linux_dsm_epyc7002/fs/9p/v9fs.c

757 lines
17 KiB
C
Raw Normal View History

/*
* linux/fs/9p/v9fs.c
*
* This file contains functions assisting in mapping VFS to 9P2000
*
* Copyright (C) 2004-2008 by Eric Van Hensbergen <ericvh@gmail.com>
* Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov>
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2
* as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to:
* Free Software Foundation
* 51 Franklin Street, Fifth Floor
* Boston, MA 02111-1301 USA
*
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/fs.h>
#include <linux/sched.h>
#include <linux/cred.h>
#include <linux/parser.h>
#include <linux/idr.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 15:04:11 +07:00
#include <linux/slab.h>
#include <linux/seq_file.h>
#include <net/9p/9p.h>
#include <net/9p/client.h>
#include <net/9p/transport.h>
#include "v9fs.h"
#include "v9fs_vfs.h"
#include "cache.h"
static DEFINE_SPINLOCK(v9fs_sessionlist_lock);
static LIST_HEAD(v9fs_sessionlist);
struct kmem_cache *v9fs_inode_cache;
/*
* Option Parsing (code inspired by NFS code)
* NOTE: each transport will parse its own options
*/
enum {
/* Options that take integer arguments */
Opt_debug, Opt_dfltuid, Opt_dfltgid, Opt_afid,
/* String options */
Opt_uname, Opt_remotename, Opt_cache, Opt_cachetag,
/* Options that take no arguments */
Opt_nodevmap,
/* Cache options */
Opt_cache_loose, Opt_fscache, Opt_mmap,
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
/* Access options */
Opt_access, Opt_posixacl,
/* Lock timeout option */
Opt_locktimeout,
/* Error token */
Opt_err
};
static const match_table_t tokens = {
{Opt_debug, "debug=%x"},
{Opt_dfltuid, "dfltuid=%u"},
{Opt_dfltgid, "dfltgid=%u"},
{Opt_afid, "afid=%u"},
{Opt_uname, "uname=%s"},
{Opt_remotename, "aname=%s"},
{Opt_nodevmap, "nodevmap"},
{Opt_cache, "cache=%s"},
{Opt_cache_loose, "loose"},
{Opt_fscache, "fscache"},
{Opt_mmap, "mmap"},
{Opt_cachetag, "cachetag=%s"},
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
{Opt_access, "access=%s"},
{Opt_posixacl, "posixacl"},
{Opt_locktimeout, "locktimeout=%u"},
{Opt_err, NULL}
};
static const char *const v9fs_cache_modes[nr__p9_cache_modes] = {
[CACHE_NONE] = "none",
[CACHE_MMAP] = "mmap",
[CACHE_LOOSE] = "loose",
[CACHE_FSCACHE] = "fscache",
};
/* Interpret mount options for cache mode */
static int get_cache_mode(char *s)
{
int version = -EINVAL;
if (!strcmp(s, "loose")) {
version = CACHE_LOOSE;
p9_debug(P9_DEBUG_9P, "Cache mode: loose\n");
} else if (!strcmp(s, "fscache")) {
version = CACHE_FSCACHE;
p9_debug(P9_DEBUG_9P, "Cache mode: fscache\n");
} else if (!strcmp(s, "mmap")) {
version = CACHE_MMAP;
p9_debug(P9_DEBUG_9P, "Cache mode: mmap\n");
} else if (!strcmp(s, "none")) {
version = CACHE_NONE;
p9_debug(P9_DEBUG_9P, "Cache mode: none\n");
} else
pr_info("Unknown Cache mode %s\n", s);
return version;
}
/*
* Display the mount options in /proc/mounts.
