emo/shadow
1
0
Fork 0
Code Issues Pull Requests Packages Releases Activity
dfc93b35cc
shadow/libmisc/find_new_uid.c

484 lines
13 KiB
C
Raw Normal View History

Make sure every source files are distributed with a copyright and license. Files with no license use the default 3-clauses BSD license. The copyright were mostly not recorded; they were updated according to the Changelog. "Julianne Frances Haugh and contributors" changed to "copyright holders and contributors".
2008-04-27 06:10:09 +05:30
/*
Update licensing info Closes #238 Update all files to list SPDX license shortname. Most files are BSD 3 clause license. The exceptions are: serge@sl ~/src/shadow$ git grep SPDX-License | grep -v BSD-3-Clause contrib/atudel:# SPDX-License-Identifier: BSD-4-Clause lib/tcbfuncs.c: * SPDX-License-Identifier: 0BSD libmisc/salt.c: * SPDX-License-Identifier: Unlicense src/login_nopam.c: * SPDX-License-Identifier: Unlicense src/nologin.c: * SPDX-License-Identifier: BSD-2-Clause src/vipw.c: * SPDX-License-Identifier: GPL-2.0-or-later Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-12-05 21:05:27 +05:30
* SPDX-FileCopyrightText: 1991 - 1994, Julianne Frances Haugh
* SPDX-FileCopyrightText: 2008 - 2011, Nicolas François
* SPDX-FileCopyrightText: 2014, Red Hat, Inc.
Make sure every source files are distributed with a copyright and license. Files with no license use the default 3-clauses BSD license. The copyright were mostly not recorded; they were updated according to the Changelog. "Julianne Frances Haugh and contributors" changed to "copyright holders and contributors".
2008-04-27 06:10:09 +05:30
*
Update licensing info Closes #238 Update all files to list SPDX license shortname. Most files are BSD 3 clause license. The exceptions are: serge@sl ~/src/shadow$ git grep SPDX-License | grep -v BSD-3-Clause contrib/atudel:# SPDX-License-Identifier: BSD-4-Clause lib/tcbfuncs.c: * SPDX-License-Identifier: 0BSD libmisc/salt.c: * SPDX-License-Identifier: Unlicense src/login_nopam.c: * SPDX-License-Identifier: Unlicense src/nologin.c: * SPDX-License-Identifier: BSD-2-Clause src/vipw.c: * SPDX-License-Identifier: GPL-2.0-or-later Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-12-05 21:05:27 +05:30
* SPDX-License-Identifier: BSD-3-Clause
Make sure every source files are distributed with a copyright and license. Files with no license use the default 3-clauses BSD license. The copyright were mostly not recorded; they were updated according to the Changelog. "Julianne Frances Haugh and contributors" changed to "copyright holders and contributors".
2008-04-27 06:10:09 +05:30
*/
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
#include <config.h>
#include <assert.h>
#include <stdio.h>
* libmisc/find_new_uid.c, libmisc/find_new_gid.c: Add missing memory allocation check.
2011-06-03 00:10:06 +05:30
#include <errno.h>
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
#include "prototypes.h"
#include "pwio.h"
#include "getdef.h"
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
#include "shadowlog.h"
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* get_ranges - Get the minimum and maximum ID ranges for the search
*
* This function will return the minimum and maximum ranges for IDs
*
* 0: The function completed successfully
* EINVAL: The provided ranges are impossible (such as maximum < minimum)
*
* preferred_min: The special-case minimum value for a specifically-
* requested ID, which may be lower than the standard min_id
*/
static int get_ranges (bool sys_user, uid_t *min_id, uid_t *max_id,
uid_t *preferred_min)
{
uid_t uid_def_max = 0;
if (sys_user) {
/* System users */
/* A requested ID is allowed to be below the autoselect range */
*preferred_min = (uid_t) 1;
/* Get the minimum ID range from login.defs or default to 101 */
*min_id = (uid_t) getdef_ulong ("SYS_UID_MIN", 101UL);
/*
* If SYS_UID_MAX is unspecified, we should assume it to be one
* less than the UID_MIN (which is reserved for non-system accounts)
*/
uid_def_max = (uid_t) getdef_ulong ("UID_MIN", 1000UL) - 1;
*max_id = (uid_t) getdef_ulong ("SYS_UID_MAX",
(unsigned long) uid_def_max);
/* Check that the ranges make sense */
if (*max_id < *min_id) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
(void) fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Invalid configuration: SYS_UID_MIN (%lu), "
"UID_MIN (%lu), SYS_UID_MAX (%lu)\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), (unsigned long) *min_id,
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
getdef_ulong ("UID_MIN", 1000UL),
(unsigned long) *max_id);
return EINVAL;
}
libmisc: minimum id check for system accounts The minimum id allocation for system accounts shouldn't be 0 as this is reserved for root. Signed-off-by: Tomáš Mráz <tm@t8m.info> Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
2022-10-06 14:51:18 +05:30
/*
* Zero is reserved for root and the allocation algorithm does not
* work right with it.
