mirror of
https://github.com/torvalds/linux.git
synced 2024-11-12 23:23:03 +00:00
16df6e07d6
-----BEGIN PGP SIGNATURE----- iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCZabMrQAKCRCRxhvAZXjc ovnUAQDgCOonb1tjtTvC8s8IMDUEoaVYZI91KVfsZQSJYN1sdQD+KfJmX1BhJnWG l0cEffGfnWGXMZkZqDgLPHUIPzFrmws= =1b3j -----END PGP SIGNATURE----- Merge tag 'vfs-6.8.netfs' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs Pull netfs updates from Christian Brauner: "This extends the netfs helper library that network filesystems can use to replace their own implementations. Both afs and 9p are ported. cifs is ready as well but the patches are way bigger and will be routed separately once this is merged. That will remove lots of code as well. The overal goal is to get high-level I/O and knowledge of the page cache and ouf of the filesystem drivers. This includes knowledge about the existence of pages and folios The pull request converts afs and 9p. This removes about 800 lines of code from afs and 300 from 9p. For 9p it is now possible to do writes in larger than a page chunks. Additionally, multipage folio support can be turned on for 9p. Separate patches exist for cifs removing another 2000+ lines. I've included detailed information in the individual pulls I took. Summary: - Add NFS-style (and Ceph-style) locking around DIO vs buffered I/O calls to prevent these from happening at the same time. - Support for direct and unbuffered I/O. - Support for write-through caching in the page cache. - O_*SYNC and RWF_*SYNC writes use write-through rather than writing to the page cache and then flushing afterwards. - Support for write-streaming. - Support for write grouping. - Skip reads for which the server could only return zeros or EOF. - The fscache module is now part of the netfs library and the corresponding maintainer entry is updated. - Some helpers from the fscache subsystem are renamed to mark them as belonging to the netfs library. - Follow-up fixes for the netfs library. - Follow-up fixes for the 9p conversion" * tag 'vfs-6.8.netfs' of gitolite.kernel.org:pub/scm/linux/kernel/git/vfs/vfs: (50 commits) netfs: Fix wrong #ifdef hiding wait cachefiles: Fix signed/unsigned mixup netfs: Fix the loop that unmarks folios after writing to the cache netfs: Fix interaction between write-streaming and cachefiles culling netfs: Count DIO writes netfs: Mark netfs_unbuffered_write_iter_locked() static netfs: Fix proc/fs/fscache symlink to point to "netfs" not "../netfs" netfs: Rearrange netfs_io_subrequest to put request pointer first 9p: Use length of data written to the server in preference to error 9p: Do a couple of cleanups 9p: Fix initialisation of netfs_inode for 9p cachefiles: Fix __cachefiles_prepare_write() 9p: Use netfslib read/write_iter afs: Use the netfs write helpers netfs: Export the netfs_sreq tracepoint netfs: Optimise away reads above the point at which there can be no data netfs: Implement a write-through caching option netfs: Provide a launder_folio implementation netfs: Provide a writepages implementation netfs, cachefiles: Pass upper bound length to allow expansion ...
399 lines
8.8 KiB
C
399 lines
8.8 KiB
C
// SPDX-License-Identifier: GPL-2.0-or-later
|
|
/* AFS dynamic root handling
|
|
*
|
|
* Copyright (C) 2018 Red Hat, Inc. All Rights Reserved.
|
|
* Written by David Howells (dhowells@redhat.com)
|
|
*/
|
|
|
|
#include <linux/fs.h>
|
|
#include <linux/namei.h>
|
|
#include <linux/dns_resolver.h>
|
|
#include "internal.h"
|
|
|
|
static atomic_t afs_autocell_ino;
|
|
|
|
/*
|
|
* iget5() comparator for inode created by autocell operations
|
|
*
|
|
* These pseudo inodes don't match anything.
|
|
*/
|
|
static int afs_iget5_pseudo_test(struct inode *inode, void *opaque)
|
|
{
|
|
return 0;
|
|
}
|
|
|
|
/*
|
|
* iget5() inode initialiser
|
|
*/
|
|
static int afs_iget5_pseudo_set(struct inode *inode, void *opaque)
|
|
{
|
|
struct afs_super_info *as = AFS_FS_S(inode->i_sb);
|
|
struct afs_vnode *vnode = AFS_FS_I(inode);
|
|
struct afs_fid *fid = opaque;
|
|
|
|
vnode->volume = as->volume;
|
|
vnode->fid = *fid;
|
|
inode->i_ino = fid->vnode;
|
|
inode->i_generation = fid->unique;
|
|
return 0;
|
|
}
|
|
|
|
/*
|
|
* Create an inode for a dynamic root directory or an autocell dynamic
|
|
* automount dir.
