2019-06-01 08:08:55 +00:00
|
|
|
// SPDX-License-Identifier: GPL-2.0-only
|
2009-02-04 14:06:58 +00:00
|
|
|
/*
|
2018-12-09 20:36:33 +00:00
|
|
|
* Integrity Measurement Architecture
|
|
|
|
*
|
2009-02-04 14:06:58 +00:00
|
|
|
* Copyright (C) 2005,2006,2007,2008 IBM Corporation
|
|
|
|
*
|
|
|
|
* Authors:
|
|
|
|
* Reiner Sailer <sailer@watson.ibm.com>
|
|
|
|
* Serge Hallyn <serue@us.ibm.com>
|
|
|
|
* Kylene Hall <kylene@us.ibm.com>
|
|
|
|
* Mimi Zohar <zohar@us.ibm.com>
|
|
|
|
*
|
|
|
|
* File: ima_main.c
|
2009-12-04 20:48:00 +00:00
|
|
|
* implements the IMA hooks: ima_bprm_check, ima_file_mmap,
|
2010-01-26 22:02:40 +00:00
|
|
|
* and ima_file_check.
|
2009-02-04 14:06:58 +00:00
|
|
|
*/
|
2018-03-23 13:41:08 +00:00
|
|
|
|
2018-12-14 21:48:07 +00:00
|
|
|
#include <linux/module.h>
|
2009-02-04 14:06:58 +00:00
|
|
|
#include <linux/file.h>
|
|
|
|
#include <linux/binfmts.h>
|
2020-10-02 17:38:15 +00:00
|
|
|
#include <linux/kernel_read_file.h>
|
2009-02-04 14:06:58 +00:00
|
|
|
#include <linux/mount.h>
|
|
|
|
#include <linux/mman.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
|
|
|
#include <linux/slab.h>
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
#include <linux/xattr.h>
|
2011-08-30 00:19:50 +00:00
|
|
|
#include <linux/ima.h>
|
2017-12-11 11:35:21 +00:00
|
|
|
#include <linux/iversion.h>
|
2018-02-21 16:35:20 +00:00
|
|
|
#include <linux/fs.h>
|
2009-02-04 14:06:58 +00:00
|
|
|
|
|
|
|
#include "ima.h"
|
|
|
|
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
#ifdef CONFIG_IMA_APPRAISE
|
|
|
|
int ima_appraise = IMA_APPRAISE_ENFORCE;
|
|
|
|
#else
|
|
|
|
int ima_appraise;
|
|
|
|
#endif
|
|
|
|
|
2013-04-25 07:43:56 +00:00
|
|
|
int ima_hash_algo = HASH_ALGO_SHA1;
|
2013-06-07 10:16:37 +00:00
|
|
|
static int hash_setup_done;
|
2013-04-25 07:43:56 +00:00
|
|
|
|
2019-06-14 12:20:15 +00:00
|
|
|
static struct notifier_block ima_lsm_policy_notifier = {
|
|
|
|
.notifier_call = ima_lsm_policy_change,
|
|
|
|
};
|
|
|
|
|
2009-02-04 14:06:58 +00:00
|
|
|
static int __init hash_setup(char *str)
|
|
|
|
{
|
2013-06-07 10:16:37 +00:00
|
|
|
struct ima_template_desc *template_desc = ima_template_desc_current();
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (hash_setup_done)
|
|
|
|
return 1;
|
|
|
|
|
|
|
|
if (strcmp(template_desc->name, IMA_TEMPLATE_IMA_NAME) == 0) {
|
2020-09-04 19:40:58 +00:00
|
|
|
if (strncmp(str, "sha1", 4) == 0) {
|
2013-06-07 10:16:37 +00:00
|
|
|
ima_hash_algo = HASH_ALGO_SHA1;
|
2020-09-04 19:40:58 +00:00
|
|
|
} else if (strncmp(str, "md5", 3) == 0) {
|
2013-06-07 10:16:37 +00:00
|
|
|
ima_hash_algo = HASH_ALGO_MD5;
|
2020-09-04 19:40:58 +00:00
|
|
|
} else {
|
|
|
|
pr_err("invalid hash algorithm \"%s\" for template \"%s\"",
|
|
|
|
str, IMA_TEMPLATE_IMA_NAME);
|
2017-10-20 08:01:03 +00:00
|
|
|
return 1;
|
2020-09-04 19:40:58 +00:00
|
|
|
}
|
2013-06-07 10:16:37 +00:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2018-05-21 11:58:02 +00:00
|
|
|
i = match_string(hash_algo_name, HASH_ALGO__LAST, str);
|
2020-09-04 19:40:58 +00:00
|
|
|
if (i < 0) {
|
|
|
|
pr_err("invalid hash algorithm \"%s\"", str);
|
2017-10-20 08:01:03 +00:00
|
|
|
return 1;
|
2020-09-04 19:40:58 +00:00
|
|
|
}
|
2018-05-21 11:58:02 +00:00
|
|
|
|
|
|
|
ima_hash_algo = i;
|
2013-06-07 10:16:37 +00:00
|
|
|
out:
|
|
|
|
hash_setup_done = 1;
|
2009-02-04 14:06:58 +00:00
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
__setup("ima_hash=", hash_setup);
|
|
|
|
|
2019-04-30 12:34:44 +00:00
|
|
|
/* Prevent mmap'ing a file execute that is already mmap'ed write */
|
|
|
|
static int mmap_violation_check(enum ima_hooks func, struct file *file,
|
|
|
|
char **pathbuf, const char **pathname,
|
|
|
|
char *filename)
|
|
|
|
{
|
|
|
|
struct inode *inode;
|
|
|
|
int rc = 0;
|
|
|
|
|
|
|
|
if ((func == MMAP_CHECK) && mapping_writably_mapped(file->f_mapping)) {
|
|
|
|
rc = -ETXTBSY;
|
|
|
|
inode = file_inode(file);
|
|
|
|
|
|
|
|
if (!*pathbuf) /* ima_rdwr_violation possibly pre-fetched */
|
|
|
|
*pathname = ima_d_path(&file->f_path, pathbuf,
|
|
|
|
filename);
|
|
|
|
integrity_audit_msg(AUDIT_INTEGRITY_DATA, inode, *pathname,
|
|
|
|
"mmap_file", "mmapped_writers", rc, 0);
|
|
|
|
}
|
|
|
|
return rc;
|
|
|
|
}
|
|
|
|
|
2010-01-20 20:35:41 +00:00
|
|
|
/*
|
2010-11-02 14:13:07 +00:00
|
|
|
* ima_rdwr_violation_check
|
2010-01-20 20:35:41 +00:00
|
|
|
*
|
2010-11-02 14:13:07 +00:00
|
|
|
* Only invalidate the PCR for measured files:
|
2014-03-04 16:04:20 +00:00
|
|
|
* - Opening a file for write when already open for read,
|
2010-01-20 20:35:41 +00:00
|
|
|
* results in a time of measure, time of use (ToMToU) error.
|
|
|
|
* - Opening a file for read when already open for write,
|
2014-03-04 16:04:20 +00:00
|
|
|
* could result in a file measurement error.
|
2010-01-20 20:35:41 +00:00
|
|
|
*
|
|
|
|
*/
|
2014-09-12 17:35:55 +00:00
|
|
|
static void ima_rdwr_violation_check(struct file *file,
|
|
|
|
struct integrity_iint_cache *iint,
|
2014-09-12 17:35:56 +00:00
|
|
|
int must_measure,
|
2014-09-12 17:35:55 +00:00
|
|
|
char **pathbuf,
|
2017-11-30 10:56:02 +00:00
|
|
|
const char **pathname,
|
|
|
|
char *filename)
|
2010-01-20 20:35:41 +00:00
|
|
|
{
|
2013-06-13 22:37:49 +00:00
|
|
|
struct inode *inode = file_inode(file);
|
2010-01-20 20:35:41 +00:00
|
|
|
fmode_t mode = file->f_mode;
|
2010-10-25 18:41:45 +00:00
|
|
|
bool send_tomtou = false, send_writers = false;
|
2010-10-25 18:41:59 +00:00
|
|
|
|
2010-01-20 20:35:41 +00:00
|
|
|
if (mode & FMODE_WRITE) {
|
ima: check inode integrity cache in violation check
When IMA did not support ima-appraisal, existance of the S_IMA flag
clearly indicated that the file was measured. With IMA appraisal S_IMA
flag indicates that file was measured and/or appraised. Because of
this, when measurement is not enabled by the policy, violations are
still reported.
To differentiate between measurement and appraisal policies this
patch checks the inode integrity cache flags. The IMA_MEASURED
flag indicates whether the file was actually measured, while the
IMA_MEASURE flag indicates whether the file should be measured.
Unfortunately, the IMA_MEASURED flag is reset to indicate the file
needs to be re-measured. Thus, this patch checks the IMA_MEASURE
flag.
This patch limits the false positive violation reports, but does
not fix it entirely. The IMA_MEASURE/IMA_MEASURED flags are
indications that, at some point in time, the file opened for read
was in policy, but might not be in policy now (eg. different uid).
Other changes would be needed to further limit false positive
violation reports.
Changelog:
- expanded patch description based on conversation with Roberto (Mimi)
Signed-off-by: Dmitry Kasatkin <d.kasatkin@samsung.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2014-03-27 08:29:28 +00:00
|
|
|
if (atomic_read(&inode->i_readcount) && IS_IMA(inode)) {
|
2014-09-12 17:35:55 +00:00
|
|
|
if (!iint)
|
|
|
|
iint = integrity_iint_find(inode);
|
ima: check inode integrity cache in violation check
When IMA did not support ima-appraisal, existance of the S_IMA flag
clearly indicated that the file was measured. With IMA appraisal S_IMA
flag indicates that file was measured and/or appraised. Because of
this, when measurement is not enabled by the policy, violations are
still reported.
To differentiate between measurement and appraisal policies this
patch checks the inode integrity cache flags. The IMA_MEASURED
flag indicates whether the file was actually measured, while the
IMA_MEASURE flag indicates whether the file should be measured.
Unfortunately, the IMA_MEASURED flag is reset to indicate the file
needs to be re-measured. Thus, this patch checks the IMA_MEASURE
flag.
This patch limits the false positive violation reports, but does
not fix it entirely. The IMA_MEASURE/IMA_MEASURED flags are
indications that, at some point in time, the file opened for read
was in policy, but might not be in policy now (eg. different uid).
Other changes would be needed to further limit false positive
violation reports.
Changelog:
- expanded patch description based on conversation with Roberto (Mimi)
Signed-off-by: Dmitry Kasatkin <d.kasatkin@samsung.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2014-03-27 08:29:28 +00:00
|
|
|
/* IMA_MEASURE is set from reader side */
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (iint && test_bit(IMA_MUST_MEASURE,
|
|
|
|
&iint->atomic_flags))
|
ima: check inode integrity cache in violation check
When IMA did not support ima-appraisal, existance of the S_IMA flag
clearly indicated that the file was measured. With IMA appraisal S_IMA
flag indicates that file was measured and/or appraised. Because of
this, when measurement is not enabled by the policy, violations are
still reported.
