/*
Unix SMB/CIFS implementation.
POSIX NTVFS backend - ACL support
Copyright (C) Andrew Tridgell 2004
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
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, see .
*/
#include "includes.h"
#include "auth/auth.h"
#include "vfs_posix.h"
#include "librpc/gen_ndr/xattr.h"
#include "libcli/security/security.h"
/* the list of currently registered ACL backends */
static struct pvfs_acl_backend {
const struct pvfs_acl_ops *ops;
} *backends = NULL;
static int num_backends;
/*
register a pvfs acl backend.
The 'name' can be later used by other backends to find the operations
structure for this backend.
*/
NTSTATUS pvfs_acl_register(const struct pvfs_acl_ops *ops)
{
struct pvfs_acl_ops *new_ops;
if (pvfs_acl_backend_byname(ops->name) != NULL) {
DEBUG(0,("pvfs acl backend '%s' already registered\n", ops->name));
return NT_STATUS_OBJECT_NAME_COLLISION;
}
backends = talloc_realloc(talloc_autofree_context(), backends, struct pvfs_acl_backend, num_backends+1);
NT_STATUS_HAVE_NO_MEMORY(backends);
new_ops = talloc_memdup(backends, ops, sizeof(*ops));
new_ops->name = talloc_strdup(new_ops, ops->name);
backends[num_backends].ops = new_ops;
num_backends++;
DEBUG(3,("NTVFS backend '%s' registered\n", ops->name));
return NT_STATUS_OK;
}
/*
return the operations structure for a named backend
*/
const struct pvfs_acl_ops *pvfs_acl_backend_byname(const char *name)
{
int i;
for (i=0;iname, name) == 0) {
return backends[i].ops;
}
}
return NULL;
}
/*
map a single access_mask from generic to specific bits for files/dirs
*/
static uint32_t pvfs_translate_mask(uint32_t access_mask)
{
if (access_mask & SEC_MASK_GENERIC) {
if (access_mask & SEC_GENERIC_READ) access_mask |= SEC_RIGHTS_FILE_READ;
if (access_mask & SEC_GENERIC_WRITE) access_mask |= SEC_RIGHTS_FILE_WRITE;
if (access_mask & SEC_GENERIC_EXECUTE) access_mask |= SEC_RIGHTS_FILE_EXECUTE;
if (access_mask & SEC_GENERIC_ALL) access_mask |= SEC_RIGHTS_FILE_ALL;
access_mask &= ~SEC_MASK_GENERIC;
}
return access_mask;
}
/*
map any generic access bits in the given acl
this relies on the fact that the mappings for files and directories
are the same
*/
static void pvfs_translate_generic_bits(struct security_acl *acl)
{
unsigned i;
if (!acl) return;
for (i=0;inum_aces;i++) {
struct security_ace *ace = &acl->aces[i];
ace->access_mask = pvfs_translate_mask(ace->access_mask);
}
}
/*
setup a default ACL for a file
*/
static NTSTATUS pvfs_default_acl(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name, int fd,
struct security_descriptor **psd)
{
struct security_descriptor *sd;
NTSTATUS status;
struct security_ace ace;
mode_t mode;
struct id_mapping *ids;
struct composite_context *ctx;
*psd = security_descriptor_initialise(req);
if (*psd == NULL) {
return NT_STATUS_NO_MEMORY;
}
sd = *psd;
ids = talloc_zero_array(sd, struct id_mapping, 2);
NT_STATUS_HAVE_NO_MEMORY(ids);
ids[0].unixid = talloc(ids, struct unixid);
NT_STATUS_HAVE_NO_MEMORY(ids[0].unixid);
ids[0].unixid->id = name->st.st_uid;