*/
if (*min_id == 0) {
*min_id = (uid_t) 1;
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
} else {
/* Non-system users */
/* Get the values from login.defs or use reasonable defaults */
*min_id = (uid_t) getdef_ulong ("UID_MIN", 1000UL);
*max_id = (uid_t) getdef_ulong ("UID_MAX", 60000UL);
/*
* The preferred minimum should match the standard ID minimum
* for non-system users.
*/
*preferred_min = *min_id;
/* Check that the ranges make sense */
if (*max_id < *min_id) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
(void) fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Invalid configuration: UID_MIN (%lu), "
"UID_MAX (%lu)\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), (unsigned long) *min_id,
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
(unsigned long) *max_id);
return EINVAL;
}
}
return 0;
}
/*
* check_uid - See if the requested UID is available
*
* On success, return 0
* If the ID is in use, return EEXIST
* If the ID is outside the range, return ERANGE
* In other cases, return errno from getpwuid()
*/
static int check_uid(const uid_t uid,
const uid_t uid_min,
const uid_t uid_max,
Declare read-only parameters const Signal callers arguments are not going to be modified and allow passing const pointers.
2022-08-05 21:10:26 +05:30
const bool *used_uids)
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
{
/* First test that the preferred ID is in the range */
if (uid < uid_min || uid > uid_max) {
return ERANGE;
}
/*
* Check whether we already detected this UID
* using the pw_next() loop
*/
if (used_uids != NULL && used_uids[uid]) {
return EEXIST;
}
/* Check if the UID exists according to NSS */
errno = 0;
add --prefix option
2016-05-15 19:19:39 +05:30
if (prefix_getpwuid(uid) != NULL) {
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
return EEXIST;
} else {
/* getpwuid() was NULL
* we have to ignore errors as temporary
* failures of remote user identity services
* would completely block user/group creation
*/
}
/* If we've made it here, the UID must be available */
return 0;
}
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
/*
* find_new_uid - Find a new unused UID.
*
* If successful, find_new_uid provides an unused user ID in the
* [UID_MIN:UID_MAX] range.
* This ID should be higher than all the used UID, but if not possible,
* the lowest unused ID in the range will be returned.
fix spelling and unify whitespace
2021-08-18 23:36:02 +05:30
*
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
* Return 0 on success, -1 if no unused UIDs are available.
*/
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
int find_new_uid(bool sys_user,
uid_t *uid,
/*@null@*/uid_t const *preferred_uid)
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
{
* libmisc/find_new_gid.c: Use booleans instead of char fo used_gids. * libmisc/find_new_gid.c: Use getdef_ulong and cast to git_t to get GID values. * libmisc/find_new_gid.c: Use UL as a prefix for ulong values. * libmisc/find_new_uid.c: Likewise.
2009-04-23 16:46:38 +05:30
bool *used_uids;
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
const struct passwd *pwd;
uid_t uid_min, uid_max, preferred_min;
remove unused variables parent, user_id, and group_id are unused. Signed-off-by: Serge Hallyn <shallyn@cisco.com>
2019-10-13 06:27:12 +05:30
uid_t id;
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
uid_t lowest_found, highest_found;
int result;
int nospam = 0;
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
assert (uid != NULL);
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* First, figure out what ID range is appropriate for
* automatic assignment
*/
result = get_ranges (sys_user, &uid_min, &uid_max, &preferred_min);
if (result == EINVAL) {
return -1;
}
/* Check if the preferred UID is available */
if (preferred_uid) {
result = check_uid (*preferred_uid, preferred_min, uid_max, NULL);
if (result == 0) {
/*
* Make sure the UID isn't queued for use already
*/
if (pw_locate_uid (*preferred_uid) == NULL) {
*uid = *preferred_uid;
return 0;
}
/*
* pw_locate_uid() found the UID in an as-yet uncommitted
* entry. We'll proceed below and auto-set an UID.
*/
} else if (result == EEXIST || result == ERANGE) {
/*
* Continue on below. At this time, we won't
* treat these two cases differently.
*/
} else {
/*
* An unexpected error occurred. We should report
* this and fail the user creation.