To differentiate between measurement and appraisal policies this
patch checks the inode integrity cache flags. The IMA_MEASURED
flag indicates whether the file was actually measured, while the
IMA_MEASURE flag indicates whether the file should be measured.
Unfortunately, the IMA_MEASURED flag is reset to indicate the file
needs to be re-measured. Thus, this patch checks the IMA_MEASURE
flag.
This patch limits the false positive violation reports, but does
not fix it entirely. The IMA_MEASURE/IMA_MEASURED flags are
indications that, at some point in time, the file opened for read
was in policy, but might not be in policy now (eg. different uid).
Other changes would be needed to further limit false positive
violation reports.
Changelog:
- expanded patch description based on conversation with Roberto (Mimi)
Signed-off-by: Dmitry Kasatkin <d.kasatkin@samsung.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2014-03-27 08:29:28 +00:00
|
|
|
send_tomtou = true;
|
|
|
|
}
|
2014-03-27 08:54:11 +00:00
|
|
|
} else {
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (must_measure)
|
|
|
|
set_bit(IMA_MUST_MEASURE, &iint->atomic_flags);
|
2018-12-11 08:31:40 +00:00
|
|
|
if (inode_is_open_for_write(inode) && must_measure)
|
2014-03-27 08:54:11 +00:00
|
|
|
send_writers = true;
|
2010-01-20 20:35:41 +00:00
|
|
|
}
|
2010-10-25 18:41:45 +00:00
|
|
|
|
2012-06-20 13:32:55 +00:00
|
|
|
if (!send_tomtou && !send_writers)
|
|
|
|
return;
|
|
|
|
|
2017-01-17 11:45:41 +00:00
|
|
|
*pathname = ima_d_path(&file->f_path, pathbuf, filename);
|
2012-09-03 21:40:17 +00:00
|
|
|
|
2010-10-25 18:41:45 +00:00
|
|
|
if (send_tomtou)
|
2015-04-11 15:12:39 +00:00
|
|
|
ima_add_violation(file, *pathname, iint,
|
|
|
|
"invalid_pcr", "ToMToU");
|
2010-10-25 18:41:45 +00:00
|
|
|
if (send_writers)
|
2015-04-11 15:12:39 +00:00
|
|
|
ima_add_violation(file, *pathname, iint,
|
2012-06-20 13:32:55 +00:00
|
|
|
"invalid_pcr", "open_writers");
|
2010-01-20 20:35:41 +00:00
|
|
|
}
|
|
|
|
|
2011-03-09 19:13:22 +00:00
|
|
|
static void ima_check_last_writer(struct integrity_iint_cache *iint,
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
struct inode *inode, struct file *file)
|
2010-10-25 18:42:05 +00:00
|
|
|
{
|
2011-07-26 08:30:35 +00:00
|
|
|
fmode_t mode = file->f_mode;
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
bool update;
|
2010-10-25 18:42:05 +00:00
|
|
|
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
if (!(mode & FMODE_WRITE))
|
|
|
|
return;
|
|
|
|
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
mutex_lock(&iint->mutex);
|
2014-06-27 15:04:27 +00:00
|
|
|
if (atomic_read(&inode->i_writecount) == 1) {
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
update = test_and_clear_bit(IMA_UPDATE_XATTR,
|
|
|
|
&iint->atomic_flags);
|
2017-12-11 11:35:20 +00:00
|
|
|
if (!IS_I_VERSION(inode) ||
|
2018-02-01 13:15:25 +00:00
|
|
|
!inode_eq_iversion(inode, iint->version) ||
|
2014-06-27 15:04:27 +00:00
|
|
|
(iint->flags & IMA_NEW_FILE)) {
|
|
|
|
iint->flags &= ~(IMA_DONE_MASK | IMA_NEW_FILE);
|
2016-06-01 18:14:06 +00:00
|
|
|
iint->measured_pcrs = 0;
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (update)
|
2014-06-27 15:04:27 +00:00
|
|
|
ima_update_xattr(iint, file);
|
|
|
|
}
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
}
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
mutex_unlock(&iint->mutex);
|
2010-10-25 18:42:05 +00:00
|
|
|
}
|
|
|
|
|
2009-02-04 14:06:58 +00:00
|
|
|
/**
|
|
|
|
* ima_file_free - called on __fput()
|
|
|
|
* @file: pointer to file structure being freed
|
|
|
|
*
|
2010-11-02 14:13:07 +00:00
|
|
|
* Flag files that changed, based on i_version
|
2009-02-04 14:06:58 +00:00
|
|
|
*/
|
|
|
|
void ima_file_free(struct file *file)
|
|
|
|
{
|
2013-01-23 22:07:38 +00:00
|
|
|
struct inode *inode = file_inode(file);
|
2011-03-09 19:13:22 +00:00
|
|
|
struct integrity_iint_cache *iint;
|
2009-02-04 14:06:58 +00:00
|
|
|
|
2014-09-24 08:05:10 +00:00
|
|
|
if (!ima_policy_flag || !S_ISREG(inode->i_mode))
|
2009-02-04 14:06:58 +00:00
|
|
|
return;
|
2010-10-25 18:42:19 +00:00
|
|
|
|
2011-03-09 19:13:22 +00:00
|
|
|
iint = integrity_iint_find(inode);
|
2010-11-02 14:14:22 +00:00
|
|
|
if (!iint)
|
|
|
|
return;
|
2009-02-04 14:06:58 +00:00
|
|
|
|
2010-11-02 14:14:22 +00:00
|
|
|
ima_check_last_writer(iint, inode, file);
|
2009-02-04 14:06:58 +00:00
|
|
|
}
|
|
|
|
|
2018-01-08 21:36:20 +00:00
|
|
|
static int process_measurement(struct file *file, const struct cred *cred,
|
|
|
|
u32 secid, char *buf, loff_t size, int mask,
|
2018-06-08 17:40:10 +00:00
|
|
|
enum ima_hooks func)
|
2009-02-04 14:06:58 +00:00
|
|
|
{
|
2013-01-23 22:07:38 +00:00
|
|
|
struct inode *inode = file_inode(file);
|
2014-09-12 17:35:55 +00:00
|
|
|
struct integrity_iint_cache *iint = NULL;
|
2019-06-19 22:46:11 +00:00
|
|
|
struct ima_template_desc *template_desc = NULL;
|
2012-09-03 21:40:17 +00:00
|
|
|
char *pathbuf = NULL;
|
2017-01-17 11:45:41 +00:00
|
|
|
char filename[NAME_MAX];
|
2012-09-03 21:40:17 +00:00
|
|
|
const char *pathname = NULL;
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
int rc = 0, action, must_appraise = 0;
|
2016-06-01 18:14:02 +00:00
|
|
|
int pcr = CONFIG_IMA_MEASURE_PCR_IDX;
|
2014-10-30 10:39:39 +00:00
|
|
|
struct evm_ima_xattr_data *xattr_value = NULL;
|
2019-06-28 02:19:30 +00:00
|
|
|
struct modsig *modsig = NULL;
|
2013-04-25 07:44:04 +00:00
|
|
|
int xattr_len = 0;
|
2014-09-12 17:35:55 +00:00
|
|
|
bool violation_check;
|
2014-10-30 10:39:39 +00:00
|
|
|
enum hash_algo hash_algo;
|
2009-02-04 14:06:58 +00:00
|
|
|
|
2014-09-12 17:35:54 +00:00
|
|
|
if (!ima_policy_flag || !S_ISREG(inode->i_mode))
|
2009-02-04 14:06:58 +00:00
|
|
|
return 0;
|
2010-10-25 18:42:05 +00:00
|
|
|
|
2012-12-03 22:08:11 +00:00
|
|
|
/* Return an IMA_MEASURE, IMA_APPRAISE, IMA_AUDIT action
|
|
|
|
* bitmask based on the appraise/audit/measurement policy.
|
|
|
|
* Included is the appraise submask.
|
|
|
|
*/
|
2021-01-21 13:19:45 +00:00
|
|
|
action = ima_get_action(file_mnt_user_ns(file), inode, cred, secid,
|
|
|
|
mask, func, &pcr, &template_desc, NULL);
|
2016-01-15 01:59:14 +00:00
|
|
|
violation_check = ((func == FILE_CHECK || func == MMAP_CHECK) &&
|
2014-09-12 17:35:55 +00:00
|
|
|
(ima_policy_flag & IMA_MEASURE));
|
|
|
|
if (!action && !violation_check)
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
must_appraise = action & IMA_APPRAISE;
|
2010-10-25 18:42:05 +00:00
|
|
|
|
2012-12-05 20:14:38 +00:00
|
|
|
/* Is the appraise rule hook specific? */
|
2014-09-03 07:19:57 +00:00
|
|
|
if (action & IMA_FILE_APPRAISE)
|
2016-01-15 01:59:14 +00:00
|
|
|
func = FILE_CHECK;
|
2012-12-05 20:14:38 +00:00
|
|
|
|
2016-01-22 20:40:57 +00:00
|
|
|
inode_lock(inode);
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
|
2014-09-12 17:35:55 +00:00
|
|
|
if (action) {
|
|
|
|
iint = integrity_inode_get(inode);
|
|
|
|
if (!iint)
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
rc = -ENOMEM;
|
2014-09-12 17:35:55 +00:00
|
|
|
}
|
|
|
|
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (!rc && violation_check)
|
2014-09-12 17:35:56 +00:00
|
|
|
ima_rdwr_violation_check(file, iint, action & IMA_MEASURE,
|
2017-11-30 10:56:02 +00:00
|
|
|
&pathbuf, &pathname, filename);
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
|
|
|
|
inode_unlock(inode);
|
|
|
|
|
|
|
|
if (rc)
|
|
|
|
goto out;
|
|
|
|
if (!action)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
mutex_lock(&iint->mutex);
|
|
|
|
|
|
|
|
if (test_and_clear_bit(IMA_CHANGE_ATTR, &iint->atomic_flags))
|
|
|
|
/* reset appraisal flags if ima_inode_post_setattr was called */
|
|
|
|
iint->flags &= ~(IMA_APPRAISE | IMA_APPRAISED |
|
|
|
|
IMA_APPRAISE_SUBMASK | IMA_APPRAISED_SUBMASK |
|
|
|
|
IMA_ACTION_FLAGS);
|
|
|
|
|
2018-02-21 16:35:20 +00:00
|
|
|
/*
|
|
|
|
* Re-evaulate the file if either the xattr has changed or the
|
|
|
|
* kernel has no way of detecting file change on the filesystem.
|
|
|
|
* (Limited to privileged mounted filesystems.)
