sin/samba-mirror
1
0
Fork 0
mirror of https://github.com/samba-team/samba.git synced 2024-12-25 23:21:54 +03:00
Code
9342b3ebf7
samba-mirror/source3/modules/vfs_tru64acl.c

491 lines
13 KiB
C
Raw Normal View History

r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
/*
Unix SMB/Netbios implementation.
VFS module to get and set Tru64 acls
Remove redundant parameter fd from SMB_VFS_SYS_ACL_SET_FD(). Michael (This used to be commit 9296e93588c0e795cae770765050247ac1474a74)
2008-01-08 03:54:19 +03:00
Copyright (C) Michael Adam 2006,2008
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
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
r23779: Change from v2 or later to v3 or later. Jeremy. (This used to be commit 407e6e695b8366369b7c76af1ff76869b45347b3)
2007-07-09 23:25:36 +04:00
the Free Software Foundation; either version 3 of the License, or
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
(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
r23784: use the GPLv3 boilerplate as recommended by the FSF and the license text (This used to be commit b0132e94fc5fef936aa766fb99a306b3628e9f07)
2007-07-10 04:52:41 +04:00
along with this program. If not, see <http://www.gnu.org/licenses/>.
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
*/
#include "includes.h"
s3-vfs acl modules: more non-linux build fixes. Guenther Autobuild-User: Günther Deschner <gd@samba.org> Autobuild-Date: Wed Mar 30 16:00:02 CEST 2011 on sn-devel-104
2011-03-30 17:14:05 +04:00
#include "system/filesys.h"
s3-vfs: include smbd/smbd.h in vfs modules. Guenther
2011-03-23 00:34:22 +03:00
#include "smbd/smbd.h"
s3-includes: only include acl prototypes when needed. Guenther
2011-04-14 01:48:33 +04:00
#include "modules/vfs_tru64acl.h"
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
/* prototypes for private functions first - for clarity */
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
static struct smb_acl_t *tru64_acl_to_smb_acl(const struct acl *tru64_acl,
TALLOC_CTX *mem_ctx);
RIP BOOL. Convert BOOL -> bool. I found a few interesting bugs in various places whilst doing this (places that assumed BOOL == int). I also need to fix the Samba4 pidl generation (next checkin). Jeremy. (This used to be commit f35a266b3cbb3e5fa6a86be60f34fe340a3ca71f)
2007-10-19 04:40:25 +04:00
static bool tru64_ace_to_smb_ace(acl_entry_t tru64_ace,
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
struct smb_acl_entry *smb_ace);
static acl_t smb_acl_to_tru64_acl(const SMB_ACL_T smb_acl);
static acl_tag_t smb_tag_to_tru64(SMB_ACL_TAG_T smb_tag);
static SMB_ACL_TAG_T tru64_tag_to_smb(acl_tag_t tru64_tag);
r22918: Attempt to fix the build of the tru64acl module. Where the heck did that smb_acl_permset_t come from? I can't remember... Michael (This used to be commit e55f952a1319de8743ddfe4993cc4c798b7aa725)
2007-05-16 01:41:02 +04:00
static acl_perm_t smb_permset_to_tru64(SMB_ACL_PERM_T smb_permset);
static SMB_ACL_PERM_T tru64_permset_to_smb(const acl_perm_t tru64_permset);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
/* public functions - the api */
SMB_ACL_T tru64acl_sys_acl_get_file(vfs_handle_struct *handle,
const char *path_p,
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
SMB_ACL_TYPE_T type,
TALLOC_CTX *mem_ctx)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
struct smb_acl_t *result;
acl_type_t the_acl_type;
acl_t tru64_acl;
DEBUG(10, ("Hi! This is tru64acl_sys_acl_get_file.\n"));
switch(type) {
case SMB_ACL_TYPE_ACCESS:
the_acl_type = ACL_TYPE_ACCESS;
break;
case SMB_ACL_TYPE_DEFAULT:
the_acl_type = ACL_TYPE_DEFAULT;
break;
default:
errno = EINVAL;
return NULL;
}
tru64_acl = acl_get_file((char *)path_p, the_acl_type);
if (tru64_acl == NULL) {
return NULL;
}
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