*/
int v9fs_show_options(struct seq_file *m, struct dentry *root)
{
struct v9fs_session_info *v9ses = root->d_sb->s_fs_info;
if (v9ses->debug)
seq_printf(m, ",debug=%x", v9ses->debug);
if (!uid_eq(v9ses->dfltuid, V9FS_DEFUID))
seq_printf(m, ",dfltuid=%u",
from_kuid_munged(&init_user_ns, v9ses->dfltuid));
if (!gid_eq(v9ses->dfltgid, V9FS_DEFGID))
seq_printf(m, ",dfltgid=%u",
from_kgid_munged(&init_user_ns, v9ses->dfltgid));
if (v9ses->afid != ~0)
seq_printf(m, ",afid=%u", v9ses->afid);
if (strcmp(v9ses->uname, V9FS_DEFUSER) != 0)
seq_printf(m, ",uname=%s", v9ses->uname);
if (strcmp(v9ses->aname, V9FS_DEFANAME) != 0)
seq_printf(m, ",aname=%s", v9ses->aname);
if (v9ses->nodev)
seq_puts(m, ",nodevmap");
if (v9ses->cache)
seq_printf(m, ",%s", v9fs_cache_modes[v9ses->cache]);
#ifdef CONFIG_9P_FSCACHE
if (v9ses->cachetag && v9ses->cache == CACHE_FSCACHE)
seq_printf(m, ",cachetag=%s", v9ses->cachetag);
#endif
switch (v9ses->flags & V9FS_ACCESS_MASK) {
case V9FS_ACCESS_USER:
seq_puts(m, ",access=user");
break;
case V9FS_ACCESS_ANY:
seq_puts(m, ",access=any");
break;
case V9FS_ACCESS_CLIENT:
seq_puts(m, ",access=client");
break;
case V9FS_ACCESS_SINGLE:
seq_printf(m, ",access=%u",
from_kuid_munged(&init_user_ns, v9ses->uid));
break;
}
if (v9ses->flags & V9FS_POSIX_ACL)
seq_puts(m, ",posixacl");
return p9_show_client_options(m, v9ses->clnt);
}
/**
* v9fs_parse_options - parse mount options into session structure
* @v9ses: existing v9fs session information
*
* Return 0 upon success, -ERRNO upon failure.
*/
static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts)
{
char *options, *tmp_options;
substring_t args[MAX_OPT_ARGS];
char *p;
int option = 0;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
char *s, *e;
int ret = 0;
/* setup defaults */
v9ses->afid = ~0;
v9ses->debug = 0;
v9ses->cache = CACHE_NONE;
#ifdef CONFIG_9P_FSCACHE
v9ses->cachetag = NULL;
#endif
v9ses->session_lock_timeout = P9_LOCK_TIMEOUT;
if (!opts)
return 0;
tmp_options = kstrdup(opts, GFP_KERNEL);
if (!tmp_options) {
ret = -ENOMEM;
goto fail_option_alloc;
}
options = tmp_options;
while ((p = strsep(&options, ",")) != NULL) {
int token, r;
if (!*p)
continue;
token = match_token(p, tokens, args);
switch (token) {
case Opt_debug:
r = match_int(&args[0], &option);
if (r < 0) {
p9_debug(P9_DEBUG_ERROR,
"integer field, but no integer?\n");
ret = r;
} else {
v9ses->debug = option;
#ifdef CONFIG_NET_9P_DEBUG
p9_debug_level = option;
#endif
}
break;
case Opt_dfltuid:
r = match_int(&args[0], &option);
if (r < 0) {
p9_debug(P9_DEBUG_ERROR,
"integer field, but no integer?\n");
ret = r;
continue;
}
v9ses->dfltuid = make_kuid(current_user_ns(), option);
if (!uid_valid(v9ses->dfltuid)) {
p9_debug(P9_DEBUG_ERROR,
"uid field, but not a uid?\n");
ret = -EINVAL;
}
break;
case Opt_dfltgid:
r = match_int(&args[0], &option);
if (r < 0) {
p9_debug(P9_DEBUG_ERROR,
"integer field, but no integer?\n");
ret = r;
continue;
}
v9ses->dfltgid = make_kgid(current_user_ns(), option);
if (!gid_valid(v9ses->dfltgid)) {
p9_debug(P9_DEBUG_ERROR,
"gid field, but not a gid?\n");
ret = -EINVAL;
}
break;
case Opt_afid:
r = match_int(&args[0], &option);
if (r < 0) {
p9_debug(P9_DEBUG_ERROR,
"integer field, but no integer?\n");
ret = r;
} else {
v9ses->afid = option;
}
break;
case Opt_uname:
kfree(v9ses->uname);
v9ses->uname = match_strdup(&args[0]);
if (!v9ses->uname) {
ret = -ENOMEM;
goto free_and_return;
}
break;
case Opt_remotename:
kfree(v9ses->aname);
v9ses->aname = match_strdup(&args[0]);
if (!v9ses->aname) {
ret = -ENOMEM;
goto free_and_return;
}
break;
case Opt_nodevmap:
v9ses->nodev = 1;
break;
case Opt_cache_loose:
v9ses->cache = CACHE_LOOSE;
break;
case Opt_fscache:
v9ses->cache = CACHE_FSCACHE;
break;
case Opt_mmap:
v9ses->cache = CACHE_MMAP;
break;
case Opt_cachetag:
#ifdef CONFIG_9P_FSCACHE
kfree(v9ses->cachetag);
v9ses->cachetag = match_strdup(&args[0]);
if (!v9ses->cachetag) {
ret = -ENOMEM;
goto free_and_return;
}
#endif
break;
case Opt_cache:
s = match_strdup(&args[0]);
if (!s) {
ret = -ENOMEM;
p9_debug(P9_DEBUG_ERROR,
"problem allocating copy of cache arg\n");
goto free_and_return;
}
r = get_cache_mode(s);
if (r < 0)
ret = r;
else
v9ses->cache = r;
kfree(s);
break;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
case Opt_access:
s = match_strdup(&args[0]);
if (!s) {
ret = -ENOMEM;
p9_debug(P9_DEBUG_ERROR,
"problem allocating copy of access arg\n");
goto free_and_return;
}
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
v9ses->flags &= ~V9FS_ACCESS_MASK;
if (strcmp(s, "user") == 0)
v9ses->flags |= V9FS_ACCESS_USER;
else if (strcmp(s, "any") == 0)
v9ses->flags |= V9FS_ACCESS_ANY;
else if (strcmp(s, "client") == 0) {
v9ses->flags |= V9FS_ACCESS_CLIENT;
} else {
uid_t uid;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
v9ses->flags |= V9FS_ACCESS_SINGLE;
uid = simple_strtoul(s, &e, 10);
if (*e != '\0') {
ret = -EINVAL;
pr_info("Unknown access argument %s\n",
s);
kfree(s);
continue;
}
v9ses->uid = make_kuid(current_user_ns(), uid);
if (!uid_valid(v9ses->uid)) {
ret = -EINVAL;
pr_info("Unknown uid %s\n", s);
}
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
}
kfree(s);
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
break;
case Opt_posixacl:
#ifdef CONFIG_9P_FS_POSIX_ACL
v9ses->flags |= V9FS_POSIX_ACL;
#else
p9_debug(P9_DEBUG_ERROR,
"Not defined CONFIG_9P_FS_POSIX_ACL. Ignoring posixacl option\n");
#endif
break;
case Opt_locktimeout:
r = match_int(&args[0], &option);
if (r < 0) {
p9_debug(P9_DEBUG_ERROR,
"integer field, but no integer?\n");
ret = r;
continue;
}
if (option < 1) {
p9_debug(P9_DEBUG_ERROR,
"locktimeout must be a greater than zero integer.\n");
ret = -EINVAL;
continue;
}
v9ses->session_lock_timeout = (long)option * HZ;
break;
default:
continue;
}
}
free_and_return:
kfree(tmp_options);
fail_option_alloc:
return ret;
}
/**
* v9fs_session_init - initialize session
* @v9ses: session information structure
* @dev_name: device being mounted
* @data: options
*
*/
struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses,
const char *dev_name, char *data)
{
struct p9_fid *fid;
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
int rc = -ENOMEM;
v9ses->uname = kstrdup(V9FS_DEFUSER, GFP_KERNEL);
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
if (!v9ses->uname)
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
goto err_names;
v9ses->aname = kstrdup(V9FS_DEFANAME, GFP_KERNEL);
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
if (!v9ses->aname)
goto err_names;
init_rwsem(&v9ses->rename_sem);
v9ses->uid = INVALID_UID;
v9ses->dfltuid = V9FS_DEFUID;
v9ses->dfltgid = V9FS_DEFGID;
v9ses->clnt = p9_client_create(dev_name, data);
if (IS_ERR(v9ses->clnt)) {
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
rc = PTR_ERR(v9ses->clnt);
p9_debug(P9_DEBUG_ERROR, "problem initializing 9p client\n");
goto err_names;
}
v9ses->flags = V9FS_ACCESS_USER;
if (p9_is_proto_dotl(v9ses->clnt)) {
v9ses->flags = V9FS_ACCESS_CLIENT;
v9ses->flags |= V9FS_PROTO_2000L;
} else if (p9_is_proto_dotu(v9ses->clnt)) {
v9ses->flags |= V9FS_PROTO_2000U;
}
rc = v9fs_parse_options(v9ses, data);
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
if (rc < 0)
goto err_clnt;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ;
if (!v9fs_proto_dotl(v9ses) &&
((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)) {
/*
* We support ACCESS_CLIENT only for dotl.