|
|
|
|
*/
|
|
|
|
if (test_and_clear_bit(IMA_CHANGE_XATTR, &iint->atomic_flags) ||
|
|
|
|
((inode->i_sb->s_iflags & SB_I_IMA_UNVERIFIABLE_SIGNATURE) &&
|
2018-02-21 16:36:32 +00:00
|
|
|
!(inode->i_sb->s_iflags & SB_I_UNTRUSTED_MOUNTER) &&
|
|
|
|
!(action & IMA_FAIL_UNVERIFIABLE_SIGS))) {
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
iint->flags &= ~IMA_DONE_MASK;
|
2018-02-21 16:35:20 +00:00
|
|
|
iint->measured_pcrs = 0;
|
|
|
|
}
|
2011-10-19 09:04:40 +00:00
|
|
|
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
/* Determine if already appraised/measured based on bitmask
|
2012-12-03 22:08:11 +00:00
|
|
|
* (IMA_MEASURE, IMA_MEASURED, IMA_XXXX_APPRAISE, IMA_XXXX_APPRAISED,
|
|
|
|
* IMA_AUDIT, IMA_AUDITED)
|
|
|
|
*/
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
iint->flags |= action;
|
2012-06-08 10:58:49 +00:00
|
|
|
action &= IMA_DO_MASK;
|
2016-06-01 18:14:06 +00:00
|
|
|
action &= ~((iint->flags & (IMA_DONE_MASK ^ IMA_MEASURED)) >> 1);
|
|
|
|
|
|
|
|
/* If target pcr is already measured, unset IMA_MEASURE action */
|
|
|
|
if ((action & IMA_MEASURE) && (iint->measured_pcrs & (0x1 << pcr)))
|
|
|
|
action ^= IMA_MEASURE;
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
|
ima: support new "hash" and "dont_hash" policy actions
The builtin ima_appraise_tcb policy, which is specified on the boot
command line, can be replaced with a custom policy, normally early in
the boot process. Custom policies can be more restrictive in some ways,
like requiring file signatures, but can be less restrictive in other
ways, like not appraising mutable files. With a less restrictive policy
in place, files in the builtin policy might not be hashed and labeled
with a security.ima hash. On reboot, files which should be labeled in
the ima_appraise_tcb are not labeled, possibly preventing the system
from booting properly.
To resolve this problem, this patch extends the existing IMA policy
actions "measure", "dont_measure", "appraise", "dont_appraise", and
"audit" with "hash" and "dont_hash". The new "hash" action will write
the file hash as security.ima, but without requiring the file to be
appraised as well.
For example, the builtin ima_appraise_tcb policy includes the rule,
"appraise fowner=0". Adding the "hash fowner=0" rule to a custom
policy, will cause the needed file hashes to be calculated and written
as security.ima xattrs.
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
2016-09-29 14:04:52 +00:00
|
|
|
/* HASH sets the digital signature and update flags, nothing else */
|
|
|
|
if ((action & IMA_HASH) &&
|
|
|
|
!(test_bit(IMA_DIGSIG, &iint->atomic_flags))) {
|
|
|
|
xattr_len = ima_read_xattr(file_dentry(file), &xattr_value);
|
|
|
|
if ((xattr_value && xattr_len > 2) &&
|
|
|
|
(xattr_value->type == EVM_IMA_XATTR_DIGSIG))
|
|
|
|
set_bit(IMA_DIGSIG, &iint->atomic_flags);
|
|
|
|
iint->flags |= IMA_HASHED;
|
|
|
|
action ^= IMA_HASH;
|
|
|
|
set_bit(IMA_UPDATE_XATTR, &iint->atomic_flags);
|
|
|
|
}
|
|
|
|
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
/* Nothing to do, just return existing appraised status */
|
|
|
|
if (!action) {
|
2019-04-30 12:34:44 +00:00
|
|
|
if (must_appraise) {
|
|
|
|
rc = mmap_violation_check(func, file, &pathbuf,
|
|
|
|
&pathname, filename);
|
|
|
|
if (!rc)
|
|
|
|
rc = ima_get_cache_status(iint, func);
|
|
|
|
}
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
goto out_locked;
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
}
|
2009-02-04 14:06:58 +00:00
|
|
|
|
2014-08-22 06:43:55 +00:00
|
|
|
if ((action & IMA_APPRAISE_SUBMASK) ||
|
2019-06-28 02:19:30 +00:00
|
|
|
strcmp(template_desc->name, IMA_TEMPLATE_IMA_NAME) != 0) {
|
2014-10-30 10:39:39 +00:00
|
|
|
/* read 'security.ima' */
|
2016-09-16 10:44:20 +00:00
|
|
|
xattr_len = ima_read_xattr(file_dentry(file), &xattr_value);
|
2013-04-25 07:44:04 +00:00
|
|
|
|
2019-06-28 02:19:33 +00:00
|
|
|
/*
|
|
|
|
* Read the appended modsig if allowed by the policy, and allow
|
|
|
|
* an additional measurement list entry, if needed, based on the
|
|
|
|
* template format and whether the file was already measured.
|
|
|
|
*/
|
|
|
|
if (iint->flags & IMA_MODSIG_ALLOWED) {
|
|
|
|
rc = ima_read_modsig(func, buf, size, &modsig);
|
|
|
|
|
|
|
|
if (!rc && ima_template_has_modsig(template_desc) &&
|
|
|
|
iint->flags & IMA_MEASURED)
|
|
|
|
action |= IMA_MEASURE;
|
|
|
|
}
|
2019-06-28 02:19:30 +00:00
|
|
|
}
|
|
|
|
|
2014-10-30 10:39:39 +00:00
|
|
|
hash_algo = ima_get_hash_algo(xattr_value, xattr_len);
|
|
|
|
|
2019-06-28 02:19:31 +00:00
|
|
|
rc = ima_collect_measurement(iint, file, buf, size, hash_algo, modsig);
|
2017-06-18 03:56:23 +00:00
|
|
|
if (rc != 0 && rc != -EBADF && rc != -EINVAL)
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
goto out_locked;
|
2012-06-20 13:32:55 +00:00
|
|
|
|
2017-01-17 11:45:41 +00:00
|
|
|
if (!pathbuf) /* ima_rdwr_violation possibly pre-fetched */
|
|
|
|
pathname = ima_d_path(&file->f_path, &pathbuf, filename);
|
2012-06-20 13:32:55 +00:00
|
|
|
|
ima: integrity appraisal extension
IMA currently maintains an integrity measurement list used to assert the
integrity of the running system to a third party. The IMA-appraisal
extension adds local integrity validation and enforcement of the
measurement against a "good" value stored as an extended attribute
'security.ima'. The initial methods for validating 'security.ima' are
hashed based, which provides file data integrity, and digital signature
based, which in addition to providing file data integrity, provides
authenticity.
This patch creates and maintains the 'security.ima' xattr, containing
the file data hash measurement. Protection of the xattr is provided by
EVM, if enabled and configured.
Based on policy, IMA calls evm_verifyxattr() to verify a file's metadata
integrity and, assuming success, compares the file's current hash value
with the one stored as an extended attribute in 'security.ima'.
Changelov v4:
- changed iint cache flags to hex values
Changelog v3:
- change appraisal default for filesystems without xattr support to fail
Changelog v2:
- fix audit msg 'res' value
- removed unused 'ima_appraise=' values
Changelog v1:
- removed unused iint mutex (Dmitry Kasatkin)
- setattr hook must not reset appraised (Dmitry Kasatkin)
- evm_verifyxattr() now differentiates between no 'security.evm' xattr
(INTEGRITY_NOLABEL) and no EVM 'protected' xattrs included in the
'security.evm' (INTEGRITY_NOXATTRS).
- replace hash_status with ima_status (Dmitry Kasatkin)
- re-initialize slab element ima_status on free (Dmitry Kasatkin)
- include 'security.ima' in EVM if CONFIG_IMA_APPRAISE, not CONFIG_IMA
- merged half "ima: ima_must_appraise_or_measure API change" (Dmitry Kasatkin)
- removed unnecessary error variable in process_measurement() (Dmitry Kasatkin)
- use ima_inode_post_setattr() stub function, if IMA_APPRAISE not configured
(moved ima_inode_post_setattr() to ima_appraise.c)
- make sure ima_collect_measurement() can read file
Changelog:
- add 'iint' to evm_verifyxattr() call (Dimitry Kasatkin)
- fix the race condition between chmod, which takes the i_mutex and then
iint->mutex, and ima_file_free() and process_measurement(), which take
the locks in the reverse order, by eliminating iint->mutex. (Dmitry Kasatkin)
- cleanup of ima_appraise_measurement() (Dmitry Kasatkin)
- changes as a result of the iint not allocated for all regular files, but
only for those measured/appraised.
- don't try to appraise new/empty files
- expanded ima_appraisal description in ima/Kconfig
- IMA appraise definitions required even if IMA_APPRAISE not enabled
- add return value to ima_must_appraise() stub
- unconditionally set status = INTEGRITY_PASS *after* testing status,
not before. (Found by Joe Perches)
Signed-off-by: Mimi Zohar <zohar@us.ibm.com>
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@intel.com>
2012-02-13 15:15:05 +00:00
|
|
|
if (action & IMA_MEASURE)
|
2013-07-23 15:15:00 +00:00
|
|
|
ima_store_measurement(iint, file, pathname,
|
2019-06-28 02:19:32 +00:00
|
|
|
xattr_value, xattr_len, modsig, pcr,
|
2019-06-19 22:46:11 +00:00
|
|
|
template_desc);
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (rc == 0 && (action & IMA_APPRAISE_SUBMASK)) {
|
2019-10-31 03:31:32 +00:00
|
|
|
rc = ima_check_blacklist(iint, modsig, pcr);
|
|
|
|
if (rc != -EPERM) {
|
|
|
|
inode_lock(inode);
|
|
|
|
rc = ima_appraise_measurement(func, iint, file,
|
|
|
|
pathname, xattr_value,
|
|
|
|
xattr_len, modsig);
|
|
|
|
inode_unlock(inode);
|
|
|
|
}
|
2019-04-30 12:34:44 +00:00
|
|
|
if (!rc)
|
|
|
|
rc = mmap_violation_check(func, file, &pathbuf,
|
|
|
|
&pathname, filename);
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
}
|
2012-06-14 17:04:36 +00:00
|
|
|
if (action & IMA_AUDIT)
|
2012-09-03 21:40:17 +00:00
|
|
|
ima_audit_measurement(iint, pathname);
|
2014-09-12 17:35:55 +00:00
|
|
|
|
2017-06-18 03:56:23 +00:00
|
|
|
if ((file->f_flags & O_DIRECT) && (iint->flags & IMA_PERMIT_DIRECTIO))
|
|
|
|
rc = 0;
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
out_locked:
|
|
|
|
if ((mask & MAY_WRITE) && test_bit(IMA_DIGSIG, &iint->atomic_flags) &&
|
2016-03-01 00:52:05 +00:00
|
|
|
!(iint->flags & IMA_NEW_FILE))
|
2012-09-27 12:06:28 +00:00
|
|
|
rc = -EACCES;
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
mutex_unlock(&iint->mutex);
|
2014-09-12 17:35:55 +00:00
|
|
|
kfree(xattr_value);
|
2019-06-28 02:19:30 +00:00
|
|
|
ima_free_modsig(modsig);
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
out:
|
2014-10-01 18:43:10 +00:00
|
|
|
if (pathbuf)
|
|
|
|
__putname(pathbuf);
|
ima: re-introduce own integrity cache lock
Before IMA appraisal was introduced, IMA was using own integrity cache
lock along with i_mutex. process_measurement and ima_file_free took
the iint->mutex first and then the i_mutex, while setxattr, chmod and
chown took the locks in reverse order. To resolve the potential deadlock,
i_mutex was moved to protect entire IMA functionality and the redundant
iint->mutex was eliminated.