result = tru64_acl_to_smb_acl(tru64_acl, mem_ctx);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
acl_free(tru64_acl);
return result;
}
SMB_ACL_T tru64acl_sys_acl_get_fd(vfs_handle_struct *handle,
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
files_struct *fsp,
TALLOC_CTX *mem_ctx)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
struct smb_acl_t *result;
Remove redundant parameter fd from SMB_VFS_SYS_ACL_GET_FD(). Michael (This used to be commit 42663e8736e1a3dfb57e0aafdcbf5fec880da779)
2008-01-08 01:53:34 +03:00
acl_t tru64_acl = acl_get_fd(fsp->fh->fd, ACL_TYPE_ACCESS);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
if (tru64_acl == NULL) {
return NULL;
}
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
result = tru64_acl_to_smb_acl(tru64_acl, mem_ctx);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
acl_free(tru64_acl);
return result;
}
int tru64acl_sys_acl_set_file(vfs_handle_struct *handle,
const char *name,
SMB_ACL_TYPE_T type,
SMB_ACL_T theacl)
{
int res;
acl_type_t the_acl_type;
acl_t tru64_acl;
DEBUG(10, ("tru64acl_sys_acl_set_file called with name %s, type %d\n",
name, type));
switch(type) {
case SMB_ACL_TYPE_ACCESS:
DEBUGADD(10, ("got acl type ACL_TYPE_ACCESS\n"));
the_acl_type = ACL_TYPE_ACCESS;
break;
case SMB_ACL_TYPE_DEFAULT:
DEBUGADD(10, ("got acl type ACL_TYPE_DEFAULT\n"));
the_acl_type = ACL_TYPE_DEFAULT;
break;
default:
DEBUGADD(10, ("invalid acl type\n"));
errno = EINVAL;
goto fail;
}
tru64_acl = smb_acl_to_tru64_acl(theacl);
if (tru64_acl == NULL) {
DEBUG(10, ("smb_acl_to_tru64_acl failed!\n"));
goto fail;
}
DEBUG(10, ("got tru64 acl...\n"));
res = acl_set_file((char *)name, the_acl_type, tru64_acl);
acl_free(tru64_acl);
if (res != 0) {
DEBUG(10, ("acl_set_file failed: %s\n", strerror(errno)));
goto fail;
}
return res;
fail:
DEBUG(1, ("tru64acl_sys_acl_set_file failed!\n"));
return -1;
}
int tru64acl_sys_acl_set_fd(vfs_handle_struct *handle,
files_struct *fsp,
Remove redundant parameter fd from SMB_VFS_SYS_ACL_SET_FD(). Michael (This used to be commit 9296e93588c0e795cae770765050247ac1474a74)
2008-01-08 03:54:19 +03:00
SMB_ACL_T theacl)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
int res;
acl_t tru64_acl = smb_acl_to_tru64_acl(theacl);
if (tru64_acl == NULL) {
return -1;
}
Remove redundant parameter fd from SMB_VFS_SYS_ACL_SET_FD(). Michael (This used to be commit 9296e93588c0e795cae770765050247ac1474a74)
2008-01-08 03:54:19 +03:00
res = acl_set_fd(fsp->fh->fd, ACL_TYPE_ACCESS, tru64_acl);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
acl_free(tru64_acl);
return res;
}
int tru64acl_sys_acl_delete_def_file(vfs_handle_struct *handle,
const char *path)
{
return acl_delete_def_file((char *)path);
}
/* private functions */
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
static struct smb_acl_t *tru64_acl_to_smb_acl(const struct acl *tru64_acl,
TALLOC_CTX *mem_ctx)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
struct smb_acl_t *result;
acl_entry_t entry;
DEBUG(10, ("Hi! This is tru64_acl_to_smb_acl.\n"));
smbd: Add mem_ctx to sys_acl_init() and all callers This changes from allocation on NULL to allocation on the supplied memory context. Currently that supplied context is talloc_tos() at the the final consumer of the ACL. Andrew Bartlett
2012-10-10 03:18:32 +04:00
if ((result = sys_acl_init(mem_ctx)) == NULL) {
s3-smbd: Change allocation of smb_acl_t to talloc() The acl element is changed to be a talloc child, and is no longer one element longer than requested by virtue of the acl[1] base pointer. This also avoids one of the few remaining cases of over-allocation of a structure. Andrew Bartlett
2012-08-12 14:41:35 +04:00
DEBUG(0, ("sys_acl_init() failed in tru64_acl_to_smb_acl\n"));
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
errno = ENOMEM;
goto fail;
}
if (acl_first_entry((struct acl *)tru64_acl) != 0) {