ids[0].unixid->type = ID_TYPE_UID;
ids[0].sid = NULL;
ids[1].unixid = talloc(ids, struct unixid);
NT_STATUS_HAVE_NO_MEMORY(ids[1].unixid);
ids[1].unixid->id = name->st.st_gid;
ids[1].unixid->type = ID_TYPE_GID;
ids[1].sid = NULL;
ctx = wbc_xids_to_sids_send(pvfs->wbc_ctx, ids, 2, ids);
NT_STATUS_HAVE_NO_MEMORY(ctx);
status = wbc_xids_to_sids_recv(ctx, &ids);
NT_STATUS_NOT_OK_RETURN(status);
sd->owner_sid = talloc_steal(sd, ids[0].sid);
sd->group_sid = talloc_steal(sd, ids[1].sid);
talloc_free(ids);
sd->type |= SEC_DESC_DACL_PRESENT;
mode = name->st.st_mode;
/*
we provide up to 4 ACEs
- Owner
- Group
- Everyone
- Administrator
*/
/* setup owner ACE */
ace.type = SEC_ACE_TYPE_ACCESS_ALLOWED;
ace.flags = 0;
ace.trustee = *sd->owner_sid;
ace.access_mask = 0;
if (mode & S_IRUSR) {
if (mode & S_IWUSR) {
ace.access_mask |= SEC_RIGHTS_FILE_ALL;
} else {
ace.access_mask |= SEC_RIGHTS_FILE_READ | SEC_FILE_EXECUTE;
}
}
if (mode & S_IWUSR) {
ace.access_mask |= SEC_RIGHTS_FILE_WRITE | SEC_STD_DELETE;
}
if (ace.access_mask) {
security_descriptor_dacl_add(sd, &ace);
}
/* setup group ACE */
ace.trustee = *sd->group_sid;
ace.access_mask = 0;
if (mode & S_IRGRP) {
ace.access_mask |= SEC_RIGHTS_FILE_READ | SEC_FILE_EXECUTE;
}
if (mode & S_IWGRP) {
/* note that delete is not granted - this matches posix behaviour */
ace.access_mask |= SEC_RIGHTS_FILE_WRITE;
}
if (ace.access_mask) {
security_descriptor_dacl_add(sd, &ace);
}
/* setup other ACE */
ace.trustee = *dom_sid_parse_talloc(req, SID_WORLD);
ace.access_mask = 0;
if (mode & S_IROTH) {
ace.access_mask |= SEC_RIGHTS_FILE_READ | SEC_FILE_EXECUTE;
}
if (mode & S_IWOTH) {
ace.access_mask |= SEC_RIGHTS_FILE_WRITE;
}
if (ace.access_mask) {
security_descriptor_dacl_add(sd, &ace);
}
/* setup system ACE */
ace.trustee = *dom_sid_parse_talloc(req, SID_NT_SYSTEM);
ace.access_mask = SEC_RIGHTS_FILE_ALL;
security_descriptor_dacl_add(sd, &ace);
return NT_STATUS_OK;
}
/*
omit any security_descriptor elements not specified in the given
secinfo flags
*/
static void normalise_sd_flags(struct security_descriptor *sd, uint32_t secinfo_flags)
{
if (!(secinfo_flags & SECINFO_OWNER)) {
sd->owner_sid = NULL;
}
if (!(secinfo_flags & SECINFO_GROUP)) {
sd->group_sid = NULL;
}
if (!(secinfo_flags & SECINFO_DACL)) {
sd->dacl = NULL;
}
if (!(secinfo_flags & SECINFO_SACL)) {
sd->sacl = NULL;
}
}
/*
answer a setfileinfo for an ACL
*/
NTSTATUS pvfs_acl_set(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name, int fd,
uint32_t access_mask,
union smb_setfileinfo *info)
{
uint32_t secinfo_flags = info->set_secdesc.in.secinfo_flags;
struct security_descriptor *new_sd, *sd, orig_sd;
NTSTATUS status = NT_STATUS_NOT_FOUND;
uid_t old_uid = -1;
gid_t old_gid = -1;
uid_t new_uid = -1;
gid_t new_gid = -1;
struct id_mapping *ids;
struct composite_context *ctx;
if (pvfs->acl_ops != NULL) {
status = pvfs->acl_ops->acl_load(pvfs, name, fd, req, &sd);
}
if (NT_STATUS_EQUAL(status, NT_STATUS_NOT_FOUND)) {
status = pvfs_default_acl(pvfs, req, name, fd, &sd);
}
if (!NT_STATUS_IS_OK(status)) {