Solution was based on the assumption that filesystem code does not take
i_mutex further. But when file is opened with O_DIRECT flag, direct-io
implementation takes i_mutex and produces deadlock. Furthermore, certain
other filesystem operations, such as llseek, also take i_mutex.
More recently some filesystems have replaced their filesystem specific
lock with the global i_rwsem to read a file. As a result, when IMA
attempts to calculate the file hash, reading the file attempts to take
the i_rwsem again.
To resolve O_DIRECT related deadlock problem, this patch re-introduces
iint->mutex. But to eliminate the original chmod() related deadlock
problem, this patch eliminates the requirement for chmod hooks to take
the iint->mutex by introducing additional atomic iint->attr_flags to
indicate calling of the hooks. The allowed locking order is to take
the iint->mutex first and then the i_rwsem.
Original flags were cleared in chmod(), setxattr() or removwxattr()
hooks and tested when file was closed or opened again. New atomic flags
are set or cleared in those hooks and tested to clear iint->flags on
close or on open.
Atomic flags are following:
* IMA_CHANGE_ATTR - indicates that chATTR() was called (chmod, chown,
chgrp) and file attributes have changed. On file open, it causes IMA
to clear iint->flags to re-evaluate policy and perform IMA functions
again.
* IMA_CHANGE_XATTR - indicates that setxattr or removexattr was called
and extended attributes have changed. On file open, it causes IMA to
clear iint->flags IMA_DONE_MASK to re-appraise.
* IMA_UPDATE_XATTR - indicates that security.ima needs to be updated.
It is cleared if file policy changes and no update is needed.
* IMA_DIGSIG - indicates that file security.ima has signature and file
security.ima must not update to file has on file close.
* IMA_MUST_MEASURE - indicates the file is in the measurement policy.
Fixes: Commit 6552321831dc ("xfs: remove i_iolock and use i_rwsem in
the VFS inode instead")
Signed-off-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
2017-12-05 19:06:34 +00:00
|
|
|
if (must_appraise) {
|
|
|
|
if (rc && (ima_appraise & IMA_APPRAISE_ENFORCE))
|
|
|
|
return -EACCES;
|
|
|
|
if (file->f_mode & FMODE_WRITE)
|
|
|
|
set_bit(IMA_UPDATE_XATTR, &iint->atomic_flags);
|
|
|
|
}
|
2012-09-27 12:57:10 +00:00
|
|
|
return 0;
|
2009-02-04 14:06:58 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* ima_file_mmap - based on policy, collect/store measurement.
|
|
|
|
* @file: pointer to the file to be measured (May be NULL)
|
|
|
|
* @prot: contains the protection that will be applied by the kernel.
|
|
|
|
*
|
|
|
|
* Measure files being mmapped executable based on the ima_must_measure()
|
|
|
|
* policy decision.
|
|
|
|
*
|
2012-09-27 12:57:10 +00:00
|
|
|
* On success return 0. On integrity appraisal error, assuming the file
|
|
|
|
* is in policy and IMA-appraisal is in enforcing mode, return -EACCES.
|
2009-02-04 14:06:58 +00:00
|
|
|
*/
|
|
|
|
int ima_file_mmap(struct file *file, unsigned long prot)
|
|
|
|
{
|
2018-01-08 21:36:20 +00:00
|
|
|
u32 secid;
|
|
|
|
|
|
|
|
if (file && (prot & PROT_EXEC)) {
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2018-01-08 21:36:20 +00:00
|
|
|
return process_measurement(file, current_cred(), secid, NULL,
|
2018-06-08 17:40:10 +00:00
|
|
|
0, MAY_EXEC, MMAP_CHECK);
|
2018-01-08 21:36:20 +00:00
|
|
|
}
|
|
|
|
|
2012-09-27 12:57:10 +00:00
|
|
|
return 0;
|
2009-02-04 14:06:58 +00:00
|
|
|
}
|
|
|
|
|
2020-05-03 05:00:02 +00:00
|
|
|
/**
|
|
|
|
* ima_file_mprotect - based on policy, limit mprotect change
|
|
|
|
* @prot: contains the protection that will be applied by the kernel.
|
|
|
|
*
|
|
|
|
* Files can be mmap'ed read/write and later changed to execute to circumvent
|
|
|
|
* IMA's mmap appraisal policy rules. Due to locking issues (mmap semaphore
|
|
|
|
* would be taken before i_mutex), files can not be measured or appraised at
|
|
|
|
* this point. Eliminate this integrity gap by denying the mprotect
|
|
|
|
* PROT_EXECUTE change, if an mmap appraise policy rule exists.
|
|
|
|
*
|
|
|
|
* On mprotect change success, return 0. On failure, return -EACESS.
|
|
|
|
*/
|
|
|
|
int ima_file_mprotect(struct vm_area_struct *vma, unsigned long prot)
|
|
|
|
{
|
2020-11-12 20:39:59 +00:00
|
|
|
struct ima_template_desc *template = NULL;
|
2020-05-03 05:00:02 +00:00
|
|
|
struct file *file = vma->vm_file;
|
|
|
|
char filename[NAME_MAX];
|
|
|
|
char *pathbuf = NULL;
|
|
|
|
const char *pathname = NULL;
|
|
|
|
struct inode *inode;
|
|
|
|
int result = 0;
|
|
|
|
int action;
|
|
|
|
u32 secid;
|
|
|
|
int pcr;
|
|
|
|
|
|
|
|
/* Is mprotect making an mmap'ed file executable? */
|
2020-06-10 13:18:26 +00:00
|
|
|
if (!(ima_policy_flag & IMA_APPRAISE) || !vma->vm_file ||
|
|
|
|
!(prot & PROT_EXEC) || (vma->vm_flags & VM_EXEC))
|
2020-05-03 05:00:02 +00:00
|
|
|
return 0;
|
|
|
|
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2020-05-03 05:00:02 +00:00
|
|
|
inode = file_inode(vma->vm_file);
|
2021-01-21 13:19:45 +00:00
|
|
|
action = ima_get_action(file_mnt_user_ns(vma->vm_file), inode,
|
|
|
|
current_cred(), secid, MAY_EXEC, MMAP_CHECK,
|
|
|
|
&pcr, &template, 0);
|
2020-05-03 05:00:02 +00:00
|
|
|
|
|
|
|
/* Is the mmap'ed file in policy? */
|
|
|
|
if (!(action & (IMA_MEASURE | IMA_APPRAISE_SUBMASK)))
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
if (action & IMA_APPRAISE_SUBMASK)
|
|
|
|
result = -EPERM;
|
|
|
|
|
|
|
|
file = vma->vm_file;
|
|
|
|
pathname = ima_d_path(&file->f_path, &pathbuf, filename);
|
|
|
|
integrity_audit_msg(AUDIT_INTEGRITY_DATA, inode, pathname,
|
|
|
|
"collect_data", "failed-mprotect", result, 0);
|
|
|
|
if (pathbuf)
|
|
|
|
__putname(pathbuf);
|
|
|
|
|
|
|
|
return result;
|
|
|
|
}
|
|
|
|
|
2009-02-04 14:06:58 +00:00
|
|
|
/**
|
|
|
|
* ima_bprm_check - based on policy, collect/store measurement.
|
|
|
|
* @bprm: contains the linux_binprm structure
|
|
|
|
*
|
|
|
|
* The OS protects against an executable file, already open for write,
|
|
|
|
* from being executed in deny_write_access() and an executable file,
|
|
|
|
* already open for execute, from being modified in get_write_access().
|
|
|
|
* So we can be certain that what we verify and measure here is actually
|
|
|
|
* what is being executed.
|
|
|
|
*
|
2012-09-27 12:57:10 +00:00
|
|
|
* On success return 0. On integrity appraisal error, assuming the file
|
|
|
|
* is in policy and IMA-appraisal is in enforcing mode, return -EACCES.
|
2009-02-04 14:06:58 +00:00
|
|
|
*/
|
|
|
|
int ima_bprm_check(struct linux_binprm *bprm)
|
|
|
|
{
|
2018-01-08 21:36:20 +00:00
|
|
|
int ret;
|
|
|
|
u32 secid;
|
|
|
|
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2018-01-08 21:36:20 +00:00
|
|
|
ret = process_measurement(bprm->file, current_cred(), secid, NULL, 0,
|
2018-06-08 17:40:10 +00:00
|
|
|
MAY_EXEC, BPRM_CHECK);
|
2018-01-08 21:36:20 +00:00
|
|
|
if (ret)
|
|
|
|
return ret;
|
|
|
|
|
|
|
|
security_cred_getsecid(bprm->cred, &secid);
|
|
|
|
return process_measurement(bprm->file, bprm->cred, secid, NULL, 0,
|
2018-06-08 17:40:10 +00:00
|
|
|
MAY_EXEC, CREDS_CHECK);
|
2009-02-04 14:06:58 +00:00
|
|
|
}
|
|
|
|
|
2010-01-20 20:35:41 +00:00
|
|
|
/**
|
2021-04-06 02:12:10 +00:00
|
|
|
* ima_file_check - based on policy, collect/store measurement.
|
2010-01-20 20:35:41 +00:00
|
|
|
* @file: pointer to the file to be measured
|
2017-01-06 04:38:11 +00:00
|
|
|
* @mask: contains MAY_READ, MAY_WRITE, MAY_EXEC or MAY_APPEND
|
2010-01-20 20:35:41 +00:00
|
|
|
*
|
|
|
|
* Measure files based on the ima_must_measure() policy decision.
|
|
|
|
*
|
2012-09-27 12:57:10 +00:00
|
|
|
* On success return 0. On integrity appraisal error, assuming the file
|
|
|
|
* is in policy and IMA-appraisal is in enforcing mode, return -EACCES.
|
2010-01-20 20:35:41 +00:00
|
|
|
*/
|
2018-06-08 17:40:10 +00:00
|
|
|
int ima_file_check(struct file *file, int mask)
|
2010-01-20 20:35:41 +00:00
|
|
|
{
|
2018-01-08 21:36:20 +00:00
|
|
|
u32 secid;
|
|
|
|
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2018-01-08 21:36:20 +00:00
|
|
|
return process_measurement(file, current_cred(), secid, NULL, 0,
|
2017-01-06 04:38:11 +00:00
|
|
|
mask & (MAY_READ | MAY_WRITE | MAY_EXEC |
|
2018-06-08 17:40:10 +00:00
|
|
|
MAY_APPEND), FILE_CHECK);
|
2010-01-20 20:35:41 +00:00
|
|
|
}
|
2010-01-26 22:02:40 +00:00
|
|
|
EXPORT_SYMBOL_GPL(ima_file_check);
|
2010-01-20 20:35:41 +00:00
|
|
|
|
2020-11-24 15:12:08 +00:00
|
|
|
static int __ima_inode_hash(struct inode *inode, char *buf, size_t buf_size)
|
2020-01-13 09:42:44 +00:00
|
|
|
{
|
|
|
|
struct integrity_iint_cache *iint;
|
|
|
|
int hash_algo;
|
|
|
|
|
|
|
|
if (!ima_policy_flag)
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
|
|
|
iint = integrity_iint_find(inode);
|
|
|
|
if (!iint)
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
|
|
|
|
mutex_lock(&iint->mutex);
|
ima: Fix NULL pointer dereference in ima_file_hash
ima_file_hash can be called when there is no iint->ima_hash available
even though the inode exists in the integrity cache. It is fairly
common for a file to not have a hash. (e.g. an mknodat, prior to the
file being closed).