|
|
*/
|
|
struct inode *afs_iget_pseudo_dir(struct super_block *sb, bool root)
|
|
{
|
|
struct afs_super_info *as = AFS_FS_S(sb);
|
|
struct afs_vnode *vnode;
|
|
struct inode *inode;
|
|
struct afs_fid fid = {};
|
|
|
|
_enter("");
|
|
|
|
if (as->volume)
|
|
fid.vid = as->volume->vid;
|
|
if (root) {
|
|
fid.vnode = 1;
|
|
fid.unique = 1;
|
|
} else {
|
|
fid.vnode = atomic_inc_return(&afs_autocell_ino);
|
|
fid.unique = 0;
|
|
}
|
|
|
|
inode = iget5_locked(sb, fid.vnode,
|
|
afs_iget5_pseudo_test, afs_iget5_pseudo_set, &fid);
|
|
if (!inode) {
|
|
_leave(" = -ENOMEM");
|
|
return ERR_PTR(-ENOMEM);
|
|
}
|
|
|
|
_debug("GOT INODE %p { ino=%lu, vl=%llx, vn=%llx, u=%x }",
|
|
inode, inode->i_ino, fid.vid, fid.vnode, fid.unique);
|
|
|
|
vnode = AFS_FS_I(inode);
|
|
|
|
/* there shouldn't be an existing inode */
|
|
BUG_ON(!(inode->i_state & I_NEW));
|
|
|
|
netfs_inode_init(&vnode->netfs, NULL, false);
|
|
inode->i_size = 0;
|
|
inode->i_mode = S_IFDIR | S_IRUGO | S_IXUGO;
|
|
if (root) {
|
|
inode->i_op = &afs_dynroot_inode_operations;
|
|
inode->i_fop = &simple_dir_operations;
|
|
} else {
|
|
inode->i_op = &afs_autocell_inode_operations;
|
|
}
|
|
set_nlink(inode, 2);
|
|
inode->i_uid = GLOBAL_ROOT_UID;
|
|
inode->i_gid = GLOBAL_ROOT_GID;
|
|
simple_inode_init_ts(inode);
|
|
inode->i_blocks = 0;
|
|
inode->i_generation = 0;
|
|
|
|
set_bit(AFS_VNODE_PSEUDODIR, &vnode->flags);
|
|
if (!root) {
|
|
set_bit(AFS_VNODE_MOUNTPOINT, &vnode->flags);
|
|
inode->i_flags |= S_AUTOMOUNT;
|
|
}
|
|
|
|
inode->i_flags |= S_NOATIME;
|
|
unlock_new_inode(inode);
|
|
_leave(" = %p", inode);
|
|
return inode;
|
|
}
|
|
|
|
/*
|
|
* Probe to see if a cell may exist. This prevents positive dentries from
|
|
* being created unnecessarily.
|
|
*/
|
|
static int afs_probe_cell_name(struct dentry *dentry)
|
|
{
|
|
struct afs_cell *cell;
|
|
struct afs_net *net = afs_d2net(dentry);
|
|
const char *name = dentry->d_name.name;
|
|
size_t len = dentry->d_name.len;
|
|
char *result = NULL;
|
|
int ret;
|
|
|
|
/* Names prefixed with a dot are R/W mounts. */
|
|
if (name[0] == '.') {
|
|
if (len == 1)
|
|
return -EINVAL;
|
|
name++;
|
|
len--;
|
|
}
|
|
|
|
cell = afs_find_cell(net, name, len, afs_cell_trace_use_probe);
|
|
if (!IS_ERR(cell)) {
|
|
afs_unuse_cell(net, cell, afs_cell_trace_unuse_probe);
|
|
return 0;
|
|
}
|
|
|
|
ret = dns_query(net->net, "afsdb", name, len, "srv=1",
|
|
&result, NULL, false);
|
|
if (ret == -ENODATA || ret == -ENOKEY || ret == 0)
|
|
ret = -ENOENT;
|
|
if (ret > 0 && ret >= sizeof(struct dns_server_list_v1_header)) {
|
|
struct dns_server_list_v1_header *v1 = (void *)result;
|
|
|
|
if (v1->hdr.zero == 0 &&
|
|
v1->hdr.content == DNS_PAYLOAD_IS_SERVER_LIST &&
|
|
v1->hdr.version == 1 &&
|
|
(v1->status != DNS_LOOKUP_GOOD &&
|
|
v1->status != DNS_LOOKUP_GOOD_WITH_BAD))
|
|
return -ENOENT;
|
|
|
|
}
|
|
|
|
kfree(result);
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
* Try to auto mount the mountpoint with pseudo directory, if the autocell
|
|
* operation is setted.
