linux/fs/nfsd/export.c

1290 lines
32 KiB
C
Raw Normal View History

/*
* NFS exporting and validation.
*
* We maintain a list of clients, each of which has a list of
* exports. To export an fs to a given client, you first have
* to create the client entry with NFSCTL_ADDCLIENT, which
* creates a client control block and adds it to the hash
* table. Then, you call NFSCTL_EXPORT for each fs.
*
*
* Copyright (C) 1995, 1996 Olaf Kirch, <okir@monad.swb.de>
*/
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 11:04:11 +03:00
#include <linux/slab.h>
#include <linux/namei.h>
#include <linux/module.h>
#include <linux/exportfs.h>
#include <linux/sunrpc/svc_xprt.h>
#include "nfsd.h"
#include "nfsfh.h"
#include "netns.h"
nfsd: implement pNFS operations Add support for the GETDEVICEINFO, LAYOUTGET, LAYOUTCOMMIT and LAYOUTRETURN NFSv4.1 operations, as well as backing code to manage outstanding layouts and devices. Layout management is very straight forward, with a nfs4_layout_stateid structure that extends nfs4_stid to manage layout stateids as the top-level structure. It is linked into the nfs4_file and nfs4_client structures like the other stateids, and contains a linked list of layouts that hang of the stateid. The actual layout operations are implemented in layout drivers that are not part of this commit, but will be added later. The worst part of this commit is the management of the pNFS device IDs, which suffers from a specification that is not sanely implementable due to the fact that the device-IDs are global and not bound to an export, and have a small enough size so that we can't store the fsid portion of a file handle, and must never be reused. As we still do need perform all export authentication and validation checks on a device ID passed to GETDEVICEINFO we are caught between a rock and a hard place. To work around this issue we add a new hash that maps from a 64-bit integer to a fsid so that we can look up the export to authenticate against it, a 32-bit integer as a generation that we can bump when changing the device, and a currently unused 32-bit integer that could be used in the future to handle more than a single device per export. Entries in this hash table are never deleted as we can't reuse the ids anyway, and would have a severe lifetime problem anyway as Linux export structures are temporary structures that can go away under load. Parts of the XDR data, structures and marshaling/unmarshaling code, as well as many concepts are derived from the old pNFS server implementation from Andy Adamson, Benny Halevy, Dean Hildebrand, Marc Eshel, Fred Isaman, Mike Sager, Ricardo Labiaga and many others. Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-05-05 15:11:59 +04:00
#include "pnfs.h"
#define NFSDDBG_FACILITY NFSDDBG_EXPORT
/*
* We have two caches.
* One maps client+vfsmnt+dentry to export options - the export map
* The other maps client+filehandle-fragment to export options. - the expkey map
*
* The export options are actually stored in the first map, and the
* second map contains a reference to the entry in the first map.
*/
#define EXPKEY_HASHBITS 8
#define EXPKEY_HASHMAX (1 << EXPKEY_HASHBITS)
#define EXPKEY_HASHMASK (EXPKEY_HASHMAX -1)
static void expkey_put(struct kref *ref)
{
struct svc_expkey *key = container_of(ref, struct svc_expkey, h.ref);
if (test_bit(CACHE_VALID, &key->h.flags) &&
!test_bit(CACHE_NEGATIVE, &key->h.flags))
path_put(&key->ek_path);
auth_domain_put(key->ek_client);
kfree(key);
}
static void expkey_request(struct cache_detail *cd,
struct cache_head *h,
char **bpp, int *blen)
{
/* client fsidtype \xfsid */
struct svc_expkey *ek = container_of(h, struct svc_expkey, h);
char type[5];
qword_add(bpp, blen, ek->ek_client->name);
snprintf(type, 5, "%d", ek->ek_fsidtype);
qword_add(bpp, blen, type);
qword_addhex(bpp, blen, (char*)ek->ek_fsid, key_len(ek->ek_fsidtype));
(*bpp)[-1] = '\n';
}
static struct svc_expkey *svc_expkey_update(struct cache_detail *cd, struct svc_expkey *new,
struct svc_expkey *old);
static struct svc_expkey *svc_expkey_lookup(struct cache_detail *cd, struct svc_expkey *);
static int expkey_parse(struct cache_detail *cd, char *mesg, int mlen)
{
/* client fsidtype fsid expiry [path] */
char *buf;
int len;
struct auth_domain *dom = NULL;
int err;
int fsidtype;
char *ep;
struct svc_expkey key;
struct svc_expkey *ek = NULL;
if (mesg[mlen - 1] != '\n')
return -EINVAL;
mesg[mlen-1] = 0;
buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
err = -ENOMEM;
if (!buf)
goto out;
err = -EINVAL;
if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
goto out;
err = -ENOENT;
dom = auth_domain_find(buf);
if (!dom)
goto out;
dprintk("found domain %s\n", buf);
err = -EINVAL;
if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
goto out;
fsidtype = simple_strtoul(buf, &ep, 10);
if (*ep)
goto out;
dprintk("found fsidtype %d\n", fsidtype);
if (key_len(fsidtype)==0) /* invalid type */
goto out;
if ((len=qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
goto out;
dprintk("found fsid length %d\n", len);
if (len != key_len(fsidtype))
goto out;
/* OK, we seem to have a valid key */
key.h.flags = 0;
key.h.expiry_time = get_expiry(&mesg);
if (key.h.expiry_time == 0)
goto out;
key.ek_client = dom;
key.ek_fsidtype = fsidtype;
memcpy(key.ek_fsid, buf, len);
ek = svc_expkey_lookup(cd, &key);
err = -ENOMEM;
if (!ek)
goto out;
/* now we want a pathname, or empty meaning NEGATIVE */
err = -EINVAL;