Another example where this can happen (suggested by Jann Horn):
Process A does:
while(1) {
unlink("/tmp/imafoo");
fd = open("/tmp/imafoo", O_RDWR|O_CREAT|O_TRUNC, 0700);
if (fd == -1) {
perror("open");
continue;
}
write(fd, "A", 1);
close(fd);
}
and Process B does:
while (1) {
int fd = open("/tmp/imafoo", O_RDONLY);
if (fd == -1)
continue;
char *mapping = mmap(NULL, 0x1000, PROT_READ|PROT_EXEC,
MAP_PRIVATE, fd, 0);
if (mapping != MAP_FAILED)
munmap(mapping, 0x1000);
close(fd);
}
Due to the race to get the iint->mutex between ima_file_hash and
process_measurement iint->ima_hash could still be NULL.
Fixes: 6beea7afcc72 ("ima: add the ability to query the cached hash of a given file")
Signed-off-by: KP Singh <kpsingh@google.com>
Reviewed-by: Florent Revest <revest@chromium.org>
Signed-off-by: Mimi Zohar <zohar@linux.ibm.com>
2020-09-16 18:02:42 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* ima_file_hash can be called when ima_collect_measurement has still
|
|
|
|
* not been called, we might not always have a hash.
|
|
|
|
*/
|
|
|
|
if (!iint->ima_hash) {
|
|
|
|
mutex_unlock(&iint->mutex);
|
|
|
|
return -EOPNOTSUPP;
|
|
|
|
}
|
|
|
|
|
2020-01-13 09:42:44 +00:00
|
|
|
if (buf) {
|
|
|
|
size_t copied_size;
|
|
|
|
|
|
|
|
copied_size = min_t(size_t, iint->ima_hash->length, buf_size);
|
|
|
|
memcpy(buf, iint->ima_hash->digest, copied_size);
|
|
|
|
}
|
|
|
|
hash_algo = iint->ima_hash->algo;
|
|
|
|
mutex_unlock(&iint->mutex);
|
|
|
|
|
|
|
|
return hash_algo;
|
|
|
|
}
|
2020-11-24 15:12:08 +00:00
|
|
|
|
|
|
|
/**
|
|
|
|
* ima_file_hash - return the stored measurement if a file has been hashed and
|
|
|
|
* is in the iint cache.
|
|
|
|
* @file: pointer to the file
|
|
|
|
* @buf: buffer in which to store the hash
|
|
|
|
* @buf_size: length of the buffer
|
|
|
|
*
|
|
|
|
* On success, return the hash algorithm (as defined in the enum hash_algo).
|
|
|
|
* If buf is not NULL, this function also outputs the hash into buf.
|
|
|
|
* If the hash is larger than buf_size, then only buf_size bytes will be copied.
|
|
|
|
* It generally just makes sense to pass a buffer capable of holding the largest
|
|
|
|
* possible hash: IMA_MAX_DIGEST_SIZE.
|
|
|
|
* The file hash returned is based on the entire file, including the appended
|
|
|
|
* signature.
|
|
|
|
*
|
|
|
|
* If IMA is disabled or if no measurement is available, return -EOPNOTSUPP.
|
|
|
|
* If the parameters are incorrect, return -EINVAL.
|
|
|
|
*/
|
|
|
|
int ima_file_hash(struct file *file, char *buf, size_t buf_size)
|
|
|
|
{
|
|
|
|
if (!file)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
return __ima_inode_hash(file_inode(file), buf, buf_size);
|
|
|
|
}
|
2020-01-13 09:42:44 +00:00
|
|
|
EXPORT_SYMBOL_GPL(ima_file_hash);
|
|
|
|
|
2020-11-24 15:12:08 +00:00
|
|
|
/**
|
|
|
|
* ima_inode_hash - return the stored measurement if the inode has been hashed
|
|
|
|
* and is in the iint cache.
|
|
|
|
* @inode: pointer to the inode
|
|
|
|
* @buf: buffer in which to store the hash
|
|
|
|
* @buf_size: length of the buffer
|
|
|
|
*
|
|
|
|
* On success, return the hash algorithm (as defined in the enum hash_algo).
|
|
|
|
* If buf is not NULL, this function also outputs the hash into buf.
|
|
|
|
* If the hash is larger than buf_size, then only buf_size bytes will be copied.
|
|
|
|
* It generally just makes sense to pass a buffer capable of holding the largest
|
|
|
|
* possible hash: IMA_MAX_DIGEST_SIZE.
|
|
|
|
* The hash returned is based on the entire contents, including the appended
|
|
|
|
* signature.
|
|
|
|
*
|
|
|
|
* If IMA is disabled or if no measurement is available, return -EOPNOTSUPP.
|
|
|
|
* If the parameters are incorrect, return -EINVAL.
|
|
|
|
*/
|
|
|
|
int ima_inode_hash(struct inode *inode, char *buf, size_t buf_size)
|
|
|
|
{
|
|
|
|
if (!inode)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
return __ima_inode_hash(inode, buf, buf_size);
|
|
|
|
}
|
|
|
|
EXPORT_SYMBOL_GPL(ima_inode_hash);
|
|
|
|
|
2019-01-22 20:06:49 +00:00
|
|
|
/**
|
|
|
|
* ima_post_create_tmpfile - mark newly created tmpfile as new
|
2021-01-21 13:19:45 +00:00
|
|
|
* @mnt_userns: user namespace of the mount the inode was found from
|
2019-01-22 20:06:49 +00:00
|
|
|
* @file : newly created tmpfile
|
|
|
|
*
|
|
|
|
* No measuring, appraising or auditing of newly created tmpfiles is needed.
|
|
|
|
* Skip calling process_measurement(), but indicate which newly, created
|
|
|
|
* tmpfiles are in policy.
|
|
|
|
*/
|
2021-01-21 13:19:45 +00:00
|
|
|
void ima_post_create_tmpfile(struct user_namespace *mnt_userns,
|
|
|
|
struct inode *inode)
|
2019-01-22 20:06:49 +00:00
|
|
|
{
|
|
|
|
struct integrity_iint_cache *iint;
|
|
|
|
int must_appraise;
|
|
|
|
|
2021-03-19 15:14:25 +00:00
|
|
|
if (!ima_policy_flag || !S_ISREG(inode->i_mode))
|
|
|
|
return;
|
|
|
|
|
2021-01-21 13:19:45 +00:00
|
|
|
must_appraise = ima_must_appraise(mnt_userns, inode, MAY_ACCESS,
|
|
|
|
FILE_CHECK);
|
2019-01-22 20:06:49 +00:00
|
|
|
if (!must_appraise)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Nothing to do if we can't allocate memory */
|
|
|
|
iint = integrity_inode_get(inode);
|
|
|
|
if (!iint)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* needed for writing the security xattrs */
|
|
|
|
set_bit(IMA_UPDATE_XATTR, &iint->atomic_flags);
|
|
|
|
iint->ima_file_status = INTEGRITY_PASS;
|
|
|
|
}
|
|
|
|
|
2016-03-01 00:52:05 +00:00
|
|
|
/**
|
|
|
|
* ima_post_path_mknod - mark as a new inode
|
2021-01-21 13:19:45 +00:00
|
|
|
* @mnt_userns: user namespace of the mount the inode was found from
|
2016-03-01 00:52:05 +00:00
|
|
|
* @dentry: newly created dentry
|
|
|
|
*
|
|
|
|
* Mark files created via the mknodat syscall as new, so that the
|
|
|
|
* file data can be written later.
|
|
|
|
*/
|
2021-01-21 13:19:45 +00:00
|
|
|
void ima_post_path_mknod(struct user_namespace *mnt_userns,
|
|
|
|
struct dentry *dentry)
|
2016-03-01 00:52:05 +00:00
|
|
|
{
|
|
|
|
struct integrity_iint_cache *iint;
|
|
|
|
struct inode *inode = dentry->d_inode;
|
|
|
|
int must_appraise;
|
|
|
|
|
2021-03-19 15:14:25 +00:00
|
|
|
if (!ima_policy_flag || !S_ISREG(inode->i_mode))
|
|
|
|
return;
|
|
|
|
|
2021-01-21 13:19:45 +00:00
|
|
|
must_appraise = ima_must_appraise(mnt_userns, inode, MAY_ACCESS,
|
|
|
|
FILE_CHECK);
|
2016-03-01 00:52:05 +00:00
|
|
|
if (!must_appraise)
|
|
|
|
return;
|
|
|
|
|
2019-01-22 20:06:49 +00:00
|
|
|
/* Nothing to do if we can't allocate memory */
|
2016-03-01 00:52:05 +00:00
|
|
|
iint = integrity_inode_get(inode);
|
2019-01-22 20:06:49 +00:00
|
|
|
if (!iint)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* needed for re-opening empty files */
|
|
|
|
iint->flags |= IMA_NEW_FILE;
|
2016-03-01 00:52:05 +00:00
|
|
|
}
|
|
|
|
|
2016-01-31 03:23:26 +00:00
|
|
|
/**
|
|
|
|
* ima_read_file - pre-measure/appraise hook decision based on policy
|
|
|
|
* @file: pointer to the file to be measured/appraised/audit
|
|
|
|
* @read_id: caller identifier
|
2020-10-02 17:38:23 +00:00
|
|
|
* @contents: whether a subsequent call will be made to ima_post_read_file()
|
2016-01-31 03:23:26 +00:00
|
|
|
*
|
|
|
|
* Permit reading a file based on policy. The policy rules are written
|
|
|
|
* in terms of the policy identifier. Appraising the integrity of
|
|
|
|
* a file requires a file descriptor.