* Fall back to ACCESS_USER
*/
v9ses->flags &= ~V9FS_ACCESS_MASK;
v9ses->flags |= V9FS_ACCESS_USER;
}
/*FIXME !! */
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
/* for legacy mode, fall back to V9FS_ACCESS_ANY */
if (!(v9fs_proto_dotu(v9ses) || v9fs_proto_dotl(v9ses)) &&
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
((v9ses->flags&V9FS_ACCESS_MASK) == V9FS_ACCESS_USER)) {
v9ses->flags &= ~V9FS_ACCESS_MASK;
v9ses->flags |= V9FS_ACCESS_ANY;
v9ses->uid = INVALID_UID;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
}
if (!v9fs_proto_dotl(v9ses) ||
!((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)) {
/*
* We support ACL checks on clinet only if the protocol is
* 9P2000.L and access is V9FS_ACCESS_CLIENT.
*/
v9ses->flags &= ~V9FS_ACL_MASK;
}
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
fid = p9_client_attach(v9ses->clnt, NULL, v9ses->uname, INVALID_UID,
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
v9ses->aname);
if (IS_ERR(fid)) {
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
rc = PTR_ERR(fid);
p9_debug(P9_DEBUG_ERROR, "cannot attach\n");
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
goto err_clnt;
}
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
if ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_SINGLE)
fid->uid = v9ses->uid;
else
fid->uid = INVALID_UID;
9p: attach-per-user The 9P2000 protocol requires the authentication and permission checks to be done in the file server. For that reason every user that accesses the file server tree has to authenticate and attach to the server separately. Multiple users can share the same connection to the server. Currently v9fs does a single attach and executes all I/O operations as a single user. This makes using v9fs in multiuser environment unsafe as it depends on the client doing the permission checking. This patch improves the 9P2000 support by allowing every user to attach separately. The patch defines three modes of access (new mount option 'access'): - attach-per-user (access=user) (default mode for 9P2000.u) If a user tries to access a file served by v9fs for the first time, v9fs sends an attach command to the server (Tattach) specifying the user. If the attach succeeds, the user can access the v9fs tree. As there is no uname->uid (string->integer) mapping yet, this mode works only with the 9P2000.u dialect. - allow only one user to access the tree (access=<uid>) Only the user with uid can access the v9fs tree. Other users that attempt to access it will get EPERM error. - do all operations as a single user (access=any) (default for 9P2000) V9fs does a single attach and all operations are done as a single user. If this mode is selected, the v9fs behavior is identical with the current one. Signed-off-by: Latchesar Ionkov <lucho@ionkov.net> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2007-10-18 02:31:07 +07:00
#ifdef CONFIG_9P_FSCACHE
/* register the session for caching */
v9fs_cache_session_get_cookie(v9ses);
#endif
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
spin_lock(&v9fs_sessionlist_lock);
list_add(&v9ses->slist, &v9fs_sessionlist);
spin_unlock(&v9fs_sessionlist_lock);
return fid;
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
err_clnt:
#ifdef CONFIG_9P_FSCACHE
kfree(v9ses->cachetag);
#endif