len = qword_get(&mesg, buf, PAGE_SIZE);
if (len < 0)
goto out;
dprintk("Path seems to be <%s>\n", buf);
err = 0;
if (len == 0) {
set_bit(CACHE_NEGATIVE, &key.h.flags);
ek = svc_expkey_update(cd, &key, ek);
if (!ek)
err = -ENOMEM;
} else {
err = kern_path(buf, 0, &key.ek_path);
if (err)
goto out;
dprintk("Found the path %s\n", buf);
ek = svc_expkey_update(cd, &key, ek);
if (!ek)
err = -ENOMEM;
path_put(&key.ek_path);
}
cache_flush();
out:
if (ek)
cache_put(&ek->h, cd);
if (dom)
auth_domain_put(dom);
kfree(buf);
return err;
}
static int expkey_show(struct seq_file *m,
struct cache_detail *cd,
struct cache_head *h)
{
struct svc_expkey *ek ;
int i;
if (h ==NULL) {
seq_puts(m, "#domain fsidtype fsid [path]\n");
return 0;
}
ek = container_of(h, struct svc_expkey, h);
seq_printf(m, "%s %d 0x", ek->ek_client->name,
ek->ek_fsidtype);
for (i=0; i < key_len(ek->ek_fsidtype)/4; i++)
seq_printf(m, "%08x", ek->ek_fsid[i]);
if (test_bit(CACHE_VALID, &h->flags) &&
!test_bit(CACHE_NEGATIVE, &h->flags)) {
seq_printf(m, " ");
seq_path(m, &ek->ek_path, "\\ \t\n");
}
seq_printf(m, "\n");
return 0;
}
static inline int expkey_match (struct cache_head *a, struct cache_head *b)
{
struct svc_expkey *orig = container_of(a, struct svc_expkey, h);
struct svc_expkey *new = container_of(b, struct svc_expkey, h);
if (orig->ek_fsidtype != new->ek_fsidtype ||
orig->ek_client != new->ek_client ||
memcmp(orig->ek_fsid, new->ek_fsid, key_len(orig->ek_fsidtype)) != 0)
return 0;
return 1;
}
static inline void expkey_init(struct cache_head *cnew,
struct cache_head *citem)
{
struct svc_expkey *new = container_of(cnew, struct svc_expkey, h);
struct svc_expkey *item = container_of(citem, struct svc_expkey, h);
kref_get(&item->ek_client->ref);
new->ek_client = item->ek_client;
new->ek_fsidtype = item->ek_fsidtype;
memcpy(new->ek_fsid, item->ek_fsid, sizeof(new->ek_fsid));
}
static inline void expkey_update(struct cache_head *cnew,
struct cache_head *citem)
{
struct svc_expkey *new = container_of(cnew, struct svc_expkey, h);
struct svc_expkey *item = container_of(citem, struct svc_expkey, h);
new->ek_path = item->ek_path;
path_get(&item->ek_path);
}
static struct cache_head *expkey_alloc(void)
{
struct svc_expkey *i = kmalloc(sizeof(*i), GFP_KERNEL);
if (i)
return &i->h;
else
return NULL;
}
static struct cache_detail svc_expkey_cache_template = {
.owner = THIS_MODULE,
.hash_size = EXPKEY_HASHMAX,
.name = "nfsd.fh",
.cache_put = expkey_put,
.cache_request = expkey_request,
.cache_parse = expkey_parse,
.cache_show = expkey_show,
.match = expkey_match,
.init = expkey_init,
.update = expkey_update,
.alloc = expkey_alloc,
};
static int
svc_expkey_hash(struct svc_expkey *item)
{
int hash = item->ek_fsidtype;
char * cp = (char*)item->ek_fsid;
int len = key_len(item->ek_fsidtype);
hash ^= hash_mem(cp, len, EXPKEY_HASHBITS);
hash ^= hash_ptr(item->ek_client, EXPKEY_HASHBITS);
hash &= EXPKEY_HASHMASK;
return hash;
}
static struct svc_expkey *
svc_expkey_lookup(struct cache_detail *cd, struct svc_expkey *item)
{
struct cache_head *ch;
int hash = svc_expkey_hash(item);
ch = sunrpc_cache_lookup(cd, &item->h, hash);
if (ch)
return container_of(ch, struct svc_expkey, h);
else
return NULL;
}
static struct svc_expkey *
svc_expkey_update(struct cache_detail *cd, struct svc_expkey *new,
struct svc_expkey *old)
{
struct cache_head *ch;
int hash = svc_expkey_hash(new);
ch = sunrpc_cache_update(cd, &new->h, &old->h, hash);
if (ch)
return container_of(ch, struct svc_expkey, h);
else
return NULL;
}
#define EXPORT_HASHBITS 8
#define EXPORT_HASHMAX (1<< EXPORT_HASHBITS)
static void nfsd4_fslocs_free(struct nfsd4_fs_locations *fsloc)
{
struct nfsd4_fs_location *locations = fsloc->locations;
int i;
if (!locations)
return;
for (i = 0; i < fsloc->locations_count; i++) {
kfree(locations[i].path);
kfree(locations[i].hosts);
}
kfree(locations);
fsloc->locations = NULL;
}
static void svc_export_put(struct kref *ref)
{
struct svc_export *exp = container_of(ref, struct svc_export, h.ref);
path_put(&exp->ex_path);
auth_domain_put(exp->ex_client);
nfsd4_fslocs_free(&exp->ex_fslocs);
kfree(exp->ex_uuid);
kfree(exp);
}
static void svc_export_request(struct cache_detail *cd,
struct cache_head *h,
char **bpp, int *blen)
{
/* client path */
struct svc_export *exp = container_of(h, struct svc_export, h);
char *pth;
qword_add(bpp, blen, exp->ex_client->name);
pth = d_path(&exp->ex_path, *bpp, *blen);
if (IS_ERR(pth)) {
/* is this correct? */
(*bpp)[0] = '\n';
return;
}
qword_add(bpp, blen, pth);
(*bpp)[-1] = '\n';
}
static struct svc_export *svc_export_update(struct svc_export *new,
struct svc_export *old);
static struct svc_export *svc_export_lookup(struct svc_export *);
static int check_export(struct inode *inode, int *flags, unsigned char *uuid)
{
/*
* We currently export only dirs, regular files, and (for v4
* pseudoroot) symlinks.
*/
if (!S_ISDIR(inode->i_mode) &&
!S_ISLNK(inode->i_mode) &&
!S_ISREG(inode->i_mode))
return -ENOTDIR;
/*
* Mountd should never pass down a writeable V4ROOT export, but,
* just to make sure:
*/
if (*flags & NFSEXP_V4ROOT)
*flags |= NFSEXP_READONLY;
/* There are two requirements on a filesystem to be exportable.
* 1: We must be able to identify the filesystem from a number.
* either a device number (so FS_REQUIRES_DEV needed)
* or an FSID number (so NFSEXP_FSID or ->uuid is needed).
* 2: We must be able to find an inode from a filehandle.
* This means that s_export_op must be set.