DEBUG(10, ("acl_first_entry failed: %s\n", strerror(errno)));
goto fail;
}
while ((entry = acl_get_entry((struct acl *)tru64_acl)) != NULL) {
s3-smbd: Change allocation of smb_acl_t to talloc() The acl element is changed to be a talloc child, and is no longer one element longer than requested by virtue of the acl[1] base pointer. This also avoids one of the few remaining cases of over-allocation of a structure. Andrew Bartlett
2012-08-12 14:41:35 +04:00
result->acl = talloc_realloc(result, result->acl, struct smb_acl_entry,
result->count + 1);
if (result->acl == NULL) {
TALLOC_FREE(result);
DEBUG(0, ("talloc_realloc failed in tru64_acl_to_smb_acl\n"));
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
errno = ENOMEM;
goto fail;
}
/* XYZ */
if (!tru64_ace_to_smb_ace(entry, &result->acl[result->count])) {
s3-smbd: Change allocation of smb_acl_t to talloc() The acl element is changed to be a talloc child, and is no longer one element longer than requested by virtue of the acl[1] base pointer. This also avoids one of the few remaining cases of over-allocation of a structure. Andrew Bartlett
2012-08-12 14:41:35 +04:00
TALLOC_FREE(result);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
goto fail;
}
result->count += 1;
}
return result;
fail:
s3-smbd: Change allocation of smb_acl_t to talloc() The acl element is changed to be a talloc child, and is no longer one element longer than requested by virtue of the acl[1] base pointer. This also avoids one of the few remaining cases of over-allocation of a structure. Andrew Bartlett
2012-08-12 14:41:35 +04:00
TALLOC_FREE(result);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
DEBUG(1, ("tru64_acl_to_smb_acl failed!\n"));
return NULL;
}
RIP BOOL. Convert BOOL -> bool. I found a few interesting bugs in various places whilst doing this (places that assumed BOOL == int). I also need to fix the Samba4 pidl generation (next checkin). Jeremy. (This used to be commit f35a266b3cbb3e5fa6a86be60f34fe340a3ca71f)
2007-10-19 04:40:25 +04:00
static bool tru64_ace_to_smb_ace(acl_entry_t tru64_ace,
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
struct smb_acl_entry *smb_ace)
{
acl_tag_t tru64_tag;
acl_permset_t permset;
SMB_ACL_TAG_T smb_tag_type;
r22918: Attempt to fix the build of the tru64acl module. Where the heck did that smb_acl_permset_t come from? I can't remember... Michael (This used to be commit e55f952a1319de8743ddfe4993cc4c798b7aa725)
2007-05-16 01:41:02 +04:00
SMB_ACL_PERM_T smb_permset;
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
void *qualifier;
if (acl_get_tag_type(tru64_ace, &tru64_tag) != 0) {
DEBUG(0, ("acl_get_tag_type failed: %s\n", strerror(errno)));
return False;
}
/* On could set the tag type directly to save a function call,
* but I like this better... */
smb_tag_type = tru64_tag_to_smb(tru64_tag);
if (smb_tag_type == 0) {
DEBUG(3, ("invalid tag type given: %d\n", tru64_tag));
return False;
}
if (sys_acl_set_tag_type(smb_ace, smb_tag_type) != 0) {
DEBUG(3, ("sys_acl_set_tag_type failed: %s\n",
strerror(errno)));
return False;
}
qualifier = acl_get_qualifier(tru64_ace);
if (qualifier != NULL) {
if (sys_acl_set_qualifier(smb_ace, qualifier) != 0) {
DEBUG(3, ("sys_acl_set_qualifier failed\n"));
return False;
}
}
if (acl_get_permset(tru64_ace, &permset) != 0) {
DEBUG(3, ("acl_get_permset failed: %s\n", strerror(errno)));
return False;
}
smb_permset = tru64_permset_to_smb(*permset);
if (sys_acl_set_permset(smb_ace, &smb_permset) != 0) {
DEBUG(3, ("sys_acl_set_permset failed: %s\n", strerror(errno)));
return False;
}
return True;
}
static acl_t smb_acl_to_tru64_acl(const SMB_ACL_T smb_acl)
{
acl_t result;
acl_entry_t tru64_entry;
int i;
char *acl_text;
ssize_t acl_text_len;
/* The tru64 acl_init function takes a size_t value
* instead of a count of entries (as with posix).