|
|
|
|
*
|
|
|
|
* For permission return 0, otherwise return -EACCES.
|
|
|
|
*/
|
2020-10-02 17:38:23 +00:00
|
|
|
int ima_read_file(struct file *file, enum kernel_read_file_id read_id,
|
|
|
|
bool contents)
|
2016-01-31 03:23:26 +00:00
|
|
|
{
|
2020-10-02 17:38:24 +00:00
|
|
|
enum ima_hooks func;
|
|
|
|
u32 secid;
|
2020-10-02 17:38:23 +00:00
|
|
|
|
2018-07-13 18:06:03 +00:00
|
|
|
/*
|
|
|
|
* Do devices using pre-allocated memory run the risk of the
|
|
|
|
* firmware being accessible to the device prior to the completion
|
|
|
|
* of IMA's signature verification any more than when using two
|
2020-10-02 17:38:13 +00:00
|
|
|
* buffers? It may be desirable to include the buffer address
|
|
|
|
* in this API and walk all the dma_map_single() mappings to check.
|
2018-07-13 18:06:03 +00:00
|
|
|
*/
|
2020-10-02 17:38:24 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* There will be a call made to ima_post_read_file() with
|
|
|
|
* a filled buffer, so we don't need to perform an extra
|
|
|
|
* read early here.
|
|
|
|
*/
|
|
|
|
if (contents)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
/* Read entire file for all partial reads. */
|
|
|
|
func = read_idmap[read_id] ?: FILE_CHECK;
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2020-10-02 17:38:24 +00:00
|
|
|
return process_measurement(file, current_cred(), secid, NULL,
|
|
|
|
0, MAY_READ, func);
|
2016-01-31 03:23:26 +00:00
|
|
|
}
|
|
|
|
|
2019-08-20 00:18:01 +00:00
|
|
|
const int read_idmap[READING_MAX_ID] = {
|
ima: support for kexec image and initramfs
Add IMA policy support for measuring/appraising the kexec image and
initramfs. Two new IMA policy identifiers KEXEC_KERNEL_CHECK and
KEXEC_INITRAMFS_CHECK are defined.
Example policy rules:
measure func=KEXEC_KERNEL_CHECK
appraise func=KEXEC_KERNEL_CHECK appraise_type=imasig
measure func=KEXEC_INITRAMFS_CHECK
appraise func=KEXEC_INITRAMFS_CHECK appraise_type=imasig
Moving the enumeration to the vfs layer simplified the patches, allowing
the IMA changes, for the most part, to be separated from the other
changes. Unfortunately, passing either a kernel_read_file_id or a
ima_hooks enumeration within IMA is messy.
Option 1: duplicate kernel_read_file enumeration in ima_hooks
enum kernel_read_file_id {
...
READING_KEXEC_IMAGE,
READING_KEXEC_INITRAMFS,
READING_MAX_ID
enum ima_hooks {
...
KEXEC_KERNEL_CHECK
KEXEC_INITRAMFS_CHECK
Option 2: define ima_hooks as extension of kernel_read_file
eg: enum ima_hooks {
FILE_CHECK = READING_MAX_ID,
MMAP_CHECK,
In order to pass both kernel_read_file_id and ima_hooks values, we
would need to specify a struct containing a union.
struct caller_id {
union {
enum ima_hooks func_id;
enum kernel_read_file_id read_id;
};
};
Option 3: incorportate the ima_hooks enumeration into kernel_read_file_id,
perhaps changing the enumeration name.
For now, duplicate the new READING_KEXEC_IMAGE/INITRAMFS in the ima_hooks.
Changelog v4:
- replaced switch statement with a kernel_read_file_id to an ima_hooks
id mapping array - Dmitry
- renamed ima_hook tokens KEXEC_CHECK and INITRAMFS_CHECK to
KEXEC_KERNEL_CHECK and KEXEC_INITRAMFS_CHECK respectively - Dave Young
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Acked-by: Petko Manolov <petkan@mip-labs.com>
Acked-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Cc: Dave Young <dyoung@redhat.com>
2016-01-15 01:59:14 +00:00
|
|
|
[READING_FIRMWARE] = FIRMWARE_CHECK,
|
|
|
|
[READING_MODULE] = MODULE_CHECK,
|
|
|
|
[READING_KEXEC_IMAGE] = KEXEC_KERNEL_CHECK,
|
|
|
|
[READING_KEXEC_INITRAMFS] = KEXEC_INITRAMFS_CHECK,
|
2016-01-15 15:17:12 +00:00
|
|
|
[READING_POLICY] = POLICY_CHECK
|
ima: support for kexec image and initramfs
Add IMA policy support for measuring/appraising the kexec image and
initramfs. Two new IMA policy identifiers KEXEC_KERNEL_CHECK and
KEXEC_INITRAMFS_CHECK are defined.
Example policy rules:
measure func=KEXEC_KERNEL_CHECK
appraise func=KEXEC_KERNEL_CHECK appraise_type=imasig
measure func=KEXEC_INITRAMFS_CHECK
appraise func=KEXEC_INITRAMFS_CHECK appraise_type=imasig
Moving the enumeration to the vfs layer simplified the patches, allowing
the IMA changes, for the most part, to be separated from the other
changes. Unfortunately, passing either a kernel_read_file_id or a
ima_hooks enumeration within IMA is messy.
Option 1: duplicate kernel_read_file enumeration in ima_hooks
enum kernel_read_file_id {
...
READING_KEXEC_IMAGE,
READING_KEXEC_INITRAMFS,
READING_MAX_ID
enum ima_hooks {
...
KEXEC_KERNEL_CHECK
KEXEC_INITRAMFS_CHECK
Option 2: define ima_hooks as extension of kernel_read_file
eg: enum ima_hooks {
FILE_CHECK = READING_MAX_ID,
MMAP_CHECK,
In order to pass both kernel_read_file_id and ima_hooks values, we
would need to specify a struct containing a union.
struct caller_id {
union {
enum ima_hooks func_id;
enum kernel_read_file_id read_id;
};
};
Option 3: incorportate the ima_hooks enumeration into kernel_read_file_id,
perhaps changing the enumeration name.
For now, duplicate the new READING_KEXEC_IMAGE/INITRAMFS in the ima_hooks.
Changelog v4:
- replaced switch statement with a kernel_read_file_id to an ima_hooks
id mapping array - Dmitry
- renamed ima_hook tokens KEXEC_CHECK and INITRAMFS_CHECK to
KEXEC_KERNEL_CHECK and KEXEC_INITRAMFS_CHECK respectively - Dave Young
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Acked-by: Petko Manolov <petkan@mip-labs.com>
Acked-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Cc: Dave Young <dyoung@redhat.com>
2016-01-15 01:59:14 +00:00
|
|
|
};
|
|
|
|
|
2016-01-14 22:57:47 +00:00
|
|
|
/**
|
|
|
|
* ima_post_read_file - in memory collect/appraise/audit measurement
|
|
|
|
* @file: pointer to the file to be measured/appraised/audit
|
|
|
|
* @buf: pointer to in memory file contents
|
|
|
|
* @size: size of in memory file contents
|
|
|
|
* @read_id: caller identifier
|
|
|
|
*
|
|
|
|
* Measure/appraise/audit in memory file based on policy. Policy rules
|
|
|
|
* are written in terms of a policy identifier.
|
|
|
|
*
|
|
|
|
* On success return 0. On integrity appraisal error, assuming the file
|
|
|
|
* is in policy and IMA-appraisal is in enforcing mode, return -EACCES.
|
|
|
|
*/
|
|
|
|
int ima_post_read_file(struct file *file, void *buf, loff_t size,
|
|
|
|
enum kernel_read_file_id read_id)
|
|
|
|
{
|
ima: support for kexec image and initramfs
Add IMA policy support for measuring/appraising the kexec image and
initramfs. Two new IMA policy identifiers KEXEC_KERNEL_CHECK and
KEXEC_INITRAMFS_CHECK are defined.
Example policy rules:
measure func=KEXEC_KERNEL_CHECK
appraise func=KEXEC_KERNEL_CHECK appraise_type=imasig
measure func=KEXEC_INITRAMFS_CHECK
appraise func=KEXEC_INITRAMFS_CHECK appraise_type=imasig
Moving the enumeration to the vfs layer simplified the patches, allowing
the IMA changes, for the most part, to be separated from the other
changes. Unfortunately, passing either a kernel_read_file_id or a
ima_hooks enumeration within IMA is messy.
Option 1: duplicate kernel_read_file enumeration in ima_hooks
enum kernel_read_file_id {
...
READING_KEXEC_IMAGE,
READING_KEXEC_INITRAMFS,
READING_MAX_ID
enum ima_hooks {
...
KEXEC_KERNEL_CHECK
KEXEC_INITRAMFS_CHECK
Option 2: define ima_hooks as extension of kernel_read_file
eg: enum ima_hooks {
FILE_CHECK = READING_MAX_ID,
MMAP_CHECK,
In order to pass both kernel_read_file_id and ima_hooks values, we
would need to specify a struct containing a union.
struct caller_id {
union {
enum ima_hooks func_id;
enum kernel_read_file_id read_id;
};
};
Option 3: incorportate the ima_hooks enumeration into kernel_read_file_id,
perhaps changing the enumeration name.
For now, duplicate the new READING_KEXEC_IMAGE/INITRAMFS in the ima_hooks.
Changelog v4:
- replaced switch statement with a kernel_read_file_id to an ima_hooks
id mapping array - Dmitry
- renamed ima_hook tokens KEXEC_CHECK and INITRAMFS_CHECK to
KEXEC_KERNEL_CHECK and KEXEC_INITRAMFS_CHECK respectively - Dave Young
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Acked-by: Petko Manolov <petkan@mip-labs.com>
Acked-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Cc: Dave Young <dyoung@redhat.com>
2016-01-15 01:59:14 +00:00
|
|
|
enum ima_hooks func;
|
2018-01-08 21:36:20 +00:00
|
|
|
u32 secid;
|
2016-01-14 22:57:47 +00:00
|
|
|
|
2017-09-10 07:49:45 +00:00
|
|
|
/* permit signed certs */
|
|
|
|
if (!file && read_id == READING_X509_CERTIFICATE)
|
|
|
|
return 0;
|
|
|
|
|
2016-01-14 22:57:47 +00:00
|
|
|
if (!file || !buf || size == 0) { /* should never happen */
|
|
|
|
if (ima_appraise & IMA_APPRAISE_ENFORCE)
|
|
|
|
return -EACCES;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
ima: support for kexec image and initramfs
Add IMA policy support for measuring/appraising the kexec image and
initramfs. Two new IMA policy identifiers KEXEC_KERNEL_CHECK and
KEXEC_INITRAMFS_CHECK are defined.