return status;
}
ids = talloc(req, struct id_mapping);
NT_STATUS_HAVE_NO_MEMORY(ids);
ids->unixid = NULL;
ids->sid = NULL;
ids->status = NT_STATUS_NONE_MAPPED;
new_sd = info->set_secdesc.in.sd;
orig_sd = *sd;
old_uid = name->st.st_uid;
old_gid = name->st.st_gid;
/* only set the elements that have been specified */
if (secinfo_flags & SECINFO_OWNER) {
if (!(access_mask & SEC_STD_WRITE_OWNER)) {
return NT_STATUS_ACCESS_DENIED;
}
if (!dom_sid_equal(sd->owner_sid, new_sd->owner_sid)) {
ids->sid = new_sd->owner_sid;
ctx = wbc_sids_to_xids_send(pvfs->wbc_ctx, ids, 1, ids);
NT_STATUS_HAVE_NO_MEMORY(ctx);
status = wbc_sids_to_xids_recv(ctx, &ids);
NT_STATUS_NOT_OK_RETURN(status);
if (ids->unixid->type == ID_TYPE_BOTH ||
ids->unixid->type == ID_TYPE_UID) {
new_uid = ids->unixid->id;
}
}
sd->owner_sid = new_sd->owner_sid;
}
if (secinfo_flags & SECINFO_GROUP) {
if (!(access_mask & SEC_STD_WRITE_OWNER)) {
return NT_STATUS_ACCESS_DENIED;
}
if (!dom_sid_equal(sd->group_sid, new_sd->group_sid)) {
ids->sid = new_sd->group_sid;
ctx = wbc_sids_to_xids_send(pvfs->wbc_ctx, ids, 1, ids);
NT_STATUS_HAVE_NO_MEMORY(ctx);
status = wbc_sids_to_xids_recv(ctx, &ids);
NT_STATUS_NOT_OK_RETURN(status);
if (ids->unixid->type == ID_TYPE_BOTH ||
ids->unixid->type == ID_TYPE_GID) {
new_gid = ids->unixid->id;
}
}
sd->group_sid = new_sd->group_sid;
}
if (secinfo_flags & SECINFO_DACL) {
if (!(access_mask & SEC_STD_WRITE_DAC)) {
return NT_STATUS_ACCESS_DENIED;
}
sd->dacl = new_sd->dacl;
pvfs_translate_generic_bits(sd->dacl);
}
if (secinfo_flags & SECINFO_SACL) {
if (!(access_mask & SEC_FLAG_SYSTEM_SECURITY)) {
return NT_STATUS_ACCESS_DENIED;
}
sd->sacl = new_sd->sacl;
pvfs_translate_generic_bits(sd->sacl);
}
if (new_uid == old_uid) {
new_uid = -1;
}
if (new_gid == old_gid) {
new_gid = -1;
}
/* if there's something to change try it */
if (new_uid != -1 || new_gid != -1) {
int ret;
if (fd == -1) {
ret = chown(name->full_name, new_uid, new_gid);
} else {
ret = fchown(fd, new_uid, new_gid);
}
if (ret == -1) {
return pvfs_map_errno(pvfs, errno);
}
}
/* we avoid saving if the sd is the same. This means when clients
copy files and end up copying the default sd that we don't
needlessly use xattrs */
if (!security_descriptor_equal(sd, &orig_sd) && pvfs->acl_ops) {
status = pvfs->acl_ops->acl_save(pvfs, name, fd, sd);
}
return status;
}
/*
answer a fileinfo query for the ACL
*/
NTSTATUS pvfs_acl_query(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name, int fd,
union smb_fileinfo *info)
{
NTSTATUS status = NT_STATUS_NOT_FOUND;
struct security_descriptor *sd;
if (pvfs->acl_ops) {
status = pvfs->acl_ops->acl_load(pvfs, name, fd, req, &sd);
}
if (NT_STATUS_EQUAL(status, NT_STATUS_NOT_FOUND)) {
status = pvfs_default_acl(pvfs, req, name, fd, &sd);
}
if (!NT_STATUS_IS_OK(status)) {
return status;
}
normalise_sd_flags(sd, info->query_secdesc.in.secinfo_flags);
info->query_secdesc.out.sd = sd;
return NT_STATUS_OK;
}
/*
check the read only bit against any of the write access bits
*/
