, offset);
+			if (IS_ERR(inode))
+				return ERR_CAST(inode);
+			break;
+		}
+
+		/* next entry */
+		offset = be32_to_cpu(ri.next) & ROMFH_MASK;
+	}
+
+	return d_splice_alias(inode, dentry);
+error:
+	return ERR_PTR(ret);
+}
+
+static const struct file_operations romfs_dir_operations = {
+	.read		= generic_read_dir,
+	.iterate_shared	= romfs_readdir,
+	.llseek		= generic_file_llseek,
+};
+
+static const struct inode_operations romfs_dir_inode_operations = {
+	.lookup		= romfs_lookup,
+};
+
+#define ROMFS_MAX_HARDLINK_DEPTH 64
+
+/*
+ * get a romfs inode based on its position in the image (which doubles as the
+ * inode number)
+ */
+static struct inode *romfs_iget(struct super_block *sb, unsigned long pos)
+{
+	struct romfs_inode_info *inode;
+	struct romfs_inode ri;
+	struct inode *i;
+	unsigned long nlen;
+	unsigned nextfh;
+	unsigned int depth = 0;
+	int ret;
+	umode_t mode;
+
+	/* we might have to traverse a chain of "hard link" file entries to get
+	 * to the actual file */
+	for (;;) {
+		ret = romfs_dev_read(sb, pos, &ri, sizeof(ri));
+		if (ret < 0)
+			goto error;
+
+		/* XXX: do romfs_checksum here too (with name) */
+
+		nextfh = be32_to_cpu(ri.next);
+		if ((nextfh & ROMFH_TYPE) != ROMFH_HRD)
+			break;
+
+		if (++depth > ROMFS_MAX_HARDLINK_DEPTH)
+			return ERR_PTR(-ELOOP);
+
+		pos = be32_to_cpu(ri.spec) & ROMFH_MASK;
+	}
+
+	/* determine the length of the filename */
+	nlen = romfs_dev_strnlen(sb, pos + ROMFH_SIZE, ROMFS_MAXFN);
+	if (IS_ERR_VALUE(nlen))
+		goto eio;
+
+	/* get an inode for this image position */
+	i = iget_locked(sb, pos);
+	if (!i)
+		return ERR_PTR(-ENOMEM);
+
+	if (!(inode_state_read_once(i) & I_NEW))
+		return i;
+
+	/* precalculate the data offset */
+	inode = ROMFS_I(i);
+	inode->i_metasize = (ROMFH_SIZE + nlen + 1 + ROMFH_PAD) & ROMFH_MASK;
+	inode->i_dataoffset = pos + inode->i_metasize;
+
+	set_nlink(i, 1);		/* Hard to decide.. */
+	i->i_size = be32_to_cpu(ri.size);
+	inode_set_mtime_to_ts(i,
+			      inode_set_atime_to_ts(i, inode_set_ctime(i, 0, 0)));
+
+	/* set up mode and ops */
+	mode = romfs_modemap[nextfh & ROMFH_TYPE];
+
+	switch (nextfh & ROMFH_TYPE) {
+	case ROMFH_DIR:
+		i->i_size = ROMFS_I(i)->i_metasize;
+		i->i_op = &romfs_dir_inode_operations;
+		i->i_fop = &romfs_dir_operations;
+		if (nextfh & ROMFH_EXEC)
+			mode |= S_IXUGO;
+		break;
+	case ROMFH_REG:
+		i->i_fop = &romfs_ro_fops;
+		i->i_data.a_ops = &romfs_aops;
+		if (nextfh & ROMFH_EXEC)
+			mode |= S_IXUGO;
+		break;
+	case ROMFH_SYM:
+		i->i_op = &page_symlink_inode_operations;
+		inode_nohighmem(i);
+		i->i_data.a_ops = &romfs_aops;
+		mode |= S_IRWXUGO;
+		break;
+	default:
+		/* depending on MBZ for sock/fifos */
+		nextfh = be32_to_cpu(ri.spec);
+		init_special_inode(i, mode, MKDEV(nextfh >> 16,
+						  nextfh & 0xffff));
+		break;
+	}
+
+	i->i_mode = mode;
+	i->i_blocks = (i->i_size + 511) >> 9;
+
+	unlock_new_inode(i);
+	return i;
+
+eio:
+	ret = -EIO;
+error:
+	pr_err("read error for inode 0x%lx\n", pos);
+	return ERR_PTR(ret);
+}
+
+/*
+ * allocate a new inode
+ */
+static struct inode *romfs_alloc_inode(struct super_block *sb)
+{
+	struct romfs_inode_info *inode;
+
+	inode = alloc_inode_sb(sb, romfs_inode_cachep, GFP_KERNEL);
+	return inode ? &inode->vfs_inode : NULL;