*/
if (!(inode->i_sb->s_type->fs_flags & FS_REQUIRES_DEV) &&
!(*flags & NFSEXP_FSID) &&
uuid == NULL) {
dprintk("exp_export: export of non-dev fs without fsid\n");
return -EINVAL;
}
if (!inode->i_sb->s_export_op ||
!inode->i_sb->s_export_op->fh_to_dentry) {
dprintk("exp_export: export of invalid fs type.\n");
return -EINVAL;
}
return 0;
}
#ifdef CONFIG_NFSD_V4
static int
fsloc_parse(char **mesg, char *buf, struct nfsd4_fs_locations *fsloc)
{
int len;
int migrated, i, err;
/* more than one fsloc */
if (fsloc->locations)
return -EINVAL;
/* listsize */
err = get_uint(mesg, &fsloc->locations_count);
if (err)
return err;
if (fsloc->locations_count > MAX_FS_LOCATIONS)
return -EINVAL;
if (fsloc->locations_count == 0)
return 0;
fsloc->locations = kzalloc(fsloc->locations_count
* sizeof(struct nfsd4_fs_location), GFP_KERNEL);
if (!fsloc->locations)
return -ENOMEM;
for (i=0; i < fsloc->locations_count; i++) {
/* colon separated host list */
err = -EINVAL;
len = qword_get(mesg, buf, PAGE_SIZE);
if (len <= 0)
goto out_free_all;
err = -ENOMEM;
fsloc->locations[i].hosts = kstrdup(buf, GFP_KERNEL);
if (!fsloc->locations[i].hosts)
goto out_free_all;
err = -EINVAL;
/* slash separated path component list */
len = qword_get(mesg, buf, PAGE_SIZE);
if (len <= 0)
goto out_free_all;
err = -ENOMEM;
fsloc->locations[i].path = kstrdup(buf, GFP_KERNEL);
if (!fsloc->locations[i].path)
goto out_free_all;
}
/* migrated */
err = get_int(mesg, &migrated);
if (err)
goto out_free_all;
err = -EINVAL;
if (migrated < 0 || migrated > 1)
goto out_free_all;
fsloc->migrated = migrated;
return 0;
out_free_all:
nfsd4_fslocs_free(fsloc);
return err;
}
static int secinfo_parse(char **mesg, char *buf, struct svc_export *exp)
{
struct exp_flavor_info *f;
u32 listsize;
int err;
/* more than one secinfo */
if (exp->ex_nflavors)
return -EINVAL;
err = get_uint(mesg, &listsize);
if (err)
return err;
if (listsize > MAX_SECINFO_LIST)
return -EINVAL;
for (f = exp->ex_flavors; f < exp->ex_flavors + listsize; f++) {
err = get_uint(mesg, &f->pseudoflavor);
if (err)
return err;
/*
* XXX: It would be nice to also check whether this
* pseudoflavor is supported, so we can discover the
* problem at export time instead of when a client fails
* to authenticate.
*/
err = get_uint(mesg, &f->flags);
if (err)
return err;
/* Only some flags are allowed to differ between flavors: */
if (~NFSEXP_SECINFO_FLAGS & (f->flags ^ exp->ex_flags))
return -EINVAL;
}
exp->ex_nflavors = listsize;
return 0;
}
#else /* CONFIG_NFSD_V4 */
static inline int
fsloc_parse(char **mesg, char *buf, struct nfsd4_fs_locations *fsloc){return 0;}
static inline int
secinfo_parse(char **mesg, char *buf, struct svc_export *exp) { return 0; }
#endif
static inline int
nfsd_uuid_parse(char **mesg, char *buf, unsigned char **puuid)
{
int len;
/* more than one uuid */
if (*puuid)
return -EINVAL;
/* expect a 16 byte uuid encoded as \xXXXX... */
len = qword_get(mesg, buf, PAGE_SIZE);
if (len != EX_UUID_LEN)
return -EINVAL;
*puuid = kmemdup(buf, EX_UUID_LEN, GFP_KERNEL);
if (*puuid == NULL)
return -ENOMEM;
return 0;
}
static int svc_export_parse(struct cache_detail *cd, char *mesg, int mlen)
{
/* client path expiry [flags anonuid anongid fsid] */
char *buf;
int len;
int err;
struct auth_domain *dom = NULL;
struct svc_export exp = {}, *expp;
int an_int;
if (mesg[mlen-1] != '\n')
return -EINVAL;
mesg[mlen-1] = 0;
buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
if (!buf)
return -ENOMEM;
/* client */
err = -EINVAL;
len = qword_get(&mesg, buf, PAGE_SIZE);
if (len <= 0)
goto out;
err = -ENOENT;
dom = auth_domain_find(buf);
if (!dom)
goto out;
/* path */
err = -EINVAL;
if ((len = qword_get(&mesg, buf, PAGE_SIZE)) <= 0)
goto out1;
err = kern_path(buf, 0, &exp.ex_path);
if (err)
goto out1;
exp.ex_client = dom;
exp.cd = cd;
nfsd: implement pNFS operations Add support for the GETDEVICEINFO, LAYOUTGET, LAYOUTCOMMIT and LAYOUTRETURN NFSv4.1 operations, as well as backing code to manage outstanding layouts and devices. Layout management is very straight forward, with a nfs4_layout_stateid structure that extends nfs4_stid to manage layout stateids as the top-level structure. It is linked into the nfs4_file and nfs4_client structures like the other stateids, and contains a linked list of layouts that hang of the stateid. The actual layout operations are implemented in layout drivers that are not part of this commit, but will be added later. The worst part of this commit is the management of the pNFS device IDs, which suffers from a specification that is not sanely implementable due to the fact that the device-IDs are global and not bound to an export, and have a small enough size so that we can't store the fsid portion of a file handle, and must never be reused. As we still do need perform all export authentication and validation checks on a device ID passed to GETDEVICEINFO we are caught between a rock and a hard place. To work around this issue we add a new hash that maps from a 64-bit integer to a fsid so that we can look up the export to authenticate against it, a 32-bit integer as a generation that we can bump when changing the device, and a currently unused 32-bit integer that could be used in the future to handle more than a single device per export. Entries in this hash table are never deleted as we can't reuse the ids anyway, and would have a severe lifetime problem anyway as Linux export structures are temporary structures that can go away under load. Parts of the XDR data, structures and marshaling/unmarshaling code, as well as many concepts are derived from the old pNFS server implementation from Andy Adamson, Benny Halevy, Dean Hildebrand, Marc Eshel, Fred Isaman, Mike Sager, Ricardo Labiaga and many others. Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-05-05 15:11:59 +04:00
exp.ex_devid_map = NULL;
/* expiry */
err = -EINVAL;
exp.h.expiry_time = get_expiry(&mesg);
if (exp.h.expiry_time == 0)
goto out3;
/* flags */
err = get_int(&mesg, &an_int);
if (err == -ENOENT) {
err = 0;
set_bit(CACHE_NEGATIVE, &exp.h.flags);
} else {
if (err || an_int < 0)
goto out3;
exp.ex_flags= an_int;
/* anon uid */
err = get_int(&mesg, &an_int);
if (err)
goto out3;
exp.ex_anon_uid= make_kuid(&init_user_ns, an_int);
/* anon gid */
err = get_int(&mesg, &an_int);
if (err)
goto out3;
exp.ex_anon_gid= make_kgid(&init_user_ns, an_int);
/* fsid */
err = get_int(&mesg, &an_int);
if (err)
goto out3;
exp.ex_fsid = an_int;
while ((len = qword_get(&mesg, buf, PAGE_SIZE)) > 0) {
if (strcmp(buf, "fsloc") == 0)
err = fsloc_parse(&mesg, buf, &exp.ex_fslocs);
else if (strcmp(buf, "uuid") == 0)
err = nfsd_uuid_parse(&mesg, buf, &exp.ex_uuid);
else if (strcmp(buf, "secinfo") == 0)
err = secinfo_parse(&mesg, buf, &exp);
else
/* quietly ignore unknown words and anything
* following. Newer user-space can try to set
* new values, then see what the result was.