v9fs: fix error handling in v9fs_session_init() On failure, v9fs_session_init() returns with the v9fs_session_info struct partially initialized and expects the caller to invoke v9fs_session_close() to clean it up; however, it doesn't track whether the bdi is initialized or not and curiously invokes bdi_destroy() in both vfs_session_init() failure path too. A. If v9fs_session_init() fails before the bdi is initialized, the follow-up v9fs_session_close() will invoke bdi_destroy() on an uninitialized bdi. B. If v9fs_session_init() fails after the bdi is initialized, bdi_destroy() will be called twice on the same bdi - once in the failure path of v9fs_session_init() and then by v9fs_session_close(). A is broken no matter what. B used to be okay because bdi_destroy() allowed being invoked multiple times on the same bdi, which BTW was broken in its own way - if bdi_destroy() was invoked on an initialiezd but !registered bdi, it'd fail to free percpu counters. Since f0054bb1e1f3 ("writeback: move backing_dev_info->wb_lock and ->worklist into bdi_writeback"), this no longer work - bdi_destroy() on an initialized but not registered bdi works correctly but multiple invocations of bdi_destroy() is no longer allowed. The obvious culprit here is v9fs_session_init()'s odd and broken error behavior. It should simply clean up after itself on failures. This patch makes the following updates to v9fs_session_init(). * @rc -> @retval error return propagation removed. It didn't serve any purpose. Just use @rc. * Move addition to v9fs_sessionlist to the end of the function so that incomplete sessions are not put on the list or iterated and error path doesn't have to worry about it. * Update error handling so that it cleans up after itself. Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Sasha Levin <sasha.levin@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-06-08 12:57:31 +07:00
p9_client_destroy(v9ses->clnt);
err_names:
kfree(v9ses->uname);
kfree(v9ses->aname);
return ERR_PTR(rc);
}
/**
* v9fs_session_close - shutdown a session
* @v9ses: session information structure
*
*/
void v9fs_session_close(struct v9fs_session_info *v9ses)
{
if (v9ses->clnt) {
p9_client_destroy(v9ses->clnt);
v9ses->clnt = NULL;
}
#ifdef CONFIG_9P_FSCACHE
if (v9ses->fscache) {
v9fs_cache_session_put_cookie(v9ses);
kfree(v9ses->cachetag);
}
#endif
kfree(v9ses->uname);
kfree(v9ses->aname);
spin_lock(&v9fs_sessionlist_lock);
list_del(&v9ses->slist);
spin_unlock(&v9fs_sessionlist_lock);
}
/**
* v9fs_session_cancel - terminate a session
* @v9ses: session to terminate
*
* mark transport as disconnected and cancel all pending requests.
*/
void v9fs_session_cancel(struct v9fs_session_info *v9ses) {
p9_debug(P9_DEBUG_ERROR, "cancel session %p\n", v9ses);
p9_client_disconnect(v9ses->clnt);
}
/**
* v9fs_session_begin_cancel - Begin terminate of a session
* @v9ses: session to terminate
*
* After this call we don't allow any request other than clunk.
*/
void v9fs_session_begin_cancel(struct v9fs_session_info *v9ses)
{
p9_debug(P9_DEBUG_ERROR, "begin cancel session %p\n", v9ses);
p9_client_begin_disconnect(v9ses->clnt);
}
extern int v9fs_error_init(void);
static struct kobject *v9fs_kobj;
#ifdef CONFIG_9P_FSCACHE
/**
* caches_show - list caches associated with a session
*
* Returns the size of buffer written.