* the size parameter "Specifies the size of the working
* storage in bytes" (according to the man page).
* But it is unclear to me, how this size is to be
* calculated.
*
* It should not matter, since acl_create_entry enlarges
* the working storage at need. ... */
DEBUG(10, ("Hi! This is smb_acl_to_tru64_acl.\n"));
result = acl_init(1);
if (result == NULL) {
DEBUG(3, ("acl_init failed!\n"));
goto fail;
}
DEBUGADD(10, ("parsing acl entries...\n"));
for (i = 0; i < smb_acl->count; i++) {
/* XYZ - maybe eliminate this direct access? */
const struct smb_acl_entry *smb_entry = &smb_acl->acl[i];
acl_tag_t tru64_tag;
acl_perm_t tru64_permset;
tru64_tag = smb_tag_to_tru64(smb_entry->a_type);
if (tru64_tag == -1) {
DEBUG(3, ("smb_tag_to_tru64 failed!\n"));
goto fail;
}
if (tru64_tag == ACL_MASK) {
DEBUGADD(10, (" - acl type ACL_MASK: not implemented on Tru64 ==> skipping\n"));
continue;
}
tru64_entry = acl_create_entry(&result);
if (tru64_entry == NULL) {
DEBUG(3, ("acl_create_entry failed: %s\n",
strerror(errno)));
goto fail;
}
if (acl_set_tag_type(tru64_entry, tru64_tag) != 0) {
DEBUG(3, ("acl_set_tag_type(%d) failed: %s\n",
strerror(errno)));
goto fail;
}
switch (smb_entry->a_type) {
case SMB_ACL_USER:
if (acl_set_qualifier(tru64_entry,
s3: fix the tru64 acl module This was not adopted to the recent VFS acl structure changes.
2012-09-14 11:40:07 +04:00
(int *)&smb_entry->info.user.uid) != 0)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
DEBUG(3, ("acl_set_qualifier failed: %s\n",
strerror(errno)));
goto fail;
}
s3: fix the tru64 acl module This was not adopted to the recent VFS acl structure changes.
2012-09-14 11:40:07 +04:00
DEBUGADD(10, (" - setting uid to %d\n", smb_entry->info.user.uid));
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
break;
case SMB_ACL_GROUP:
if (acl_set_qualifier(tru64_entry,
s3: fix the tru64 acl module This was not adopted to the recent VFS acl structure changes.
2012-09-14 11:40:07 +04:00
(int *)&smb_entry->info.group.gid) != 0)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
DEBUG(3, ("acl_set_qualifier failed: %s\n",
strerror(errno)));
goto fail;
}
s3: fix the tru64 acl module This was not adopted to the recent VFS acl structure changes.