*/
break;
if (err)
goto out4;
}
err = check_export(d_inode(exp.ex_path.dentry), &exp.ex_flags,
exp.ex_uuid);
if (err)
goto out4;
/*
* No point caching this if it would immediately expire.
* Also, this protects exportfs's dummy export from the
* anon_uid/anon_gid checks:
*/
if (exp.h.expiry_time < seconds_since_boot())
goto out4;
/*
* For some reason exportfs has been passing down an
* invalid (-1) uid & gid on the "dummy" export which it
* uses to test export support. To make sure exportfs
* sees errors from check_export we therefore need to
* delay these checks till after check_export:
*/
err = -EINVAL;
if (!uid_valid(exp.ex_anon_uid))
goto out4;
if (!gid_valid(exp.ex_anon_gid))
goto out4;
err = 0;
nfsd: implement pNFS operations Add support for the GETDEVICEINFO, LAYOUTGET, LAYOUTCOMMIT and LAYOUTRETURN NFSv4.1 operations, as well as backing code to manage outstanding layouts and devices. Layout management is very straight forward, with a nfs4_layout_stateid structure that extends nfs4_stid to manage layout stateids as the top-level structure. It is linked into the nfs4_file and nfs4_client structures like the other stateids, and contains a linked list of layouts that hang of the stateid. The actual layout operations are implemented in layout drivers that are not part of this commit, but will be added later. The worst part of this commit is the management of the pNFS device IDs, which suffers from a specification that is not sanely implementable due to the fact that the device-IDs are global and not bound to an export, and have a small enough size so that we can't store the fsid portion of a file handle, and must never be reused. As we still do need perform all export authentication and validation checks on a device ID passed to GETDEVICEINFO we are caught between a rock and a hard place. To work around this issue we add a new hash that maps from a 64-bit integer to a fsid so that we can look up the export to authenticate against it, a 32-bit integer as a generation that we can bump when changing the device, and a currently unused 32-bit integer that could be used in the future to handle more than a single device per export. Entries in this hash table are never deleted as we can't reuse the ids anyway, and would have a severe lifetime problem anyway as Linux export structures are temporary structures that can go away under load. Parts of the XDR data, structures and marshaling/unmarshaling code, as well as many concepts are derived from the old pNFS server implementation from Andy Adamson, Benny Halevy, Dean Hildebrand, Marc Eshel, Fred Isaman, Mike Sager, Ricardo Labiaga and many others. Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-05-05 15:11:59 +04:00
nfsd4_setup_layout_type(&exp);
}
expp = svc_export_lookup(&exp);
if (expp)
expp = svc_export_update(&exp, expp);
else
err = -ENOMEM;
cache_flush();
if (expp == NULL)
err = -ENOMEM;
else
exp_put(expp);
out4:
nfsd4_fslocs_free(&exp.ex_fslocs);
kfree(exp.ex_uuid);
out3:
path_put(&exp.ex_path);
out1:
auth_domain_put(dom);
out:
kfree(buf);
return err;
}
static void exp_flags(struct seq_file *m, int flag, int fsid,
kuid_t anonu, kgid_t anong, struct nfsd4_fs_locations *fslocs);
static void show_secinfo(struct seq_file *m, struct svc_export *exp);
static int svc_export_show(struct seq_file *m,
struct cache_detail *cd,
struct cache_head *h)
{
struct svc_export *exp ;
if (h ==NULL) {
seq_puts(m, "#path domain(flags)\n");
return 0;
}
exp = container_of(h, struct svc_export, h);
seq_path(m, &exp->ex_path, " \t\n\\");
seq_putc(m, '\t');
seq_escape(m, exp->ex_client->name, " \t\n\\");
seq_putc(m, '(');
if (test_bit(CACHE_VALID, &h->flags) &&
!test_bit(CACHE_NEGATIVE, &h->flags)) {
exp_flags(m, exp->ex_flags, exp->ex_fsid,
exp->ex_anon_uid, exp->ex_anon_gid, &exp->ex_fslocs);
if (exp->ex_uuid) {
int i;
seq_puts(m, ",uuid=");
for (i = 0; i < EX_UUID_LEN; i++) {
if ((i&3) == 0 && i)
seq_putc(m, ':');
seq_printf(m, "%02x", exp->ex_uuid[i]);
}
}
show_secinfo(m, exp);
}
seq_puts(m, ")\n");
return 0;
}
static int svc_export_match(struct cache_head *a, struct cache_head *b)
{
struct svc_export *orig = container_of(a, struct svc_export, h);
struct svc_export *new = container_of(b, struct svc_export, h);
return orig->ex_client == new->ex_client &&
path_equal(&orig->ex_path, &new->ex_path);
}
static void svc_export_init(struct cache_head *cnew, struct cache_head *citem)
{
struct svc_export *new = container_of(cnew, struct svc_export, h);
struct svc_export *item = container_of(citem, struct svc_export, h);
kref_get(&item->ex_client->ref);
new->ex_client = item->ex_client;
new->ex_path = item->ex_path;
path_get(&item->ex_path);
new->ex_fslocs.locations = NULL;
new->ex_fslocs.locations_count = 0;
new->ex_fslocs.migrated = 0;
new->ex_layout_types = 0;
nfsd: initialize the exp->ex_uuid field in svc_export_init commit 885c91f7466 in Bruce's tree was causing oopses for me: general protection fault: 0000 [#1] SMP Modules linked in: nfsd(OF) nfs_acl(OF) auth_rpcgss(OF) lockd(OF) sunrpc(OF) kvm_amd kvm microcode i2c_piix4 virtio_net virtio_balloon cirrus drm_kms_helper ttm drm virtio_blk i2c_core CPU 0 Pid: 564, comm: exportfs Tainted: GF O 3.8.0-0.rc5.git2.1.fc19.x86_64 #1 Bochs Bochs RIP: 0010:[<ffffffff811b1509>] [<ffffffff811b1509>] kfree+0x49/0x280 RSP: 0018:ffff88007a3d7c50 EFLAGS: 00010203 RAX: 01adaf8dadadad80 RBX: 6b6b6b6b6b6b6b6b RCX: 0000000000000001 RDX: ffffffff7fffffff RSI: 0000000000000000 RDI: 6b6b6b6b6b6b6b6b RBP: ffff88007a3d7c80 R08: 6b6b6b6b6b6b6b6b R09: 0000000000000000 R10: 0000000000000018 R11: 0000000000000000 R12: ffff88006a117b50 R13: ffffffffa01a589c R14: ffff8800631b0f50 R15: 01ad998dadadad80 FS: 00007fcaa3616740(0000) GS:ffff88007fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 00007f5d84b6fdd8 CR3: 0000000064db4000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process exportfs (pid: 564, threadinfo ffff88007a3d6000, task ffff88006af28000) Stack: ffff88007a3d7c80 ffff88006a117b68 ffff88006a117b50 0000000000000000 ffff8800631b0f50 ffff88006a117b50 ffff88007a3d7ca0 ffffffffa01a589c ffff880036be1148 ffff88007a3d7cf8 ffff88007a3d7e28 ffffffffa01a6a98 Call Trace: [<ffffffffa01a589c>] svc_export_put+0x5c/0x70 [nfsd] [<ffffffffa01a6a98>] svc_export_parse+0x328/0x7e0 [nfsd] [<ffffffffa016f1c7>] cache_do_downcall+0x57/0x70 [sunrpc] [<ffffffffa016f25e>] cache_downcall+0x7e/0x100 [sunrpc] [<ffffffffa016f338>] cache_write_procfs+0x58/0x90 [sunrpc] [<ffffffffa016f2e0>] ? cache_downcall+0x100/0x100 [sunrpc] [<ffffffff8123b0e5>] proc_reg_write+0x75/0xb0 [<ffffffff811ccecf>] vfs_write+0x9f/0x170 [<ffffffff811cd089>] sys_write+0x49/0xa0 [<ffffffff816e0919>] system_call_fastpath+0x16/0x1b Code: 66 66 66 90 48 83 fb 10 0f 86 c3 00 00 00 48 89 df 49 bf 00 00 00 00 00 ea ff ff e8 f2 12 ea ff 48 c1 e8 0c 48 c1 e0 06 49 01 c7 <49> 8b 07 f6 c4 80 0f 85 1d 02 00 00 49 8b 07 a8 80 0f 84 ee 01 RIP [<ffffffff811b1509>] kfree+0x49/0x280 RSP <ffff88007a3d7c50> I think Majianpeng's patch is correct, but incomplete. In order for it to be safe to free the ex_uuid unconditionally in svc_export_put, we need to make sure it's initialized to NULL in the init routine. Cc: majianpeng <majianpeng@gmail.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com>
2013-02-02 14:55:08 +04:00
new->ex_uuid = NULL;
new->cd = item->cd;
}
static void export_update(struct cache_head *cnew, struct cache_head *citem)
{
struct svc_export *new = container_of(cnew, struct svc_export, h);
struct svc_export *item = container_of(citem, struct svc_export, h);
int i;
new->ex_flags = item->ex_flags;
new->ex_anon_uid = item->ex_anon_uid;
new->ex_anon_gid = item->ex_anon_gid;
new->ex_fsid = item->ex_fsid;
nfsd: implement pNFS operations Add support for the GETDEVICEINFO, LAYOUTGET, LAYOUTCOMMIT and LAYOUTRETURN NFSv4.1 operations, as well as backing code to manage outstanding layouts and devices. Layout management is very straight forward, with a nfs4_layout_stateid structure that extends nfs4_stid to manage layout stateids as the top-level structure. It is linked into the nfs4_file and nfs4_client structures like the other stateids, and contains a linked list of layouts that hang of the stateid. The actual layout operations are implemented in layout drivers that are not part of this commit, but will be added later. The worst part of this commit is the management of the pNFS device IDs, which suffers from a specification that is not sanely implementable due to the fact that the device-IDs are global and not bound to an export, and have a small enough size so that we can't store the fsid portion of a file handle, and must never be reused. As we still do need perform all export authentication and validation checks on a device ID passed to GETDEVICEINFO we are caught between a rock and a hard place. To work around this issue we add a new hash that maps from a 64-bit integer to a fsid so that we can look up the export to authenticate against it, a 32-bit integer as a generation that we can bump when changing the device, and a currently unused 32-bit integer that could be used in the future to handle more than a single device per export. Entries in this hash table are never deleted as we can't reuse the ids anyway, and would have a severe lifetime problem anyway as Linux export structures are temporary structures that can go away under load. Parts of the XDR data, structures and marshaling/unmarshaling code, as well as many concepts are derived from the old pNFS server implementation from Andy Adamson, Benny Halevy, Dean Hildebrand, Marc Eshel, Fred Isaman, Mike Sager, Ricardo Labiaga and many others. Signed-off-by: Christoph Hellwig <hch@lst.de>
2014-05-05 15:11:59 +04:00
new->ex_devid_map = item->ex_devid_map;
item->ex_devid_map = NULL;
new->ex_uuid = item->ex_uuid;
item->ex_uuid = NULL;
new->ex_fslocs.locations = item->ex_fslocs.locations;
item->ex_fslocs.locations = NULL;
new->ex_fslocs.locations_count = item->ex_fslocs.locations_count;
item->ex_fslocs.locations_count = 0;
new->ex_fslocs.migrated = item->ex_fslocs.migrated;
item->ex_fslocs.migrated = 0;
new->ex_layout_types = item->ex_layout_types;
new->ex_nflavors = item->ex_nflavors;
for (i = 0; i < MAX_SECINFO_LIST; i++) {
new->ex_flavors[i] = item->ex_flavors[i];
}
}
static struct cache_head *svc_export_alloc(void)
{
struct svc_export *i = kmalloc(sizeof(*i), GFP_KERNEL);
if (i)
return &i->h;
else
return NULL;
}
static struct cache_detail svc_export_cache_template = {
.owner = THIS_MODULE,
.hash_size = EXPORT_HASHMAX,
.name = "nfsd.export",
.cache_put = svc_export_put,
.cache_request = svc_export_request,
.cache_parse = svc_export_parse,
.cache_show = svc_export_show,
.match = svc_export_match,
.init = svc_export_init,
.update = export_update,
.alloc = svc_export_alloc,
};
static int
svc_export_hash(struct svc_export *exp)
{
int hash;
hash = hash_ptr(exp->ex_client, EXPORT_HASHBITS);
hash ^= hash_ptr(exp->ex_path.dentry, EXPORT_HASHBITS);
hash ^= hash_ptr(exp->ex_path.mnt, EXPORT_HASHBITS);
return hash;
}
static struct svc_export *
svc_export_lookup(struct svc_export *exp)
{
struct cache_head *ch;
int hash = svc_export_hash(exp);
ch = sunrpc_cache_lookup(exp->cd, &exp->h, hash);
if (ch)
return container_of(ch, struct svc_export, h);
else
return NULL;
}
static struct svc_export *
svc_export_update(struct svc_export *new, struct svc_export *old)
{
struct cache_head *ch;
int hash = svc_export_hash(old);
ch = sunrpc_cache_update(old->cd, &new->h, &old->h, hash);
if (ch)
return container_of(ch, struct svc_export, h);
else
return NULL;
}
static struct svc_expkey *
exp_find_key(struct cache_detail *cd, struct auth_domain *clp, int fsid_type,
u32 *fsidv, struct cache_req *reqp)
{
struct svc_expkey key, *ek;
int err;
if (!clp)
return ERR_PTR(-ENOENT);
key.ek_client = clp;
key.ek_fsidtype = fsid_type;
memcpy(key.ek_fsid, fsidv, key_len(fsid_type));
ek = svc_expkey_lookup(cd, &key);
if (ek == NULL)
return ERR_PTR(-ENOMEM);
err = cache_check(cd, &ek->h, reqp);
if (err)
return ERR_PTR(err);
return ek;
}
static struct svc_export *
exp_get_by_name(struct cache_detail *cd, struct auth_domain *clp,
const struct path *path, struct cache_req *reqp)
{
struct svc_export *exp, key;
int err;
if (!clp)
return ERR_PTR(-ENOENT);
key.ex_client = clp;
key.ex_path = *path;
key.cd = cd;
exp = svc_export_lookup(&key);
if (exp == NULL)
return ERR_PTR(-ENOMEM);
err = cache_check(cd, &exp->h, reqp);
if (err)
return ERR_PTR(err);
return exp;
}
/*
* Find the export entry for a given dentry.