static bool pvfs_read_only(struct pvfs_state *pvfs, uint32_t access_mask)
{
if ((pvfs->flags & PVFS_FLAG_READONLY) &&
(access_mask & (SEC_FILE_WRITE_DATA |
SEC_FILE_APPEND_DATA |
SEC_FILE_WRITE_EA |
SEC_FILE_WRITE_ATTRIBUTE |
SEC_STD_DELETE |
SEC_STD_WRITE_DAC |
SEC_STD_WRITE_OWNER |
SEC_DIR_DELETE_CHILD))) {
return true;
}
return false;
}
/*
default access check function based on unix permissions
doing this saves on building a full security descriptor
for the common case of access check on files with no
specific NT ACL
*/
NTSTATUS pvfs_access_check_unix(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t *access_mask)
{
uid_t uid = geteuid();
uint32_t max_bits = SEC_RIGHTS_FILE_READ | SEC_FILE_ALL;
if (pvfs_read_only(pvfs, *access_mask)) {
return NT_STATUS_ACCESS_DENIED;
}
/* owner and root get extra permissions */
if (uid == 0) {
max_bits |= SEC_STD_ALL | SEC_FLAG_SYSTEM_SECURITY;
} else if (uid == name->st.st_uid) {
max_bits |= SEC_STD_ALL;
}
if (*access_mask == SEC_FLAG_MAXIMUM_ALLOWED) {
*access_mask = max_bits;
return NT_STATUS_OK;
}
if (uid != 0 && (*access_mask & SEC_FLAG_SYSTEM_SECURITY)) {
return NT_STATUS_ACCESS_DENIED;
}
if (*access_mask & ~max_bits) {
return NT_STATUS_ACCESS_DENIED;
}
if (pvfs->ntvfs->ctx->protocol != PROTOCOL_SMB2) {
/* on SMB, this bit is always granted, even if not
asked for */
*access_mask |= SEC_FILE_READ_ATTRIBUTE;
}
return NT_STATUS_OK;
}
/*
check the security descriptor on a file, if any
*access_mask is modified with the access actually granted
*/
NTSTATUS pvfs_access_check(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t *access_mask)
{
struct security_token *token = req->session_info->security_token;
struct xattr_NTACL *acl;
NTSTATUS status;
struct security_descriptor *sd;
if (pvfs_read_only(pvfs, *access_mask)) {
return NT_STATUS_ACCESS_DENIED;
}
acl = talloc(req, struct xattr_NTACL);
if (acl == NULL) {
return NT_STATUS_NO_MEMORY;
}
/* expand the generic access bits to file specific bits */
*access_mask = pvfs_translate_mask(*access_mask);
*access_mask &= ~SEC_FILE_READ_ATTRIBUTE;
status = pvfs_acl_load(pvfs, name, -1, acl);
if (NT_STATUS_EQUAL(status, NT_STATUS_NOT_FOUND)) {
talloc_free(acl);
return pvfs_access_check_unix(pvfs, req, name, access_mask);
}
if (!NT_STATUS_IS_OK(status)) {
return status;
}
switch (acl->version) {
case 1:
sd = acl->info.sd;
break;
default:
return NT_STATUS_INVALID_ACL;
}
/* check the acl against the required access mask */
status = sec_access_check(sd, token, *access_mask, access_mask);
if (pvfs->ntvfs->ctx->protocol != PROTOCOL_SMB2) {
/* on SMB, this bit is always granted, even if not
asked for */
*access_mask |= SEC_FILE_READ_ATTRIBUTE;
}
talloc_free(acl);
return status;
}
/*
a simplified interface to access check, designed for calls that
do not take or return an access check mask
*/
NTSTATUS pvfs_access_check_simple(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t access_needed)
{
if (access_needed == 0) {
return NT_STATUS_OK;
}
return pvfs_access_check(pvfs, req, name, &access_needed);
}
/*
access check for creating a new file/directory