Example policy rules:
measure func=KEXEC_KERNEL_CHECK
appraise func=KEXEC_KERNEL_CHECK appraise_type=imasig
measure func=KEXEC_INITRAMFS_CHECK
appraise func=KEXEC_INITRAMFS_CHECK appraise_type=imasig
Moving the enumeration to the vfs layer simplified the patches, allowing
the IMA changes, for the most part, to be separated from the other
changes. Unfortunately, passing either a kernel_read_file_id or a
ima_hooks enumeration within IMA is messy.
Option 1: duplicate kernel_read_file enumeration in ima_hooks
enum kernel_read_file_id {
...
READING_KEXEC_IMAGE,
READING_KEXEC_INITRAMFS,
READING_MAX_ID
enum ima_hooks {
...
KEXEC_KERNEL_CHECK
KEXEC_INITRAMFS_CHECK
Option 2: define ima_hooks as extension of kernel_read_file
eg: enum ima_hooks {
FILE_CHECK = READING_MAX_ID,
MMAP_CHECK,
In order to pass both kernel_read_file_id and ima_hooks values, we
would need to specify a struct containing a union.
struct caller_id {
union {
enum ima_hooks func_id;
enum kernel_read_file_id read_id;
};
};
Option 3: incorportate the ima_hooks enumeration into kernel_read_file_id,
perhaps changing the enumeration name.
For now, duplicate the new READING_KEXEC_IMAGE/INITRAMFS in the ima_hooks.
Changelog v4:
- replaced switch statement with a kernel_read_file_id to an ima_hooks
id mapping array - Dmitry
- renamed ima_hook tokens KEXEC_CHECK and INITRAMFS_CHECK to
KEXEC_KERNEL_CHECK and KEXEC_INITRAMFS_CHECK respectively - Dave Young
Signed-off-by: Mimi Zohar <zohar@linux.vnet.ibm.com>
Acked-by: Petko Manolov <petkan@mip-labs.com>
Acked-by: Dmitry Kasatkin <dmitry.kasatkin@huawei.com>
Cc: Dave Young <dyoung@redhat.com>
2016-01-15 01:59:14 +00:00
|
|
|
func = read_idmap[read_id] ?: FILE_CHECK;
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2018-01-08 21:36:20 +00:00
|
|
|
return process_measurement(file, current_cred(), secid, buf, size,
|
2018-06-08 17:40:10 +00:00
|
|
|
MAY_READ, func);
|
2014-07-22 14:39:48 +00:00
|
|
|
}
|
|
|
|
|
2018-07-13 18:05:58 +00:00
|
|
|
/**
|
|
|
|
* ima_load_data - appraise decision based on policy
|
|
|
|
* @id: kernel load data caller identifier
|
2020-10-02 17:38:20 +00:00
|
|
|
* @contents: whether the full contents will be available in a later
|
|
|
|
* call to ima_post_load_data().
|
2018-07-13 18:05:58 +00:00
|
|
|
*
|
|
|
|
* Callers of this LSM hook can not measure, appraise, or audit the
|
|
|
|
* data provided by userspace. Enforce policy rules requring a file
|
|
|
|
* signature (eg. kexec'ed kernel image).
|
|
|
|
*
|
|
|
|
* For permission return 0, otherwise return -EACCES.
|
|
|
|
*/
|
2020-10-02 17:38:20 +00:00
|
|
|
int ima_load_data(enum kernel_load_data_id id, bool contents)
|
2018-07-13 18:05:58 +00:00
|
|
|
{
|
2018-10-09 17:30:34 +00:00
|
|
|
bool ima_enforce, sig_enforce;
|
2018-07-13 18:06:02 +00:00
|
|
|
|
2018-10-09 17:30:34 +00:00
|
|
|
ima_enforce =
|
|
|
|
(ima_appraise & IMA_APPRAISE_ENFORCE) == IMA_APPRAISE_ENFORCE;
|
2018-07-13 18:05:58 +00:00
|
|
|
|
|
|
|
switch (id) {
|
|
|
|
case LOADING_KEXEC_IMAGE:
|
2019-08-20 00:17:44 +00:00
|
|
|
if (IS_ENABLED(CONFIG_KEXEC_SIG)
|
2018-10-09 17:30:34 +00:00
|
|
|
&& arch_ima_get_secureboot()) {
|
|
|
|
pr_err("impossible to appraise a kernel image without a file descriptor; try using kexec_file_load syscall.\n");
|
|
|
|
return -EACCES;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ima_enforce && (ima_appraise & IMA_APPRAISE_KEXEC)) {
|
2018-07-13 18:05:58 +00:00
|
|
|
pr_err("impossible to appraise a kernel image without a file descriptor; try using kexec_file_load syscall.\n");
|
|
|
|
return -EACCES; /* INTEGRITY_UNKNOWN */
|
|
|
|
}
|
2018-07-13 18:06:00 +00:00
|
|
|
break;
|
|
|
|
case LOADING_FIRMWARE:
|
2020-10-02 17:38:21 +00:00
|
|
|
if (ima_enforce && (ima_appraise & IMA_APPRAISE_FIRMWARE) && !contents) {
|
2018-07-13 18:06:00 +00:00
|
|
|
pr_err("Prevent firmware sysfs fallback loading.\n");
|
|
|
|
return -EACCES; /* INTEGRITY_UNKNOWN */
|
|
|
|
}
|
2018-07-13 18:06:02 +00:00
|
|
|
break;
|
|
|
|
case LOADING_MODULE:
|
|
|
|
sig_enforce = is_module_sig_enforced();
|
|
|
|
|
2018-10-09 17:30:34 +00:00
|
|
|
if (ima_enforce && (!sig_enforce
|
|
|
|
&& (ima_appraise & IMA_APPRAISE_MODULES))) {
|
2018-07-13 18:06:02 +00:00
|
|
|
pr_err("impossible to appraise a module without a file descriptor. sig_enforce kernel parameter might help\n");
|
|
|
|
return -EACCES; /* INTEGRITY_UNKNOWN */
|
|
|
|
}
|
2020-11-20 18:25:46 +00:00
|
|
|
break;
|
2018-07-13 18:05:58 +00:00
|
|
|
default:
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2020-10-02 17:38:20 +00:00
|
|
|
/**
|
|
|
|
* ima_post_load_data - appraise decision based on policy
|
|
|
|
* @buf: pointer to in memory file contents
|
|
|
|
* @size: size of in memory file contents
|
|
|
|
* @id: kernel load data caller identifier
|
|
|
|
* @description: @id-specific description of contents
|
|
|
|
*
|
|
|
|
* Measure/appraise/audit in memory buffer based on policy. Policy rules
|
|
|
|
* are written in terms of a policy identifier.
|
|
|
|
*
|
|
|
|
* On success return 0. On integrity appraisal error, assuming the file
|
|
|
|
* is in policy and IMA-appraisal is in enforcing mode, return -EACCES.
|
|
|
|
*/
|
|
|
|
int ima_post_load_data(char *buf, loff_t size,
|
|
|
|
enum kernel_load_data_id load_id,
|
|
|
|
char *description)
|
|
|
|
{
|
2020-10-02 17:38:21 +00:00
|
|
|
if (load_id == LOADING_FIRMWARE) {
|
|
|
|
if ((ima_appraise & IMA_APPRAISE_FIRMWARE) &&
|
|
|
|
(ima_appraise & IMA_APPRAISE_ENFORCE)) {
|
|
|
|
pr_err("Prevent firmware loading_store.\n");
|
|
|
|
return -EACCES; /* INTEGRITY_UNKNOWN */
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2020-10-02 17:38:20 +00:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2019-06-24 06:23:29 +00:00
|
|
|
/*
|
2021-01-08 04:07:02 +00:00
|
|
|
* process_buffer_measurement - Measure the buffer or the buffer data hash
|
2021-01-21 13:19:45 +00:00
|
|
|
* @mnt_userns: user namespace of the mount the inode was found from
|
2020-07-09 06:19:11 +00:00
|
|
|
* @inode: inode associated with the object being measured (NULL for KEY_CHECK)
|
2019-06-24 06:23:29 +00:00
|
|
|
* @buf: pointer to the buffer that needs to be added to the log.
|
|
|
|
* @size: size of buffer(in bytes).
|
|
|
|
* @eventname: event name to be used for the buffer entry.
|
2019-10-31 03:31:30 +00:00
|
|
|
* @func: IMA hook
|
|
|
|
* @pcr: pcr to extend the measurement
|
2021-01-08 04:07:01 +00:00
|
|
|
* @func_data: func specific data, may be NULL
|
2021-01-08 04:07:02 +00:00
|
|
|
* @buf_hash: measure buffer data hash
|
2019-06-24 06:23:29 +00:00
|
|
|
*
|
2021-01-08 04:07:02 +00:00
|
|
|
* Based on policy, either the buffer data or buffer data hash is measured
|
2019-06-24 06:23:29 +00:00
|
|
|
*/
|
2021-01-21 13:19:45 +00:00
|
|
|
void process_buffer_measurement(struct user_namespace *mnt_userns,
|
|
|
|
struct inode *inode, const void *buf, int size,
|
2019-10-31 03:31:30 +00:00
|
|
|
const char *eventname, enum ima_hooks func,
|
2021-01-08 04:07:02 +00:00
|
|
|
int pcr, const char *func_data,
|
|
|
|
bool buf_hash)
|
2019-06-24 06:23:29 +00:00
|
|
|
{
|
|
|
|
int ret = 0;
|
2020-06-18 21:10:12 +00:00
|
|
|
const char *audit_cause = "ENOMEM";
|
2019-06-24 06:23:29 +00:00
|
|
|
struct ima_template_entry *entry = NULL;
|
|
|
|
struct integrity_iint_cache iint = {};
|
|
|
|
struct ima_event_data event_data = {.iint = &iint,
|
2019-06-24 06:23:30 +00:00
|
|
|
.filename = eventname,
|
|
|
|
.buf = buf,
|
|
|
|
.buf_len = size};
|
2020-11-12 20:39:59 +00:00
|
|
|
struct ima_template_desc *template;
|
2019-06-24 06:23:29 +00:00
|
|
|
struct {
|
|
|
|
struct ima_digest_data hdr;
|
|
|
|
char digest[IMA_MAX_DIGEST_SIZE];
|
|
|
|
} hash = {};
|
2021-01-08 04:07:02 +00:00
|
|
|
char digest_hash[IMA_MAX_DIGEST_SIZE];
|
|
|
|
int digest_hash_len = hash_digest_size[ima_hash_algo];
|
2019-06-24 06:23:29 +00:00
|
|
|
int violation = 0;
|
|
|
|
int action = 0;
|
2019-10-31 03:31:30 +00:00
|
|
|
u32 secid;
|
2019-06-24 06:23:29 +00:00
|
|
|
|
2019-12-11 16:47:02 +00:00
|
|
|
if (!ima_policy_flag)
|
|
|
|
return;
|
|
|
|
|
2020-11-12 20:39:59 +00:00
|
|
|
template = ima_template_desc_buf();
|
|
|
|
if (!template) {
|
|
|
|
ret = -EINVAL;
|
|
|
|
audit_cause = "ima_template_desc_buf";
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
2019-10-31 03:31:30 +00:00
|
|
|
/*
|
|
|
|
* Both LSM hooks and auxilary based buffer measurements are
|
|
|
|
* based on policy. To avoid code duplication, differentiate
|
|
|
|
* between the LSM hooks and auxilary buffer measurements,
|
|
|
|
* retrieving the policy rule information only for the LSM hook
|
|
|
|
* buffer measurements.