* This differs from the automatic creation
* behavior below, since if a specific UID was
* requested and generated an error, the user is
* more likely to want to stop and address the
* issue.
*/
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Encountered error attempting to use "
"preferred UID: %s\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (result));
* libmisc/find_new_gid.c, libmisc/find_new_uid.c: Fail in case of invalid configuration. * libmisc/find_new_gid.c, libmisc/find_new_uid.c: Updated comments. * libmisc/find_new_gid.c, libmisc/find_new_uid.c: Be more strict on the loop stop conditions. Stop if we passed the limit, even if the limit itself was never noticed.
2011-07-23 03:23:01 +05:30
return -1;
* libmisc/find_new_gid.c, libmisc/find_new_uid.c: Detect some misconfigurations of UID_MIN, UID_MAX, SYS_UID_MIN, SYS_UID_MAX (and the GID equivalents)
2010-03-21 06:50:50 +05:30
}
Added new option -r, --system for system accounts in useradd, groupadd, and newusers.
2008-02-20 02:31:38 +05:30
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* Search the entire passwd file,
* looking for the next unused value.
*
* We first check the local database with pw_rewind/pw_next to find
* all local values that are in use.
*
* We then compare the next free value to all databases (local and
* remote) and iterate until we find a free one. If there are free
* values beyond the lowest (system users) or highest (non-system
* users), we will prefer those and avoid potentially reclaiming a
* deleted user (which can be a security issue, since it may grant
* access to files belonging to that former user).
*
* If there are no UIDs available at the end of the search, we will
* have no choice but to iterate through the range looking for gaps.
*
*/
/* Create an array to hold all of the discovered UIDs */
* libmisc/find_new_uid.c, libmisc/find_new_gid.c: Add missing memory allocation check.
2011-06-03 00:10:06 +05:30
used_uids = malloc (sizeof (bool) * (uid_max +1));
if (NULL == used_uids) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: failed to allocate memory: %s\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (errno));
* libmisc/find_new_uid.c, libmisc/find_new_gid.c: Add missing memory allocation check.
2011-06-03 00:10:06 +05:30
return -1;
}
* libmisc/find_new_gid.c: Use booleans instead of char fo used_gids. * libmisc/find_new_gid.c: Use getdef_ulong and cast to git_t to get GID values. * libmisc/find_new_gid.c: Use UL as a prefix for ulong values. * libmisc/find_new_uid.c: Likewise.
2009-04-23 16:46:38 +05:30
memset (used_uids, false, sizeof (bool) * (uid_max + 1));
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/* First look for the lowest and highest value in the local database */
(void) pw_rewind ();
highest_found = uid_min;
lowest_found = uid_max;
while ((pwd = pw_next ()) != NULL) {
/*
* Does this entry have a lower UID than the lowest we've found
* so far?
*/
if ((pwd->pw_uid <= lowest_found) && (pwd->pw_uid >= uid_min)) {
lowest_found = pwd->pw_uid - 1;
}
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* Does this entry have a higher UID than the highest we've found
* so far?
*/
if ((pwd->pw_uid >= highest_found) && (pwd->pw_uid <= uid_max)) {
highest_found = pwd->pw_uid + 1;
}
/* create index of used UIDs */
if (pwd->pw_uid >= uid_min
&& pwd->pw_uid <= uid_max) {
used_uids[pwd->pw_uid] = true;
}
}
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
* NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Do not use getpwent / getgrent for system accounts. Trying the low-IDs with getpwuid / getgrgid should be more efficient on LDAP configured systems with many accounts.
2009-07-18 04:24:23 +05:30
if (sys_user) {
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* For system users, we want to start from the
* top of the range and work downwards.
* NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Do not use getpwent / getgrent for system accounts. Trying the low-IDs with getpwuid / getgrgid should be more efficient on LDAP configured systems with many accounts.
2009-07-18 04:24:23 +05:30
*/
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* At the conclusion of the pw_next() search, we will either
* have a presumed-free UID or we will be at UID_MIN - 1.
*/
if (lowest_found < uid_min) {
/*
* In this case, an UID is in use at UID_MIN.
*
* We will reset the search to UID_MAX and proceed down
* through all the UIDs (skipping those we detected with
* used_uids) for a free one. It is a known issue that
* this may result in reusing a previously-deleted UID,
* so administrators should be instructed to use this
* auto-detection with care (and prefer to assign UIDs
* explicitly).
*/
lowest_found = uid_max;
2009-07-18 Peter Vrabec <pvrabec@redhat.com> * NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Since system accounts are allocated from SYS_?ID_MIN to SYS_?ID_MAX in reverse order, accounts are packed close to SYS_?ID_MAX if SYS_?ID_MIN is already used but there are still dome gaps.