+}
+
+/*
+ * return a spent inode to the slab cache
+ */
+static void romfs_free_inode(struct inode *inode)
+{
+	kmem_cache_free(romfs_inode_cachep, ROMFS_I(inode));
+}
+
+/*
+ * get filesystem statistics
+ */
+static int romfs_statfs(struct dentry *dentry, struct kstatfs *buf)
+{
+	struct super_block *sb = dentry->d_sb;
+	u64 id = 0;
+
+	/* When calling huge_encode_dev(),
+	 * use sb->s_bdev->bd_dev when,
+	 *   - CONFIG_ROMFS_ON_BLOCK defined
+	 * use sb->s_dev when,
+	 *   - CONFIG_ROMFS_ON_BLOCK undefined and
+	 *   - CONFIG_ROMFS_ON_MTD defined
+	 * leave id as 0 when,
+	 *   - CONFIG_ROMFS_ON_BLOCK undefined and
+	 *   - CONFIG_ROMFS_ON_MTD undefined
+	 */
+	if (sb->s_bdev)
+		id = huge_encode_dev(sb->s_bdev->bd_dev);
+	else if (sb->s_dev)
+		id = huge_encode_dev(sb->s_dev);
+
+	buf->f_type = ROMFS_MAGIC;
+	buf->f_namelen = ROMFS_MAXFN;
+	buf->f_bsize = ROMBSIZE;
+	buf->f_bfree = buf->f_bavail = buf->f_ffree;
+	buf->f_blocks =
+		(romfs_maxsize(dentry->d_sb) + ROMBSIZE - 1) >> ROMBSBITS;
+	buf->f_fsid = u64_to_fsid(id);
+	return 0;
+}
+
+/*
+ * remounting must involve read-only
+ */
+static int romfs_reconfigure(struct fs_context *fc)
+{
+	sync_filesystem(fc->root->d_sb);
+	fc->sb_flags |= SB_RDONLY;
+	return 0;
+}
+
+static const struct super_operations romfs_super_ops = {
+	.alloc_inode	= romfs_alloc_inode,
+	.free_inode	= romfs_free_inode,
+	.statfs		= romfs_statfs,
+};
+
+/*
+ * checksum check on part of a romfs filesystem
+ */
+static __u32 romfs_checksum(const void *data, int size)
+{
+	const __be32 *ptr = data;
+	__u32 sum;
+
+	sum = 0;
+	size >>= 2;
+	while (size > 0) {
+		sum += be32_to_cpu(*ptr++);
+		size--;
+	}
+	return sum;
+}
+
+/*
+ * fill in the superblock
+ */
+static int romfs_fill_super(struct super_block *sb, struct fs_context *fc)
+{
+	struct romfs_super_block *rsb;
+	struct inode *root;
+	unsigned long pos, img_size;
+	const char *storage;
+	size_t len;
+	int ret;
+
+#ifdef CONFIG_BLOCK
+	if (!sb->s_mtd) {
+		if (!sb_set_blocksize(sb, ROMBSIZE)) {
+			errorf(fc, "romfs: unable to set blocksize\n");
+			return -EINVAL;
+		}
+	} else {
+		sb->s_blocksize = ROMBSIZE;
+		sb->s_blocksize_bits = blksize_bits(ROMBSIZE);
+	}
+#endif
+
+	sb->s_maxbytes = 0xFFFFFFFF;
+	sb->s_magic = ROMFS_MAGIC;
+	sb->s_flags |= SB_RDONLY | SB_NOATIME;
+	sb->s_time_min = 0;
+	sb->s_time_max = 0;
+	sb->s_op = &romfs_super_ops;
+
+#ifdef CONFIG_ROMFS_ON_MTD
+	/* Use same dev ID from the underlying mtdblock device */
+	if (sb->s_mtd)
+		sb->s_dev = MKDEV(MTD_BLOCK_MAJOR, sb->s_mtd->index);
+#endif
+	/* read the image superblock and check it */
+	rsb = kmalloc(512, GFP_KERNEL);
+	if (!rsb)
+		return -ENOMEM;
+
+	sb->s_fs_info = (void *) 512;
+	ret = romfs_dev_read(sb, 0, rsb, 512);
+	if (ret < 0)
+		goto error_rsb;
+
+	img_size = be32_to_cpu(rsb->size);
+
+	if (sb->s_mtd && img_size > sb->s_mtd->size)
+		goto error_rsb_inval;
+
+	sb->s_fs_info = (void *) img_size;
+
+	if (rsb->word0 != ROMSB_WORD0 || rsb->word1 != ROMSB_WORD1 ||
+	    img_size < ROMFH_SIZE) {
+		if (!(fc->sb_flags & SB_SILENT))
+			errorf(fc, "VFS: Can't find a romfs filesystem on dev %s.\n",
+			       sb->s_id);
+		goto error_rsb_inval;
+	}
+
+	if (romfs_checksum(rsb, min_t(size_t, img_size, 512))) {