|
|
*/
|
|
struct inode *afs_try_auto_mntpt(struct dentry *dentry, struct inode *dir)
|
|
{
|
|
struct afs_vnode *vnode = AFS_FS_I(dir);
|
|
struct inode *inode;
|
|
int ret = -ENOENT;
|
|
|
|
_enter("%p{%pd}, {%llx:%llu}",
|
|
dentry, dentry, vnode->fid.vid, vnode->fid.vnode);
|
|
|
|
if (!test_bit(AFS_VNODE_AUTOCELL, &vnode->flags))
|
|
goto out;
|
|
|
|
ret = afs_probe_cell_name(dentry);
|
|
if (ret < 0)
|
|
goto out;
|
|
|
|
inode = afs_iget_pseudo_dir(dir->i_sb, false);
|
|
if (IS_ERR(inode)) {
|
|
ret = PTR_ERR(inode);
|
|
goto out;
|
|
}
|
|
|
|
_leave("= %p", inode);
|
|
return inode;
|
|
|
|
out:
|
|
_leave("= %d", ret);
|
|
return ret == -ENOENT ? NULL : ERR_PTR(ret);
|
|
}
|
|
|
|
/*
|
|
* Look up @cell in a dynroot directory. This is a substitution for the
|
|
* local cell name for the net namespace.
|
|
*/
|
|
static struct dentry *afs_lookup_atcell(struct dentry *dentry)
|
|
{
|
|
struct afs_cell *cell;
|
|
struct afs_net *net = afs_d2net(dentry);
|
|
struct dentry *ret;
|
|
char *name;
|
|
int len;
|
|
|
|
if (!net->ws_cell)
|
|
return ERR_PTR(-ENOENT);
|
|
|
|
ret = ERR_PTR(-ENOMEM);
|
|
name = kmalloc(AFS_MAXCELLNAME + 1, GFP_KERNEL);
|
|
if (!name)
|
|
goto out_p;
|
|
|
|
down_read(&net->cells_lock);
|
|
cell = net->ws_cell;
|
|
if (cell) {
|
|
len = cell->name_len;
|
|
memcpy(name, cell->name, len + 1);
|
|
}
|
|
up_read(&net->cells_lock);
|
|
|
|
ret = ERR_PTR(-ENOENT);
|
|
if (!cell)
|
|
goto out_n;
|
|
|
|
ret = lookup_one_len(name, dentry->d_parent, len);
|
|
|
|
/* We don't want to d_add() the @cell dentry here as we don't want to
|
|
* the cached dentry to hide changes to the local cell name.
|
|
*/
|
|
|
|
out_n:
|
|
kfree(name);
|
|
out_p:
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
* Look up an entry in a dynroot directory.
|
|
*/
|
|
static struct dentry *afs_dynroot_lookup(struct inode *dir, struct dentry *dentry,
|
|
unsigned int flags)
|
|
{
|
|
_enter("%pd", dentry);
|
|
|
|
ASSERTCMP(d_inode(dentry), ==, NULL);
|
|
|
|
if (flags & LOOKUP_CREATE)
|
|
return ERR_PTR(-EOPNOTSUPP);
|
|
|
|
if (dentry->d_name.len >= AFSNAMEMAX) {
|
|
_leave(" = -ENAMETOOLONG");
|
|
return ERR_PTR(-ENAMETOOLONG);
|
|
}
|
|
|
|
if (dentry->d_name.len == 5 &&
|
|
memcmp(dentry->d_name.name, "@cell", 5) == 0)
|
|
return afs_lookup_atcell(dentry);
|
|
|
|
return d_splice_alias(afs_try_auto_mntpt(dentry, dir), dentry);
|
|
}
|
|
|
|
const struct inode_operations afs_dynroot_inode_operations = {
|
|
.lookup = afs_dynroot_lookup,
|
|
};
|
|
|
|
/*
|
|
* Dirs in the dynamic root don't need revalidation.
|
|
*/
|
|
static int afs_dynroot_d_revalidate(struct dentry *dentry, unsigned int flags)
|
|
{
|
|
return 1;
|
|
}
|
|
|
|
const struct dentry_operations afs_dynroot_dentry_operations = {
|
|
.d_revalidate = afs_dynroot_d_revalidate,
|
|
.d_delete = always_delete_dentry,
|
|
.d_release = afs_d_release,
|
|
.d_automount = afs_d_automount,
|
|
};
|
|
|
|
/*
|
|
* Create a manually added cell mount directory.