*/
static struct svc_export *
exp_parent(struct cache_detail *cd, struct auth_domain *clp, struct path *path)
{
struct dentry *saved = dget(path->dentry);
struct svc_export *exp = exp_get_by_name(cd, clp, path, NULL);
while (PTR_ERR(exp) == -ENOENT && !IS_ROOT(path->dentry)) {
struct dentry *parent = dget_parent(path->dentry);
dput(path->dentry);
path->dentry = parent;
exp = exp_get_by_name(cd, clp, path, NULL);
}
dput(path->dentry);
path->dentry = saved;
return exp;
}
/*
* Obtain the root fh on behalf of a client.
* This could be done in user space, but I feel that it adds some safety
* since its harder to fool a kernel module than a user space program.
*/
int
exp_rootfh(struct net *net, struct auth_domain *clp, char *name,
struct knfsd_fh *f, int maxsize)
{
struct svc_export *exp;
struct path path;
struct inode *inode;
struct svc_fh fh;
int err;
struct nfsd_net *nn = net_generic(net, nfsd_net_id);
struct cache_detail *cd = nn->svc_export_cache;
err = -EPERM;
/* NB: we probably ought to check that it's NUL-terminated */
if (kern_path(name, 0, &path)) {
printk("nfsd: exp_rootfh path not found %s", name);
return err;
}
inode = d_inode(path.dentry);
dprintk("nfsd: exp_rootfh(%s [%p] %s:%s/%ld)\n",
name, path.dentry, clp->name,
inode->i_sb->s_id, inode->i_ino);
exp = exp_parent(cd, clp, &path);
if (IS_ERR(exp)) {
err = PTR_ERR(exp);
goto out;
}
/*
* fh must be initialized before calling fh_compose
*/
fh_init(&fh, maxsize);
if (fh_compose(&fh, exp, path.dentry, NULL))
err = -EINVAL;
else
err = 0;
memcpy(f, &fh.fh_handle, sizeof(struct knfsd_fh));
fh_put(&fh);
exp_put(exp);
out:
path_put(&path);
return err;
}
static struct svc_export *exp_find(struct cache_detail *cd,
struct auth_domain *clp, int fsid_type,
u32 *fsidv, struct cache_req *reqp)
{
struct svc_export *exp;
struct nfsd_net *nn = net_generic(cd->net, nfsd_net_id);
struct svc_expkey *ek = exp_find_key(nn->svc_expkey_cache, clp, fsid_type, fsidv, reqp);
if (IS_ERR(ek))
return ERR_CAST(ek);
exp = exp_get_by_name(cd, clp, &ek->ek_path, reqp);
cache_put(&ek->h, nn->svc_expkey_cache);
if (IS_ERR(exp))
return ERR_CAST(exp);
return exp;
}
__be32 check_nfsd_access(struct svc_export *exp, struct svc_rqst *rqstp)
{
struct exp_flavor_info *f;
struct exp_flavor_info *end = exp->ex_flavors + exp->ex_nflavors;
/* legacy gss-only clients are always OK: */
if (exp->ex_client == rqstp->rq_gssclient)
return 0;
/* ip-address based client; check sec= export option: */
for (f = exp->ex_flavors; f < end; f++) {
if (f->pseudoflavor == rqstp->rq_cred.cr_flavor)
return 0;
}
/* defaults in absence of sec= options: */
if (exp->ex_nflavors == 0) {
if (rqstp->rq_cred.cr_flavor == RPC_AUTH_NULL ||
rqstp->rq_cred.cr_flavor == RPC_AUTH_UNIX)
return 0;
}
/* If the compound op contains a spo_must_allowed op,
* it will be sent with integrity/protection which
* will have to be expressly allowed on mounts that
* don't support it
*/
if (nfsd4_spo_must_allow(rqstp))
return 0;
return nfserr_wrongsec;
}
/*
* Uses rq_client and rq_gssclient to find an export; uses rq_client (an
* auth_unix client) if it's available and has secinfo information;
* otherwise, will try to use rq_gssclient.
*
* Called from functions that handle requests; functions that do work on
* behalf of mountd are passed a single client name to use, and should
* use exp_get_by_name() or exp_find().