2012-09-14 11:40:07 +04:00
DEBUGADD(10, (" - setting gid to %d\n", smb_entry->info.group.gid));
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
break;
default:
break;
}
tru64_permset = smb_permset_to_tru64(smb_entry->a_perm);
if (tru64_permset == -1) {
DEBUG(3, ("smb_permset_to_tru64 failed!\n"));
goto fail;
}
DEBUGADD(10, (" - setting perms to %0d\n", tru64_permset));
if (acl_set_permset(tru64_entry, &tru64_permset) != 0)
{
DEBUG(3, ("acl_set_permset failed: %s\n", strerror(errno)));
goto fail;
}
} /* for */
DEBUGADD(10, ("done parsing acl entries\n"));
tru64_entry = NULL;
if (acl_valid(result, &tru64_entry) != 0) {
DEBUG(1, ("smb_acl_to_tru64_acl: ACL is invalid (%s)\n",
strerror(errno)));
if (tru64_entry != NULL) {
DEBUGADD(1, ("the acl contains duplicate entries\n"));
}
goto fail;
}
DEBUGADD(10, ("acl is valid\n"));
acl_text = acl_to_text(result, &acl_text_len);
if (acl_text == NULL) {
DEBUG(3, ("acl_to_text failed: %s\n", strerror(errno)));
goto fail;
}
DEBUG(1, ("acl_text: %s\n", acl_text));
free(acl_text);
return result;
fail:
if (result != NULL) {
acl_free(result);
}
DEBUG(1, ("smb_acl_to_tru64_acl failed!\n"));
return NULL;
}
static acl_tag_t smb_tag_to_tru64(SMB_ACL_TAG_T smb_tag)
{
acl_tag_t result;
switch (smb_tag) {
case SMB_ACL_USER:
result = ACL_USER;
DEBUGADD(10, ("got acl type ACL_USER\n"));
break;
case SMB_ACL_USER_OBJ:
result = ACL_USER_OBJ;
DEBUGADD(10, ("got acl type ACL_USER_OBJ\n"));
break;
case SMB_ACL_GROUP:
result = ACL_GROUP;
DEBUGADD(10, ("got acl type ACL_GROUP\n"));
break;
case SMB_ACL_GROUP_OBJ:
result = ACL_GROUP_OBJ;
DEBUGADD(10, ("got acl type ACL_GROUP_OBJ\n"));
break;
case SMB_ACL_OTHER:
result = ACL_OTHER;
DEBUGADD(10, ("got acl type ACL_OTHER\n"));
break;
case SMB_ACL_MASK:
result = ACL_MASK;
DEBUGADD(10, ("got acl type ACL_MASK\n"));
break;
default:
DEBUG(1, ("Unknown tag type %d\n", smb_tag));
result = -1;
}
return result;
}
static SMB_ACL_TAG_T tru64_tag_to_smb(acl_tag_t tru64_tag)
{
SMB_ACL_TAG_T smb_tag_type;
switch(tru64_tag) {
case ACL_USER:
smb_tag_type = SMB_ACL_USER;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_USER\n"));
break;
case ACL_USER_OBJ:
smb_tag_type = SMB_ACL_USER_OBJ;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_USER_OBJ\n"));
break;
case ACL_GROUP:
smb_tag_type = SMB_ACL_GROUP;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_GROUP\n"));
break;
case ACL_GROUP_OBJ:
smb_tag_type = SMB_ACL_GROUP_OBJ;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_GROUP_OBJ\n"));
break;
case ACL_OTHER:
smb_tag_type = SMB_ACL_OTHER;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_OTHER\n"));
break;
case ACL_MASK:
smb_tag_type = SMB_ACL_MASK;
DEBUGADD(10, ("got smb acl tag type SMB_ACL_MASK\n"));
break;
default:
DEBUG(0, ("Unknown tag type %d\n", (unsigned int)tru64_tag));
smb_tag_type = 0;
}
return smb_tag_type;
}
r22918: Attempt to fix the build of the tru64acl module. Where the heck did that smb_acl_permset_t come from? I can't remember... Michael (This used to be commit e55f952a1319de8743ddfe4993cc4c798b7aa725)
2007-05-16 01:41:02 +04:00
static acl_perm_t smb_permset_to_tru64(SMB_ACL_PERM_T smb_permset)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
/* originally, I thought that acl_clear_perm was the
* proper way to reset the permset to 0. but without
* initializing it to 0, acl_clear_perm fails.