|
|
* - The caller must hold net->proc_cells_lock
|
|
*/
|
|
int afs_dynroot_mkdir(struct afs_net *net, struct afs_cell *cell)
|
|
{
|
|
struct super_block *sb = net->dynroot_sb;
|
|
struct dentry *root, *subdir;
|
|
int ret;
|
|
|
|
if (!sb || atomic_read(&sb->s_active) == 0)
|
|
return 0;
|
|
|
|
/* Let the ->lookup op do the creation */
|
|
root = sb->s_root;
|
|
inode_lock(root->d_inode);
|
|
subdir = lookup_one_len(cell->name, root, cell->name_len);
|
|
if (IS_ERR(subdir)) {
|
|
ret = PTR_ERR(subdir);
|
|
goto unlock;
|
|
}
|
|
|
|
/* Note that we're retaining an extra ref on the dentry */
|
|
subdir->d_fsdata = (void *)1UL;
|
|
ret = 0;
|
|
unlock:
|
|
inode_unlock(root->d_inode);
|
|
return ret;
|
|
}
|
|
|
|
/*
|
|
* Remove a manually added cell mount directory.
|
|
* - The caller must hold net->proc_cells_lock
|
|
*/
|
|
void afs_dynroot_rmdir(struct afs_net *net, struct afs_cell *cell)
|
|
{
|
|
struct super_block *sb = net->dynroot_sb;
|
|
struct dentry *root, *subdir;
|
|
|
|
if (!sb || atomic_read(&sb->s_active) == 0)
|
|
return;
|
|
|
|
root = sb->s_root;
|
|
inode_lock(root->d_inode);
|
|
|
|
/* Don't want to trigger a lookup call, which will re-add the cell */
|
|
subdir = try_lookup_one_len(cell->name, root, cell->name_len);
|
|
if (IS_ERR_OR_NULL(subdir)) {
|
|
_debug("lookup %ld", PTR_ERR(subdir));
|
|
goto no_dentry;
|
|
}
|
|
|
|
_debug("rmdir %pd %u", subdir, d_count(subdir));
|
|
|
|
if (subdir->d_fsdata) {
|
|
_debug("unpin %u", d_count(subdir));
|
|
subdir->d_fsdata = NULL;
|
|
dput(subdir);
|
|
}
|
|
dput(subdir);
|
|
no_dentry:
|
|
inode_unlock(root->d_inode);
|
|
_leave("");
|
|
}
|
|
|
|
/*
|
|
* Populate a newly created dynamic root with cell names.
|
|
*/
|
|
int afs_dynroot_populate(struct super_block *sb)
|
|
{
|
|
struct afs_cell *cell;
|
|
struct afs_net *net = afs_sb2net(sb);
|
|
int ret;
|
|
|
|
mutex_lock(&net->proc_cells_lock);
|
|
|
|
net->dynroot_sb = sb;
|
|
hlist_for_each_entry(cell, &net->proc_cells, proc_link) {
|
|
ret = afs_dynroot_mkdir(net, cell);
|
|
if (ret < 0)
|
|
goto error;
|
|
}
|
|
|
|
ret = 0;
|
|
out:
|
|
mutex_unlock(&net->proc_cells_lock);
|
|
return ret;
|
|
|
|
error:
|
|
net->dynroot_sb = NULL;
|
|
goto out;
|
|
}
|
|
|
|
/*
|
|
* When a dynamic root that's in the process of being destroyed, depopulate it
|
|
* of pinned directories.
|
|
*/
|
|
void afs_dynroot_depopulate(struct super_block *sb)
|
|
{
|
|
struct afs_net *net = afs_sb2net(sb);
|
|
struct dentry *root = sb->s_root, *subdir;
|
|
|
|
/* Prevent more subdirs from being created */
|
|
mutex_lock(&net->proc_cells_lock);
|
|
if (net->dynroot_sb == sb)
|
|
net->dynroot_sb = NULL;
|
|
mutex_unlock(&net->proc_cells_lock);
|
|
|
|
if (root) {
|
|
struct hlist_node *n;
|
|
inode_lock(root->d_inode);
|
|
|
|
/* Remove all the pins for dirs created for manually added cells */
|
|
hlist_for_each_entry_safe(subdir, n, &root->d_children, d_sib) {
|
|
if (subdir->d_fsdata) {
|
|
subdir->d_fsdata = NULL;
|
|
dput(subdir);
|
|
}
|
|
}
|
|
|
|
inode_unlock(root->d_inode);
|
|
}
|
|
}
|