+		pr_err("bad initial checksum on dev %s.\n", sb->s_id);
+		goto error_rsb_inval;
+	}
+
+	storage = sb->s_mtd ? "MTD" : "the block layer";
+
+	len = strnlen(rsb->name, ROMFS_MAXFN);
+	if (!(fc->sb_flags & SB_SILENT))
+		pr_notice("Mounting image '%*.*s' through %s\n",
+			  (unsigned) len, (unsigned) len, rsb->name, storage);
+
+	kfree(rsb);
+	rsb = NULL;
+
+	/* find the root directory */
+	pos = (ROMFH_SIZE + len + 1 + ROMFH_PAD) & ROMFH_MASK;
+
+	root = romfs_iget(sb, pos);
+	if (IS_ERR(root))
+		return PTR_ERR(root);
+
+	sb->s_root = d_make_root(root);
+	if (!sb->s_root)
+		return -ENOMEM;
+
+	return 0;
+
+error_rsb_inval:
+	ret = -EINVAL;
+error_rsb:
+	kfree(rsb);
+	return ret;
+}
+
+/*
+ * get a superblock for mounting
+ */
+static int romfs_get_tree(struct fs_context *fc)
+{
+	int ret = -EINVAL;
+
+#ifdef CONFIG_ROMFS_ON_MTD
+	ret = get_tree_mtd(fc, romfs_fill_super);
+#endif
+#ifdef CONFIG_ROMFS_ON_BLOCK
+	if (ret == -EINVAL)
+		ret = get_tree_bdev(fc, romfs_fill_super);
+#endif
+	return ret;
+}
+
+static const struct fs_context_operations romfs_context_ops = {
+	.get_tree	= romfs_get_tree,
+	.reconfigure	= romfs_reconfigure,
+};
+
+/*
+ * Set up the filesystem mount context.
+ */
+static int romfs_init_fs_context(struct fs_context *fc)
+{
+	fc->ops = &romfs_context_ops;
+	return 0;
+}
+
+/*
+ * destroy a romfs superblock in the appropriate manner
+ */
+static void romfs_kill_sb(struct super_block *sb)
+{
+	generic_shutdown_super(sb);
+
+#ifdef CONFIG_ROMFS_ON_MTD
+	if (sb->s_mtd) {
+		put_mtd_device(sb->s_mtd);
+		sb->s_mtd = NULL;
+	}
+#endif
+#ifdef CONFIG_ROMFS_ON_BLOCK
+	if (sb->s_bdev) {
+		sync_blockdev(sb->s_bdev);
+		fs_bdev_file_release(sb->s_bdev_file, sb);
+	}
+#endif
+}
+
+static struct file_system_type romfs_fs_type = {
+	.owner		= THIS_MODULE,
+	.name		= "romfs",
+	.init_fs_context = romfs_init_fs_context,
+	.kill_sb	= romfs_kill_sb,
+	.fs_flags	= FS_REQUIRES_DEV,
+};
+MODULE_ALIAS_FS("romfs");
+
+/*
+ * inode storage initialiser
+ */
+static void romfs_i_init_once(void *_inode)
+{
+	struct romfs_inode_info *inode = _inode;
+
+	inode_init_once(&inode->vfs_inode);
+}
+
+/*
+ * romfs module initialisation
+ */
+static int __init init_romfs_fs(void)
+{
+	int ret;
+
+	pr_info("ROMFS MTD (C) 2007 Red Hat, Inc.\n");
+
+	romfs_inode_cachep =
+		kmem_cache_create("romfs_i",
+				  sizeof(struct romfs_inode_info), 0,
+				  SLAB_RECLAIM_ACCOUNT | SLAB_ACCOUNT,
+				  romfs_i_init_once);
+
+	if (!romfs_inode_cachep) {
+		pr_err("Failed to initialise inode cache\n");
+		return -ENOMEM;
+	}
+	ret = register_filesystem(&romfs_fs_type);
+	if (ret) {
+		pr_err("Failed to register filesystem\n");
+		goto error_register;
+	}
+	return 0;
+
+error_register:
+	kmem_cache_destroy(romfs_inode_cachep);
+	return ret;
+}
+
+/*
+ * romfs module removal
+ */
+static void __exit exit_romfs_fs(void)
+{
+	unregister_filesystem(&romfs_fs_type);
+	/*
+	 * Make sure all delayed rcu free inodes are flushed before we
+	 * destroy cache.
+	 */
+	rcu_barrier();
+	kmem_cache_destroy(romfs_inode_cachep);
+}
+
+module_init(init_romfs_fs);
+module_exit(exit_romfs_fs);
+
+MODULE_DESCRIPTION("Direct-MTD Capable RomFS");
+MODULE_AUTHOR("Red Hat, Inc.");
+MODULE_LICENSE("GPL"); /* Actually dual-licensed, but it doesn't matter for */