2009-07-18 06:05:35 +05:30
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/* Search through all of the IDs in the range */
for (id = lowest_found; id >= uid_min; id--) {
result = check_uid (id, uid_min, uid_max, used_uids);
if (result == 0) {
/* This UID is available. Return it. */
*uid = id;
free (used_uids);
return 0;
} else if (result == EEXIST) {
/* This UID is in use, we'll continue to the next */
} else {
/*
* An unexpected error occurred.
*
* Only report it the first time to avoid spamming
* the logs
*
*/
if (!nospam) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Can't get unique system UID (%s). "
"Suppressing additional messages.\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (result));
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
SYSLOG ((LOG_ERR,
"Error checking available UIDs: %s",
strerror (result)));
nospam = 1;
}
/*
* We will continue anyway. Hopefully a later UID
* will work properly.
*/
* NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Do not use getpwent / getgrent for system accounts. Trying the low-IDs with getpwuid / getgrgid should be more efficient on LDAP configured systems with many accounts.
2009-07-18 04:24:23 +05:30
}
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* If we get all the way through the loop, try again from UID_MAX,
* unless that was where we previously started. (NOTE: the worst-case
* scenario here is that we will run through (UID_MAX - UID_MIN - 1)
* cycles *again* if we fall into this case with lowest_found as
* UID_MAX - 1, all users in the range in use and maintained by
* network services such as LDAP.)
*/
if (lowest_found != uid_max) {
for (id = uid_max; id >= uid_min; id--) {
result = check_uid (id, uid_min, uid_max, used_uids);
if (result == 0) {
/* This UID is available. Return it. */
*uid = id;
free (used_uids);
return 0;
} else if (result == EEXIST) {
/* This UID is in use, we'll continue to the next */
} else {
/*
* An unexpected error occurred.
*
* Only report it the first time to avoid spamming
* the logs
*
*/
if (!nospam) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Can't get unique system UID (%s). "
"Suppressing additional messages.\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (result));
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
SYSLOG((LOG_ERR,
"Error checking available UIDs: %s",
strerror (result)));
nospam = 1;
}
/*
* We will continue anyway. Hopefully a later UID
* will work properly.
*/
}
* NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Do not use getpwent / getgrent for system accounts. Trying the low-IDs with getpwuid / getgrgid should be more efficient on LDAP configured systems with many accounts.
2009-07-18 04:24:23 +05:30
}
From RedHat's patch shadow-4.1.2-sysAccountDownhill.patch Thanks to Peter Vrabec. * NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Build an index of used IDs to avoid a database request for each id in the allowed range (when the highest allowed ID is already used). This speedups the addition of users or groups when the highest allowed ID is already used. The additional memory usage of the tools should be acceptable when UID_MAX/SYS_UID_MAX are set to a reasonable number.
2008-09-13 17:24:49 +05:30
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
} else { /* !sys_user */
/*
* For non-system users, we want to start from the
* bottom of the range and work upwards.
*/
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* At the conclusion of the pw_next() search, we will either
* have a presumed-free UID or we will be at UID_MAX + 1.
*/
if (highest_found > uid_max) {
/*
* In this case, a UID is in use at UID_MAX.
*
* We will reset the search to UID_MIN and proceed up
* through all the UIDs (skipping those we detected with
* used_uids) for a free one. It is a known issue that
* this may result in reusing a previously-deleted UID,
* so administrators should be instructed to use this
* auto-detection with care (and prefer to assign UIDs
* explicitly).
*/
highest_found = uid_min;
* libmisc/find_new_gid.c, libmisc/find_new_uid.c: For system accounts, return the first unused ID, starting from the max value. This could be useful later to increase the static IDs range.
2009-04-11 21:30:45 +05:30
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/* Search through all of the IDs in the range */
for (id = highest_found; id <= uid_max; id++) {
result = check_uid (id, uid_min, uid_max, used_uids);
if (result == 0) {
/* This UID is available. Return it. */
*uid = id;
* libmisc/find_new_uid.c: free (used_uids) on return. * libmisc/find_new_gid.c: free (used_gids) on return.
2011-07-30 07:08:07 +05:30
free (used_uids);
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
return 0;
} else if (result == EEXIST) {
/* This UID is in use, we'll continue to the next */
} else {
/*
* An unexpected error occurred.