* so probably, acl_clear_perm is not necessary here... ?! */
acl_perm_t tru64_permset = 0;
if (acl_clear_perm(&tru64_permset) != 0) {
DEBUG(5, ("acl_clear_perm failed: %s\n", strerror(errno)));
return -1;
}
/* according to original lib/sysacls.c, acl_add_perm is
* broken on tru64 ... */
tru64_permset |= ((smb_permset & SMB_ACL_READ) ? ACL_READ : 0);
tru64_permset |= ((smb_permset & SMB_ACL_WRITE) ? ACL_WRITE : 0);
tru64_permset |= ((smb_permset & SMB_ACL_EXECUTE) ? ACL_EXECUTE : 0);
return tru64_permset;
}
r22918: Attempt to fix the build of the tru64acl module. Where the heck did that smb_acl_permset_t come from? I can't remember... Michael (This used to be commit e55f952a1319de8743ddfe4993cc4c798b7aa725)
2007-05-16 01:41:02 +04:00
static SMB_ACL_PERM_T tru64_permset_to_smb(const acl_perm_t tru64_permset)
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
{
r22918: Attempt to fix the build of the tru64acl module. Where the heck did that smb_acl_permset_t come from? I can't remember... Michael (This used to be commit e55f952a1319de8743ddfe4993cc4c798b7aa725)
2007-05-16 01:41:02 +04:00
SMB_ACL_PERM_T smb_permset = 0;
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
smb_permset |= ((tru64_permset & ACL_READ) ? SMB_ACL_READ : 0);
smb_permset |= ((tru64_permset & ACL_WRITE) ? SMB_ACL_WRITE : 0);
smb_permset |= ((tru64_permset & ACL_EXECUTE) ? SMB_ACL_EXECUTE : 0);
return smb_permset;
}
/* VFS operations structure */
Make the smbd VFS typesafe
2009-07-24 04:28:58 +04:00
static struct vfs_fn_pointers tru64acl_fns = {
vfs: Make function pointer names consistent. They all end in _fn Autobuild-User: Richard Sharpe <sharpe@samba.org> Autobuild-Date: Mon Dec 12 04:58:40 CET 2011 on sn-devel-104
2011-12-04 08:45:04 +04:00
.sys_acl_get_file_fn = tru64acl_sys_acl_get_file,
.sys_acl_get_fd_fn = tru64acl_sys_acl_get_fd,
vfs: Use posix_sys_acl_blob_get_file in vfs_tru64acl for posix ACLs
2012-10-10 09:59:53 +04:00
.sys_acl_blob_get_file_fn = posix_sys_acl_blob_get_file,
.sys_acl_blob_get_fd_fn = posix_sys_acl_blob_get_fd,
vfs: Make function pointer names consistent. They all end in _fn Autobuild-User: Richard Sharpe <sharpe@samba.org> Autobuild-Date: Mon Dec 12 04:58:40 CET 2011 on sn-devel-104
2011-12-04 08:45:04 +04:00
.sys_acl_set_file_fn = tru64acl_sys_acl_set_file,
.sys_acl_set_fd_fn = tru64acl_sys_acl_set_fd,
.sys_acl_delete_def_file_fn = tru64acl_sys_acl_delete_def_file,
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
};
r20261: merge 20260 from samba_3_0_24 clean up a bunch of no previous prototype warnings (This used to be commit c60687db112405262adf26dbf267804b04074e67)
2006-12-19 23:16:52 +03:00
NTSTATUS vfs_tru64acl_init(void);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
NTSTATUS vfs_tru64acl_init(void)
{
return smb_register_vfs(SMB_VFS_INTERFACE_VERSION, "tru64acl",
Make the smbd VFS typesafe
2009-07-24 04:28:58 +04:00
&tru64acl_fns);
r17179: Merge the vl-posixacls tmp branch into mainline. It modularizes our interface into the special posix API used on the system. Without this patch the specific API flavor is determined at compile time, something which severely limits usability on systems with more than one file system. Our first targets are AIX with its JFS and JFS2 APIs, at a later stage also GPFS. But it's certainly not limited to IBM stuff, this abstraction is also necessary for anything that copes with NFSv4 ACLs. For this we will check in handling very soon. Major contributions can be found in the copyright notices as well as the checkin log of the vl-posixacls branch. The final merge to 3_0 post-3.0.23 was done by Peter Somogyi <psomogyi@gamax.hu> (This used to be commit ca0c73f281a2a65a988094a46bb3e46a94011a53)
2006-07-21 19:51:34 +04:00
}
/* ENTE */
Copy Permalink