*/
struct svc_export *
rqst_exp_get_by_name(struct svc_rqst *rqstp, struct path *path)
{
struct svc_export *gssexp, *exp = ERR_PTR(-ENOENT);
struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
struct cache_detail *cd = nn->svc_export_cache;
if (rqstp->rq_client == NULL)
goto gss;
knfsd: nfsd: set rq_client to ip-address-determined-domain We want it to be possible for users to restrict exports both by IP address and by pseudoflavor. The pseudoflavor information has previously been passed using special auth_domains stored in the rq_client field. After the preceding patch that stored the pseudoflavor in rq_pflavor, that's now superfluous; so now we use rq_client for the ip information, as auth_null and auth_unix do. However, we keep around the special auth_domain in the rq_gssclient field for backwards compatibility purposes, so we can still do upcalls using the old "gss/pseudoflavor" auth_domain if upcalls using the unix domain to give us an appropriate export. This allows us to continue supporting old mountd. In fact, for this first patch, we always use the "gss/pseudoflavor" auth_domain (and only it) if it is available; thus rq_client is ignored in the auth_gss case, and this patch on its own makes no change in behavior; that will be left to later patches. Note on idmap: I'm almost tempted to just replace the auth_domain in the idmap upcall by a dummy value--no version of idmapd has ever used it, and it's unlikely anyone really wants to perform idmapping differently depending on the where the client is (they may want to perform *credential* mapping differently, but that's a different matter--the idmapper just handles id's used in getattr and setattr). But I'm updating the idmapd code anyway, just out of general backwards-compatibility paranoia. Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 15:04:46 +04:00
/* First try the auth_unix client: */
exp = exp_get_by_name(cd, rqstp->rq_client, path, &rqstp->rq_chandle);
if (PTR_ERR(exp) == -ENOENT)
goto gss;
if (IS_ERR(exp))
return exp;
/* If it has secinfo, assume there are no gss/... clients */
if (exp->ex_nflavors > 0)
return exp;
gss:
/* Otherwise, try falling back on gss client */
if (rqstp->rq_gssclient == NULL)
return exp;
gssexp = exp_get_by_name(cd, rqstp->rq_gssclient, path, &rqstp->rq_chandle);
if (PTR_ERR(gssexp) == -ENOENT)
return exp;
if (!IS_ERR(exp))
exp_put(exp);
return gssexp;
}
struct svc_export *
rqst_exp_find(struct svc_rqst *rqstp, int fsid_type, u32 *fsidv)
{
struct svc_export *gssexp, *exp = ERR_PTR(-ENOENT);
struct nfsd_net *nn = net_generic(SVC_NET(rqstp), nfsd_net_id);
struct cache_detail *cd = nn->svc_export_cache;
if (rqstp->rq_client == NULL)
goto gss;
knfsd: nfsd: set rq_client to ip-address-determined-domain We want it to be possible for users to restrict exports both by IP address and by pseudoflavor. The pseudoflavor information has previously been passed using special auth_domains stored in the rq_client field. After the preceding patch that stored the pseudoflavor in rq_pflavor, that's now superfluous; so now we use rq_client for the ip information, as auth_null and auth_unix do. However, we keep around the special auth_domain in the rq_gssclient field for backwards compatibility purposes, so we can still do upcalls using the old "gss/pseudoflavor" auth_domain if upcalls using the unix domain to give us an appropriate export. This allows us to continue supporting old mountd. In fact, for this first patch, we always use the "gss/pseudoflavor" auth_domain (and only it) if it is available; thus rq_client is ignored in the auth_gss case, and this patch on its own makes no change in behavior; that will be left to later patches. Note on idmap: I'm almost tempted to just replace the auth_domain in the idmap upcall by a dummy value--no version of idmapd has ever used it, and it's unlikely anyone really wants to perform idmapping differently depending on the where the client is (they may want to perform *credential* mapping differently, but that's a different matter--the idmapper just handles id's used in getattr and setattr). But I'm updating the idmapd code anyway, just out of general backwards-compatibility paranoia. Signed-off-by: "J. Bruce Fields" <bfields@citi.umich.edu> Signed-off-by: Neil Brown <neilb@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-17 15:04:46 +04:00
/* First try the auth_unix client: */
exp = exp_find(cd, rqstp->rq_client, fsid_type,
fsidv, &rqstp->rq_chandle);
if (PTR_ERR(exp) == -ENOENT)
goto gss;
if (IS_ERR(exp))
return exp;
/* If it has secinfo, assume there are no gss/... clients */
if (exp->ex_nflavors > 0)
return exp;
gss:
/* Otherwise, try falling back on gss client */
if (rqstp->rq_gssclient == NULL)
return exp;
gssexp = exp_find(cd, rqstp->rq_gssclient, fsid_type, fsidv,
&rqstp->rq_chandle);
if (PTR_ERR(gssexp) == -ENOENT)
return exp;
if (!IS_ERR(exp))
exp_put(exp);
return gssexp;
}
struct svc_export *
rqst_exp_parent(struct svc_rqst *rqstp, struct path *path)
{
struct dentry *saved = dget(path->dentry);
struct svc_export *exp = rqst_exp_get_by_name(rqstp, path);
while (PTR_ERR(exp) == -ENOENT && !IS_ROOT(path->dentry)) {
struct dentry *parent = dget_parent(path->dentry);
dput(path->dentry);
path->dentry = parent;
exp = rqst_exp_get_by_name(rqstp, path);
}
dput(path->dentry);
path->dentry = saved;
return exp;
}
struct svc_export *rqst_find_fsidzero_export(struct svc_rqst *rqstp)
{
u32 fsidv[2];
mk_fsid(FSID_NUM, fsidv, 0, 0, 0, NULL);
return rqst_exp_find(rqstp, FSID_NUM, fsidv);
}
/*
* Called when we need the filehandle for the root of the pseudofs,
* for a given NFSv4 client. The root is defined to be the
* export point with fsid==0
*/
__be32
exp_pseudoroot(struct svc_rqst *rqstp, struct svc_fh *fhp)
{
struct svc_export *exp;
__be32 rv;
exp = rqst_find_fsidzero_export(rqstp);
if (IS_ERR(exp))
return nfserrno(PTR_ERR(exp));
rv = fh_compose(fhp, exp, exp->ex_path.dentry, NULL);
exp_put(exp);
return rv;
}
static struct flags {
int flag;
char *name[2];
} expflags[] = {
{ NFSEXP_READONLY, {"ro", "rw"}},
{ NFSEXP_INSECURE_PORT, {"insecure", ""}},
{ NFSEXP_ROOTSQUASH, {"root_squash", "no_root_squash"}},
{ NFSEXP_ALLSQUASH, {"all_squash", ""}},
{ NFSEXP_ASYNC, {"async", "sync"}},
{ NFSEXP_GATHERED_WRITES, {"wdelay", "no_wdelay"}},
{ NFSEXP_NOREADDIRPLUS, {"nordirplus", ""}},
{ NFSEXP_NOHIDE, {"nohide", ""}},
