remarkable-linux/fs/filesystems.c
Nick Piggin 31e6b01f41 fs: rcu-walk for path lookup
Perform common cases of path lookups without any stores or locking in the
ancestor dentry elements. This is called rcu-walk, as opposed to the current
algorithm which is a refcount based walk, or ref-walk.

This results in far fewer atomic operations on every path element,
significantly improving path lookup performance. It also avoids cacheline
bouncing on common dentries, significantly improving scalability.

The overall design is like this:
* LOOKUP_RCU is set in nd->flags, which distinguishes rcu-walk from ref-walk.
* Take the RCU lock for the entire path walk, starting with the acquiring
  of the starting path (eg. root/cwd/fd-path). So now dentry refcounts are
  not required for dentry persistence.
* synchronize_rcu is called when unregistering a filesystem, so we can
  access d_ops and i_ops during rcu-walk.
* Similarly take the vfsmount lock for the entire path walk. So now mnt
  refcounts are not required for persistence. Also we are free to perform mount
  lookups, and to assume dentry mount points and mount roots are stable up and
  down the path.
* Have a per-dentry seqlock to protect the dentry name, parent, and inode,
  so we can load this tuple atomically, and also check whether any of its
  members have changed.
* Dentry lookups (based on parent, candidate string tuple) recheck the parent
  sequence after the child is found in case anything changed in the parent
  during the path walk.
* inode is also RCU protected so we can load d_inode and use the inode for
  limited things.
* i_mode, i_uid, i_gid can be tested for exec permissions during path walk.
* i_op can be loaded.

When we reach the destination dentry, we lock it, recheck lookup sequence,
and increment its refcount and mountpoint refcount. RCU and vfsmount locks
are dropped. This is termed "dropping rcu-walk". If the dentry refcount does
not match, we can not drop rcu-walk gracefully at the current point in the
lokup, so instead return -ECHILD (for want of a better errno). This signals the
path walking code to re-do the entire lookup with a ref-walk.

Aside from the final dentry, there are other situations that may be encounted
where we cannot continue rcu-walk. In that case, we drop rcu-walk (ie. take
a reference on the last good dentry) and continue with a ref-walk. Again, if
we can drop rcu-walk gracefully, we return -ECHILD and do the whole lookup
using ref-walk. But it is very important that we can continue with ref-walk
for most cases, particularly to avoid the overhead of double lookups, and to
gain the scalability advantages on common path elements (like cwd and root).

The cases where rcu-walk cannot continue are:
* NULL dentry (ie. any uncached path element)
* parent with d_inode->i_op->permission or ACLs
* dentries with d_revalidate
* Following links

In future patches, permission checks and d_revalidate become rcu-walk aware. It
may be possible eventually to make following links rcu-walk aware.

Uncached path elements will always require dropping to ref-walk mode, at the
very least because i_mutex needs to be grabbed, and objects allocated.

Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 17:50:27 +11:00

289 lines
6.4 KiB
C

/*
* linux/fs/filesystems.c
*
* Copyright (C) 1991, 1992 Linus Torvalds
*
* table of configured filesystems
*/
#include <linux/syscalls.h>
#include <linux/fs.h>
#include <linux/proc_fs.h>
#include <linux/seq_file.h>
#include <linux/kmod.h>
#include <linux/init.h>
#include <linux/module.h>
#include <linux/slab.h>
#include <asm/uaccess.h>
/*
* Handling of filesystem drivers list.
* Rules:
* Inclusion to/removals from/scanning of list are protected by spinlock.
* During the unload module must call unregister_filesystem().
* We can access the fields of list element if:
* 1) spinlock is held or
* 2) we hold the reference to the module.
* The latter can be guaranteed by call of try_module_get(); if it
* returned 0 we must skip the element, otherwise we got the reference.
* Once the reference is obtained we can drop the spinlock.
*/
static struct file_system_type *file_systems;
static DEFINE_RWLOCK(file_systems_lock);
/* WARNING: This can be used only if we _already_ own a reference */
void get_filesystem(struct file_system_type *fs)
{
__module_get(fs->owner);
}
void put_filesystem(struct file_system_type *fs)
{
module_put(fs->owner);
}
static struct file_system_type **find_filesystem(const char *name, unsigned len)
{
struct file_system_type **p;
for (p=&file_systems; *p; p=&(*p)->next)
if (strlen((*p)->name) == len &&
strncmp((*p)->name, name, len) == 0)
break;
return p;
}
/**
* register_filesystem - register a new filesystem
* @fs: the file system structure
*
* Adds the file system passed to the list of file systems the kernel
* is aware of for mount and other syscalls. Returns 0 on success,
* or a negative errno code on an error.
*
* The &struct file_system_type that is passed is linked into the kernel
* structures and must not be freed until the file system has been
* unregistered.
*/
int register_filesystem(struct file_system_type * fs)
{
int res = 0;
struct file_system_type ** p;
BUG_ON(strchr(fs->name, '.'));
if (fs->next)
return -EBUSY;
INIT_LIST_HEAD(&fs->fs_supers);
write_lock(&file_systems_lock);
p = find_filesystem(fs->name, strlen(fs->name));
if (*p)
res = -EBUSY;
else
*p = fs;
write_unlock(&file_systems_lock);
return res;
}
EXPORT_SYMBOL(register_filesystem);
/**
* unregister_filesystem - unregister a file system
* @fs: filesystem to unregister
*
* Remove a file system that was previously successfully registered
* with the kernel. An error is returned if the file system is not found.
* Zero is returned on a success.
*
* Once this function has returned the &struct file_system_type structure
* may be freed or reused.
*/
int unregister_filesystem(struct file_system_type * fs)
{
struct file_system_type ** tmp;
write_lock(&file_systems_lock);
tmp = &file_systems;
while (*tmp) {
if (fs == *tmp) {
*tmp = fs->next;
fs->next = NULL;
write_unlock(&file_systems_lock);
return 0;
}
tmp = &(*tmp)->next;
}
write_unlock(&file_systems_lock);
synchronize_rcu();
return -EINVAL;
}
EXPORT_SYMBOL(unregister_filesystem);
static int fs_index(const char __user * __name)
{
struct file_system_type * tmp;
char * name;
int err, index;
name = getname(__name);
err = PTR_ERR(name);
if (IS_ERR(name))
return err;
err = -EINVAL;
read_lock(&file_systems_lock);
for (tmp=file_systems, index=0 ; tmp ; tmp=tmp->next, index++) {
if (strcmp(tmp->name,name) == 0) {
err = index;
break;
}
}
read_unlock(&file_systems_lock);
putname(name);
return err;
}
static int fs_name(unsigned int index, char __user * buf)
{
struct file_system_type * tmp;
int len, res;
read_lock(&file_systems_lock);
for (tmp = file_systems; tmp; tmp = tmp->next, index--)
if (index <= 0 && try_module_get(tmp->owner))
break;
read_unlock(&file_systems_lock);
if (!tmp)
return -EINVAL;
/* OK, we got the reference, so we can safely block */
len = strlen(tmp->name) + 1;
res = copy_to_user(buf, tmp->name, len) ? -EFAULT : 0;
put_filesystem(tmp);
return res;
}
static int fs_maxindex(void)
{
struct file_system_type * tmp;
int index;
read_lock(&file_systems_lock);
for (tmp = file_systems, index = 0 ; tmp ; tmp = tmp->next, index++)
;
read_unlock(&file_systems_lock);
return index;
}
/*
* Whee.. Weird sysv syscall.
*/
SYSCALL_DEFINE3(sysfs, int, option, unsigned long, arg1, unsigned long, arg2)
{
int retval = -EINVAL;
switch (option) {
case 1:
retval = fs_index((const char __user *) arg1);
break;
case 2:
retval = fs_name(arg1, (char __user *) arg2);
break;
case 3:
retval = fs_maxindex();
break;
}
return retval;
}
int __init get_filesystem_list(char *buf)
{
int len = 0;
struct file_system_type * tmp;
read_lock(&file_systems_lock);
tmp = file_systems;
while (tmp && len < PAGE_SIZE - 80) {
len += sprintf(buf+len, "%s\t%s\n",
(tmp->fs_flags & FS_REQUIRES_DEV) ? "" : "nodev",
tmp->name);
tmp = tmp->next;
}
read_unlock(&file_systems_lock);
return len;
}
#ifdef CONFIG_PROC_FS
static int filesystems_proc_show(struct seq_file *m, void *v)
{
struct file_system_type * tmp;
read_lock(&file_systems_lock);
tmp = file_systems;
while (tmp) {
seq_printf(m, "%s\t%s\n",
(tmp->fs_flags & FS_REQUIRES_DEV) ? "" : "nodev",
tmp->name);
tmp = tmp->next;
}
read_unlock(&file_systems_lock);
return 0;
}
static int filesystems_proc_open(struct inode *inode, struct file *file)
{
return single_open(file, filesystems_proc_show, NULL);
}
static const struct file_operations filesystems_proc_fops = {
.open = filesystems_proc_open,
.read = seq_read,
.llseek = seq_lseek,
.release = single_release,
};
static int __init proc_filesystems_init(void)
{
proc_create("filesystems", 0, NULL, &filesystems_proc_fops);
return 0;
}
module_init(proc_filesystems_init);
#endif
static struct file_system_type *__get_fs_type(const char *name, int len)
{
struct file_system_type *fs;
read_lock(&file_systems_lock);
fs = *(find_filesystem(name, len));
if (fs && !try_module_get(fs->owner))
fs = NULL;
read_unlock(&file_systems_lock);
return fs;
}
struct file_system_type *get_fs_type(const char *name)
{
struct file_system_type *fs;
const char *dot = strchr(name, '.');
int len = dot ? dot - name : strlen(name);
fs = __get_fs_type(name, len);
if (!fs && (request_module("%.*s", len, name) == 0))
fs = __get_fs_type(name, len);
if (dot && fs && !(fs->fs_flags & FS_HAS_SUBTYPE)) {
put_filesystem(fs);
fs = NULL;
}
return fs;
}
EXPORT_SYMBOL(get_fs_type);