*/
NTSTATUS pvfs_access_check_create(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t *access_mask)
{
struct pvfs_filename *parent;
NTSTATUS status;
status = pvfs_resolve_parent(pvfs, req, name, &parent);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
status = pvfs_access_check(pvfs, req, parent, access_mask);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
if (! ((*access_mask) & SEC_DIR_ADD_FILE)) {
return pvfs_access_check_simple(pvfs, req, parent, SEC_DIR_ADD_FILE);
}
return status;
}
/*
access check for creating a new file/directory - no access mask supplied
*/
NTSTATUS pvfs_access_check_parent(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t access_mask)
{
struct pvfs_filename *parent;
NTSTATUS status;
status = pvfs_resolve_parent(pvfs, req, name, &parent);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
return pvfs_access_check_simple(pvfs, req, parent, access_mask);
}
/*
determine if an ACE is inheritable
*/
static bool pvfs_inheritable_ace(struct pvfs_state *pvfs,
const struct security_ace *ace,
bool container)
{
if (!container) {
return (ace->flags & SEC_ACE_FLAG_OBJECT_INHERIT) != 0;
}
if (ace->flags & SEC_ACE_FLAG_CONTAINER_INHERIT) {
return true;
}
if ((ace->flags & SEC_ACE_FLAG_OBJECT_INHERIT) &&
!(ace->flags & SEC_ACE_FLAG_NO_PROPAGATE_INHERIT)) {
return true;
}
return false;
}
/*
this is the core of ACL inheritance. It copies any inheritable
aces from the parent SD to the child SD. Note that the algorithm
depends on whether the child is a container or not
*/
static NTSTATUS pvfs_acl_inherit_aces(struct pvfs_state *pvfs,
struct security_descriptor *parent_sd,
struct security_descriptor *sd,
bool container)
{
int i;
for (i=0;idacl->num_aces;i++) {
struct security_ace ace = parent_sd->dacl->aces[i];
NTSTATUS status;
const struct dom_sid *creator = NULL, *new_id = NULL;
uint32_t orig_flags;
if (!pvfs_inheritable_ace(pvfs, &ace, container)) {
continue;
}
orig_flags = ace.flags;
/* see the RAW-ACLS inheritance test for details on these rules */
if (!container) {
ace.flags = 0;
} else {
ace.flags &= ~SEC_ACE_FLAG_INHERIT_ONLY;
if (!(ace.flags & SEC_ACE_FLAG_CONTAINER_INHERIT)) {
ace.flags |= SEC_ACE_FLAG_INHERIT_ONLY;
}
if (ace.flags & SEC_ACE_FLAG_NO_PROPAGATE_INHERIT) {
ace.flags = 0;
}
}
/* the CREATOR sids are special when inherited */
if (dom_sid_equal(&ace.trustee, pvfs->sid_cache.creator_owner)) {
creator = pvfs->sid_cache.creator_owner;
new_id = sd->owner_sid;
} else if (dom_sid_equal(&ace.trustee, pvfs->sid_cache.creator_group)) {
creator = pvfs->sid_cache.creator_group;
new_id = sd->group_sid;
} else {
new_id = &ace.trustee;
}
if (creator && container &&
(ace.flags & SEC_ACE_FLAG_CONTAINER_INHERIT)) {
uint32_t flags = ace.flags;
ace.trustee = *new_id;
ace.flags = 0;
status = security_descriptor_dacl_add(sd, &ace);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
ace.trustee = *creator;
ace.flags = flags | SEC_ACE_FLAG_INHERIT_ONLY;
status = security_descriptor_dacl_add(sd, &ace);
} else if (container &&