{ NFSEXP_CROSSMOUNT, {"crossmnt", ""}},
{ NFSEXP_NOSUBTREECHECK, {"no_subtree_check", ""}},
{ NFSEXP_NOAUTHNLM, {"insecure_locks", ""}},
nfsd: introduce export flag for v4 pseudoroot NFSv4 differs from v2 and v3 in that it presents a single unified filesystem tree, whereas v2 and v3 exported multiple filesystem (whose roots could be found using a separate mount protocol). Our original NFSv4 server implementation asked the administrator to designate a single filesystem as the NFSv4 root, then to mount filesystems they wished to export underneath. (Often using bind mounts of already-existing filesystems.) This was conceptually simple, and allowed easy implementation, but created a serious obstacle to upgrading between v2/v3: since the paths to v4 filesystems were different, administrators would have to adjust all the paths in client-side mount commands when switching to v4. Various workarounds are possible. For example, the administrator could export "/" and designate it as the v4 root. However, the security risks of that approach are obvious, and in any case we shouldn't be requiring the administrator to take extra steps to fix this problem; instead, the server should present consistent paths across different versions by default. These patches take a modified version of that approach: we provide a new export option which exports only a subset of a filesystem. With this flag, it becomes safe for mountd to export "/" by default, with no need for additional configuration. We begin just by defining the new flag. Signed-off-by: Steve Dickson <steved@redhat.com> Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
2009-09-09 22:58:22 +04:00
{ NFSEXP_V4ROOT, {"v4root", ""}},
{ NFSEXP_PNFS, {"pnfs", ""}},
{ NFSEXP_SECURITY_LABEL, {"security_label", ""}},
{ 0, {"", ""}}
};
static void show_expflags(struct seq_file *m, int flags, int mask)
{
struct flags *flg;
int state, first = 0;
for (flg = expflags; flg->flag; flg++) {
if (flg->flag & ~mask)
continue;
state = (flg->flag & flags) ? 0 : 1;
if (*flg->name[state])
seq_printf(m, "%s%s", first++?",":"", flg->name[state]);
}
}
static void show_secinfo_flags(struct seq_file *m, int flags)
{
seq_printf(m, ",");
show_expflags(m, flags, NFSEXP_SECINFO_FLAGS);
}
static bool secinfo_flags_equal(int f, int g)
{
f &= NFSEXP_SECINFO_FLAGS;
g &= NFSEXP_SECINFO_FLAGS;
return f == g;
}
static int show_secinfo_run(struct seq_file *m, struct exp_flavor_info **fp, struct exp_flavor_info *end)
{
int flags;
flags = (*fp)->flags;
seq_printf(m, ",sec=%d", (*fp)->pseudoflavor);
(*fp)++;
while (*fp != end && secinfo_flags_equal(flags, (*fp)->flags)) {
seq_printf(m, ":%d", (*fp)->pseudoflavor);
(*fp)++;
}
return flags;
}
static void show_secinfo(struct seq_file *m, struct svc_export *exp)
{
struct exp_flavor_info *f;
struct exp_flavor_info *end = exp->ex_flavors + exp->ex_nflavors;
int flags;
if (exp->ex_nflavors == 0)
return;
f = exp->ex_flavors;
flags = show_secinfo_run(m, &f, end);
if (!secinfo_flags_equal(flags, exp->ex_flags))
show_secinfo_flags(m, flags);
while (f != end) {
flags = show_secinfo_run(m, &f, end);
show_secinfo_flags(m, flags);
}
}
static void exp_flags(struct seq_file *m, int flag, int fsid,
kuid_t anonu, kgid_t anong, struct nfsd4_fs_locations *fsloc)
{
show_expflags(m, flag, NFSEXP_ALLFLAGS);
if (flag & NFSEXP_FSID)
seq_printf(m, ",fsid=%d", fsid);
if (!uid_eq(anonu, make_kuid(&init_user_ns, (uid_t)-2)) &&
!uid_eq(anonu, make_kuid(&init_user_ns, 0x10000-2)))
seq_printf(m, ",anonuid=%u", from_kuid(&init_user_ns, anonu));
if (!gid_eq(anong, make_kgid(&init_user_ns, (gid_t)-2)) &&
!gid_eq(anong, make_kgid(&init_user_ns, 0x10000-2)))
seq_printf(m, ",anongid=%u", from_kgid(&init_user_ns, anong));
if (fsloc && fsloc->locations_count > 0) {
char *loctype = (fsloc->migrated) ? "refer" : "replicas";
int i;
seq_printf(m, ",%s=", loctype);
seq_escape(m, fsloc->locations[0].path, ",;@ \t\n\\");
seq_putc(m, '@');
seq_escape(m, fsloc->locations[0].hosts, ",;@ \t\n\\");
for (i = 1; i < fsloc->locations_count; i++) {
seq_putc(m, ';');
seq_escape(m, fsloc->locations[i].path, ",;@ \t\n\\");
seq_putc(m, '@');
seq_escape(m, fsloc->locations[i].hosts, ",;@ \t\n\\");
}
}
}
static int e_show(struct seq_file *m, void *p)
{
struct cache_head *cp = p;
struct svc_export *exp = container_of(cp, struct svc_export, h);
struct cache_detail *cd = m->private;
if (p == SEQ_START_TOKEN) {
seq_puts(m, "# Version 1.1\n");
seq_puts(m, "# Path Client(Flags) # IPs\n");
return 0;
}
exp_get(exp);
if (cache_check(cd, &exp->h, NULL))
return 0;
exp_put(exp);
return svc_export_show(m, cd, cp);
}
const struct seq_operations nfs_exports_op = {
.start = cache_seq_start,
.next = cache_seq_next,
.stop = cache_seq_stop,
.show = e_show,
};
/*
* Initialize the exports module.
*/
int
nfsd_export_init(struct net *net)
{
int rv;
struct nfsd_net *nn = net_generic(net, nfsd_net_id);
dprintk("nfsd: initializing export module (net: %p).\n", net);
nn->svc_export_cache = cache_create_net(&svc_export_cache_template, net);
if (IS_ERR(nn->svc_export_cache))
return PTR_ERR(nn->svc_export_cache);
rv = cache_register_net(nn->svc_export_cache, net);
if (rv)
goto destroy_export_cache;
nn->svc_expkey_cache = cache_create_net(&svc_expkey_cache_template, net);
if (IS_ERR(nn->svc_expkey_cache)) {
rv = PTR_ERR(nn->svc_expkey_cache);
goto unregister_export_cache;
}
rv = cache_register_net(nn->svc_expkey_cache, net);
if (rv)
goto destroy_expkey_cache;
return 0;
destroy_expkey_cache:
cache_destroy_net(nn->svc_expkey_cache, net);
unregister_export_cache:
cache_unregister_net(nn->svc_export_cache, net);
destroy_export_cache:
cache_destroy_net(nn->svc_export_cache, net);
return rv;
}
/*
* Flush exports table - called when last nfsd thread is killed
*/
void
nfsd_export_flush(struct net *net)
{
struct nfsd_net *nn = net_generic(net, nfsd_net_id);
cache_purge(nn->svc_expkey_cache);
cache_purge(nn->svc_export_cache);
}
/*
* Shutdown the exports module.
*/
void
nfsd_export_shutdown(struct net *net)
{
struct nfsd_net *nn = net_generic(net, nfsd_net_id);
dprintk("nfsd: shutting down export module (net: %p).\n", net);
cache_unregister_net(nn->svc_expkey_cache, net);
cache_unregister_net(nn->svc_export_cache, net);
cache_destroy_net(nn->svc_expkey_cache, net);
cache_destroy_net(nn->svc_export_cache, net);
svcauth_unix_purge(net);
dprintk("nfsd: export shutdown complete (net: %p).\n", net);
}