*
* Only report it the first time to avoid spamming
* the logs
*
*/
if (!nospam) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Can't get unique UID (%s). "
"Suppressing additional messages.\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (result));
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
SYSLOG ((LOG_ERR,
"Error checking available UIDs: %s",
strerror (result)));
nospam = 1;
}
/*
* We will continue anyway. Hopefully a later UID
* will work properly.
*/
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
}
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/*
* If we get all the way through the loop, try again from UID_MIN,
* unless that was where we previously started. (NOTE: the worst-case
* scenario here is that we will run through (UID_MAX - UID_MIN - 1)
* cycles *again* if we fall into this case with highest_found as
* UID_MIN + 1, all users in the range in use and maintained by
* network services such as LDAP.)
*/
if (highest_found != uid_min) {
for (id = uid_min; id <= uid_max; id++) {
result = check_uid (id, uid_min, uid_max, used_uids);
if (result == 0) {
/* This UID is available. Return it. */
*uid = id;
free (used_uids);
return 0;
} else if (result == EEXIST) {
/* This UID is in use, we'll continue to the next */
} else {
/*
* An unexpected error occurred.
*
* Only report it the first time to avoid spamming
* the logs
*
*/
if (!nospam) {
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Can't get unique UID (%s). "
"Suppressing additional messages.\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname(), strerror (result));
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
SYSLOG ((LOG_ERR,
"Error checking available UIDs: %s",
strerror (result)));
nospam = 1;
}
/*
* We will continue anyway. Hopefully a later UID
* will work properly.
*/
2009-07-18 Peter Vrabec <pvrabec@redhat.com> * NEWS, libmisc/find_new_gid.c, libmisc/find_new_uid.c: Since system accounts are allocated from SYS_?ID_MIN to SYS_?ID_MAX in reverse order, accounts are packed close to SYS_?ID_MAX if SYS_?ID_MIN is already used but there are still dome gaps.
2009-07-18 06:05:35 +05:30
}
}
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
}
}
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
/* The code reached here and found no available IDs in the range */
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
fprintf (log_get_logfd(),
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
_("%s: Can't get unique UID (no more available UIDs)\n"),
Make shadow_logfd and Prog not extern Closes #444 Closes #465 Signed-off-by: Serge Hallyn <serge@hallyn.com>
2021-11-29 05:07:53 +05:30
log_get_progname());
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
SYSLOG ((LOG_WARN, "no more available UIDs on the system"));
* libmisc/find_new_uid.c, libmisc/find_new_gid.c: Add missing memory allocation check.
2011-06-03 00:10:06 +05:30
free (used_uids);
Make the group and user allocation more effective. Previously, the allocation was optimized for an outdated deployment style (that of /etc/group alongside nss_db). The issue here is that this results in extremely poor performance when using SSSD, Winbind or nss_ldap. There were actually two serious bugs here that have been addressed: 1) Running getgrent() loops won't work in most SSSD or Winbind environments, as full group enumeration is disabled by default. This could easily result in auto-allocating a group that was already in use. (This might result in a security issue as well, if the shared GID is a privileged group). 2) For system groups, the loop was always iterating through the complete SYS_GID_MIN->SYS_GID_MAX range. On SSSD and Winbind, this means hundreds of round-trips to LDAP (unless the GIDs were specifically configured to be ignored by the SSSD or winbindd). To a user with a slow connection to their LDAP server, this would appear as if groupadd -r was hung. (Though it would eventually complete). This patch changes the algorithm to be more favorable for LDAP environments, at the expense of some performance when using nss_db. Given that the DB is a local service, this should have a negligible effect from a user's perspective. With the new algorithm, we simply first iterate through all entries in the local database with gr_next(), recording the IDs that are in use. We then start from the highest presumed-available entry and call getgrgid() to see if it is available. We continue this until we come to the first unused GID. We then select that and return it. If we make it through all the remaining IDs without finding a free one, we start over from the beginning of the range and try to find room in one of the gaps in the range. The patch was originally written by Stephen Gallagher and applied identically also to the user allocation by Tomáš Mráz. Signed-off-by: Serge Hallyn <serge@hallyn.com>
2016-11-15 20:13:54 +05:30
return -1;
Add new generic functions to find the next user or group ID available: find_new_uid() and find_new_gid(). They work the same way as the functions with the same name of useradd or groupadd, except that they check in the local database to make sure an ID was not reserved in an uncommitted change (this is needed to be used in newusers), they report a status instead of calling exit(), and they can receive a preferred ID. They should later support system IDs. This should be a little bit slower, but not too much (if the database is not open the checks against the local database will exit immediately, and if it is already open, all the checks will be done regarding the data in memory).
2008-02-03 22:23:07 +05:30
}
Reference in New Issue Copy Permalink