*/
static ssize_t caches_show(struct kobject *kobj,
struct kobj_attribute *attr,
char *buf)
{
ssize_t n = 0, count = 0, limit = PAGE_SIZE;
struct v9fs_session_info *v9ses;
spin_lock(&v9fs_sessionlist_lock);
list_for_each_entry(v9ses, &v9fs_sessionlist, slist) {
if (v9ses->cachetag) {
n = snprintf(buf, limit, "%s\n", v9ses->cachetag);
if (n < 0) {
count = n;
break;
}
count += n;
limit -= n;
}
}
spin_unlock(&v9fs_sessionlist_lock);
return count;
}
static struct kobj_attribute v9fs_attr_cache = __ATTR_RO(caches);
#endif /* CONFIG_9P_FSCACHE */
static struct attribute *v9fs_attrs[] = {
#ifdef CONFIG_9P_FSCACHE
&v9fs_attr_cache.attr,
#endif
NULL,
};
static struct attribute_group v9fs_attr_group = {
.attrs = v9fs_attrs,
};
/**
* v9fs_sysfs_init - Initialize the v9fs sysfs interface
*
*/
static int __init v9fs_sysfs_init(void)
{
v9fs_kobj = kobject_create_and_add("9p", fs_kobj);
if (!v9fs_kobj)
return -ENOMEM;
if (sysfs_create_group(v9fs_kobj, &v9fs_attr_group)) {
kobject_put(v9fs_kobj);
return -ENOMEM;
}
return 0;
}
/**
* v9fs_sysfs_cleanup - Unregister the v9fs sysfs interface
*
*/
static void v9fs_sysfs_cleanup(void)
{
sysfs_remove_group(v9fs_kobj, &v9fs_attr_group);
kobject_put(v9fs_kobj);
}
static void v9fs_inode_init_once(void *foo)
{
struct v9fs_inode *v9inode = (struct v9fs_inode *)foo;
#ifdef CONFIG_9P_FSCACHE
v9inode->fscache = NULL;
#endif
memset(&v9inode->qid, 0, sizeof(v9inode->qid));
inode_init_once(&v9inode->vfs_inode);
}
/**
* v9fs_init_inode_cache - initialize a cache for 9P
* Returns 0 on success.
*/
static int v9fs_init_inode_cache(void)
{
v9fs_inode_cache = kmem_cache_create("v9fs_inode_cache",
sizeof(struct v9fs_inode),
0, (SLAB_RECLAIM_ACCOUNT|
2016-01-15 06:18:21 +07:00
SLAB_MEM_SPREAD|SLAB_ACCOUNT),
v9fs_inode_init_once);
if (!v9fs_inode_cache)
return -ENOMEM;
return 0;
}
/**
* v9fs_destroy_inode_cache - destroy the cache of 9P inode
*
*/
static void v9fs_destroy_inode_cache(void)
{
/*
* Make sure all delayed rcu free inodes are flushed before we
* destroy cache.
*/
rcu_barrier();
kmem_cache_destroy(v9fs_inode_cache);
}
static int v9fs_cache_register(void)
{
int ret;
ret = v9fs_init_inode_cache();
if (ret < 0)
return ret;
#ifdef CONFIG_9P_FSCACHE
ret = fscache_register_netfs(&v9fs_cache_netfs);
if (ret < 0)
v9fs_destroy_inode_cache();
#endif
return ret;
}
static void v9fs_cache_unregister(void)
{
v9fs_destroy_inode_cache();
#ifdef CONFIG_9P_FSCACHE
fscache_unregister_netfs(&v9fs_cache_netfs);
#endif
}
/**
* init_v9fs - Initialize module
*
*/
static int __init init_v9fs(void)
{
int err;
pr_info("Installing v9fs 9p2000 file system support\n");
/* TODO: Setup list of registered trasnport modules */
err = v9fs_cache_register();
if (err < 0) {
pr_err("Failed to register v9fs for caching\n");
return err;
}
err = v9fs_sysfs_init();
if (err < 0) {
pr_err("Failed to register with sysfs\n");
goto out_cache;
}
err = register_filesystem(&v9fs_fs_type);
if (err < 0) {
pr_err("Failed to register filesystem\n");
goto out_sysfs_cleanup;
}
return 0;
out_sysfs_cleanup:
v9fs_sysfs_cleanup();
out_cache:
v9fs_cache_unregister();
return err;
}
/**
* exit_v9fs - shutdown module
*
*/
static void __exit exit_v9fs(void)
{
v9fs_sysfs_cleanup();
v9fs_cache_unregister();
unregister_filesystem(&v9fs_fs_type);
}
module_init(init_v9fs)
module_exit(exit_v9fs)
MODULE_AUTHOR("Latchesar Ionkov <lucho@ionkov.net>");
MODULE_AUTHOR("Eric Van Hensbergen <ericvh@gmail.com>");
MODULE_AUTHOR("Ron Minnich <rminnich@lanl.gov>");
MODULE_LICENSE("GPL");