!(orig_flags & SEC_ACE_FLAG_NO_PROPAGATE_INHERIT)) {
status = security_descriptor_dacl_add(sd, &ace);
} else {
ace.trustee = *new_id;
status = security_descriptor_dacl_add(sd, &ace);
}
if (!NT_STATUS_IS_OK(status)) {
return status;
}
}
return NT_STATUS_OK;
}
/*
setup an ACL on a new file/directory based on the inherited ACL from
the parent. If there is no inherited ACL then we don't set anything,
as the default ACL applies anyway
*/
NTSTATUS pvfs_acl_inherit(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
int fd)
{
struct xattr_NTACL *acl;
NTSTATUS status;
struct pvfs_filename *parent;
struct security_descriptor *parent_sd, *sd;
bool container;
struct id_mapping *ids;
struct composite_context *ctx;
/* form the parents path */
status = pvfs_resolve_parent(pvfs, req, name, &parent);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
acl = talloc(req, struct xattr_NTACL);
if (acl == NULL) {
return NT_STATUS_NO_MEMORY;
}
status = pvfs_acl_load(pvfs, parent, -1, acl);
if (NT_STATUS_EQUAL(status, NT_STATUS_NOT_FOUND)) {
return NT_STATUS_OK;
}
if (!NT_STATUS_IS_OK(status)) {
return status;
}
switch (acl->version) {
case 1:
parent_sd = acl->info.sd;
break;
default:
return NT_STATUS_INVALID_ACL;
}
if (parent_sd == NULL ||
parent_sd->dacl == NULL ||
parent_sd->dacl->num_aces == 0) {
/* go with the default ACL */
return NT_STATUS_OK;
}
/* create the new sd */
sd = security_descriptor_initialise(req);
if (sd == NULL) {
return NT_STATUS_NO_MEMORY;
}
ids = talloc_array(sd, struct id_mapping, 2);
NT_STATUS_HAVE_NO_MEMORY(ids);
ids[0].unixid = talloc(ids, struct unixid);
NT_STATUS_HAVE_NO_MEMORY(ids[0].unixid);
ids[0].unixid->id = name->st.st_uid;
ids[0].unixid->type = ID_TYPE_UID;
ids[0].sid = NULL;
ids[0].status = NT_STATUS_NONE_MAPPED;
ids[1].unixid = talloc(ids, struct unixid);
NT_STATUS_HAVE_NO_MEMORY(ids[1].unixid);
ids[1].unixid->id = name->st.st_gid;
ids[1].unixid->type = ID_TYPE_GID;
ids[1].sid = NULL;
ids[1].status = NT_STATUS_NONE_MAPPED;
ctx = wbc_xids_to_sids_send(pvfs->wbc_ctx, ids, 2, ids);
NT_STATUS_HAVE_NO_MEMORY(ctx);
status = wbc_xids_to_sids_recv(ctx, &ids);
NT_STATUS_NOT_OK_RETURN(status);
sd->owner_sid = talloc_steal(sd, ids[0].sid);
sd->group_sid = talloc_steal(sd, ids[1].sid);
sd->type |= SEC_DESC_DACL_PRESENT;
container = (name->dos.attrib & FILE_ATTRIBUTE_DIRECTORY) ? true:false;
/* fill in the aces from the parent */
status = pvfs_acl_inherit_aces(pvfs, parent_sd, sd, container);
if (!NT_STATUS_IS_OK(status)) {
return status;
}
/* if there is nothing to inherit then we fallback to the
default acl */
if (sd->dacl == NULL || sd->dacl->num_aces == 0) {
return NT_STATUS_OK;
}
acl->info.sd = sd;
status = pvfs_acl_save(pvfs, name, fd, acl);
return status;
}
/*
return the maximum allowed access mask
*/
NTSTATUS pvfs_access_maximal_allowed(struct pvfs_state *pvfs,
struct ntvfs_request *req,
struct pvfs_filename *name,
uint32_t *maximal_access)
{
*maximal_access = SEC_FLAG_MAXIMUM_ALLOWED;
return pvfs_access_check(pvfs, req, name, maximal_access);
}