|
|
|
|
*/
|
|
|
|
if (func) {
|
2021-02-19 19:26:21 +00:00
|
|
|
security_task_getsecid_subj(current, &secid);
|
2021-01-21 13:19:45 +00:00
|
|
|
action = ima_get_action(mnt_userns, inode, current_cred(),
|
|
|
|
secid, 0, func, &pcr, &template,
|
idmapped-mounts-v5.12
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYCegywAKCRCRxhvAZXjc
ouJ6AQDlf+7jCQlQdeKKoN9QDFfMzG1ooemat36EpRRTONaGuAD8D9A4sUsG4+5f
4IU5Lj9oY4DEmF8HenbWK2ZHsesL2Qg=
=yPaw
-----END PGP SIGNATURE-----
Merge tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux
Pull idmapped mounts from Christian Brauner:
"This introduces idmapped mounts which has been in the making for some
time. Simply put, different mounts can expose the same file or
directory with different ownership. This initial implementation comes
with ports for fat, ext4 and with Christoph's port for xfs with more
filesystems being actively worked on by independent people and
maintainers.
Idmapping mounts handle a wide range of long standing use-cases. Here
are just a few:
- Idmapped mounts make it possible to easily share files between
multiple users or multiple machines especially in complex
scenarios. For example, idmapped mounts will be used in the
implementation of portable home directories in
systemd-homed.service(8) where they allow users to move their home
directory to an external storage device and use it on multiple
computers where they are assigned different uids and gids. This
effectively makes it possible to assign random uids and gids at
login time.
- It is possible to share files from the host with unprivileged
containers without having to change ownership permanently through
chown(2).
- It is possible to idmap a container's rootfs and without having to
mangle every file. For example, Chromebooks use it to share the
user's Download folder with their unprivileged containers in their
Linux subsystem.
- It is possible to share files between containers with
non-overlapping idmappings.
- Filesystem that lack a proper concept of ownership such as fat can
use idmapped mounts to implement discretionary access (DAC)
permission checking.
- They allow users to efficiently changing ownership on a per-mount
basis without having to (recursively) chown(2) all files. In
contrast to chown (2) changing ownership of large sets of files is
instantenous with idmapped mounts. This is especially useful when
ownership of a whole root filesystem of a virtual machine or
container is changed. With idmapped mounts a single syscall
mount_setattr syscall will be sufficient to change the ownership of
all files.
- Idmapped mounts always take the current ownership into account as
idmappings specify what a given uid or gid is supposed to be mapped
to. This contrasts with the chown(2) syscall which cannot by itself
take the current ownership of the files it changes into account. It
simply changes the ownership to the specified uid and gid. This is
especially problematic when recursively chown(2)ing a large set of
files which is commong with the aforementioned portable home
directory and container and vm scenario.
- Idmapped mounts allow to change ownership locally, restricting it
to specific mounts, and temporarily as the ownership changes only
apply as long as the mount exists.
Several userspace projects have either already put up patches and
pull-requests for this feature or will do so should you decide to pull
this:
- systemd: In a wide variety of scenarios but especially right away
in their implementation of portable home directories.
https://systemd.io/HOME_DIRECTORY/
- container runtimes: containerd, runC, LXD:To share data between
host and unprivileged containers, unprivileged and privileged
containers, etc. The pull request for idmapped mounts support in
containerd, the default Kubernetes runtime is already up for quite
a while now: https://github.com/containerd/containerd/pull/4734
- The virtio-fs developers and several users have expressed interest
in using this feature with virtual machines once virtio-fs is
ported.
- ChromeOS: Sharing host-directories with unprivileged containers.
I've tightly synced with all those projects and all of those listed
here have also expressed their need/desire for this feature on the
mailing list. For more info on how people use this there's a bunch of
talks about this too. Here's just two recent ones:
https://www.cncf.io/wp-content/uploads/2020/12/Rootless-Containers-in-Gitpod.pdf
https://fosdem.org/2021/schedule/event/containers_idmap/
This comes with an extensive xfstests suite covering both ext4 and
xfs:
https://git.kernel.org/brauner/xfstests-dev/h/idmapped_mounts
It covers truncation, creation, opening, xattrs, vfscaps, setid
execution, setgid inheritance and more both with idmapped and
non-idmapped mounts. It already helped to discover an unrelated xfs
setgid inheritance bug which has since been fixed in mainline. It will
be sent for inclusion with the xfstests project should you decide to
merge this.
In order to support per-mount idmappings vfsmounts are marked with
user namespaces. The idmapping of the user namespace will be used to
map the ids of vfs objects when they are accessed through that mount.
By default all vfsmounts are marked with the initial user namespace.
The initial user namespace is used to indicate that a mount is not
idmapped. All operations behave as before and this is verified in the
testsuite.
Based on prior discussions we want to attach the whole user namespace
and not just a dedicated idmapping struct. This allows us to reuse all
the helpers that already exist for dealing with idmappings instead of
introducing a whole new range of helpers. In addition, if we decide in
the future that we are confident enough to enable unprivileged users
to setup idmapped mounts the permission checking can take into account
whether the caller is privileged in the user namespace the mount is
currently marked with.
The user namespace the mount will be marked with can be specified by
passing a file descriptor refering to the user namespace as an
argument to the new mount_setattr() syscall together with the new
MOUNT_ATTR_IDMAP flag. The system call follows the openat2() pattern
of extensibility.
The following conditions must be met in order to create an idmapped
mount:
- The caller must currently have the CAP_SYS_ADMIN capability in the
user namespace the underlying filesystem has been mounted in.
- The underlying filesystem must support idmapped mounts.
- The mount must not already be idmapped. This also implies that the
idmapping of a mount cannot be altered once it has been idmapped.
- The mount must be a detached/anonymous mount, i.e. it must have
been created by calling open_tree() with the OPEN_TREE_CLONE flag
and it must not already have been visible in the filesystem.
The last two points guarantee easier semantics for userspace and the
kernel and make the implementation significantly simpler.
By default vfsmounts are marked with the initial user namespace and no
behavioral or performance changes are observed.
The manpage with a detailed description can be found here:
https://git.kernel.org/brauner/man-pages/c/1d7b902e2875a1ff342e036a9f866a995640aea8
In order to support idmapped mounts, filesystems need to be changed
and mark themselves with the FS_ALLOW_IDMAP flag in fs_flags. The
patches to convert individual filesystem are not very large or
complicated overall as can be seen from the included fat, ext4, and
xfs ports. Patches for other filesystems are actively worked on and
will be sent out separately. The xfstestsuite can be used to verify
that port has been done correctly.
The mount_setattr() syscall is motivated independent of the idmapped
mounts patches and it's been around since July 2019. One of the most
valuable features of the new mount api is the ability to perform
mounts based on file descriptors only.
Together with the lookup restrictions available in the openat2()
RESOLVE_* flag namespace which we added in v5.6 this is the first time
we are close to hardened and race-free (e.g. symlinks) mounting and
path resolution.
While userspace has started porting to the new mount api to mount
proper filesystems and create new bind-mounts it is currently not
possible to change mount options of an already existing bind mount in
the new mount api since the mount_setattr() syscall is missing.
With the addition of the mount_setattr() syscall we remove this last
restriction and userspace can now fully port to the new mount api,
covering every use-case the old mount api could. We also add the
crucial ability to recursively change mount options for a whole mount
tree, both removing and adding mount options at the same time. This
syscall has been requested multiple times by various people and
projects.
There is a simple tool available at
https://github.com/brauner/mount-idmapped
that allows to create idmapped mounts so people can play with this
patch series. I'll add support for the regular mount binary should you
decide to pull this in the following weeks:
Here's an example to a simple idmapped mount of another user's home
directory:
u1001@f2-vm:/$ sudo ./mount --idmap both:1000:1001:1 /home/ubuntu/ /mnt
u1001@f2-vm:/$ ls -al /home/ubuntu/
total 28
drwxr-xr-x 2 ubuntu ubuntu 4096 Oct 28 22:07 .
drwxr-xr-x 4 root root 4096 Oct 28 04:00 ..
-rw------- 1 ubuntu ubuntu 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 ubuntu ubuntu 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 ubuntu ubuntu 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 ubuntu ubuntu 807 Feb 25 2020 .profile
-rw-r--r-- 1 ubuntu ubuntu 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 ubuntu ubuntu 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ ls -al /mnt/
total 28
drwxr-xr-x 2 u1001 u1001 4096 Oct 28 22:07 .
drwxr-xr-x 29 root root 4096 Oct 28 22:01 ..
-rw------- 1 u1001 u1001 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 u1001 u1001 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 u1001 u1001 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 u1001 u1001 807 Feb 25 2020 .profile
-rw-r--r-- 1 u1001 u1001 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 u1001 u1001 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ touch /mnt/my-file
u1001@f2-vm:/$ setfacl -m u:1001:rwx /mnt/my-file
u1001@f2-vm:/$ sudo setcap -n 1001 cap_net_raw+ep /mnt/my-file
u1001@f2-vm:/$ ls -al /mnt/my-file
-rw-rwxr--+ 1 u1001 u1001 0 Oct 28 22:14 /mnt/my-file
u1001@f2-vm:/$ ls -al /home/ubuntu/my-file
-rw-rwxr--+ 1 ubuntu ubuntu 0 Oct 28 22:14 /home/ubuntu/my-file
u1001@f2-vm:/$ getfacl /mnt/my-file
getfacl: Removing leading '/' from absolute path names
# file: mnt/my-file
# owner: u1001
# group: u1001
user::rw-
user:u1001:rwx
group::rw-
mask::rwx
other::r--
u1001@f2-vm:/$ getfacl /home/ubuntu/my-file
getfacl: Removing leading '/' from absolute path names
# file: home/ubuntu/my-file
# owner: ubuntu
# group: ubuntu
user::rw-
user:ubuntu:rwx
group::rw-
mask::rwx
other::r--"
* tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux: (41 commits)
xfs: remove the possibly unused mp variable in xfs_file_compat_ioctl
xfs: support idmapped mounts
ext4: support idmapped mounts
fat: handle idmapped mounts
tests: add mount_setattr() selftests
fs: introduce MOUNT_ATTR_IDMAP
fs: add mount_setattr()
fs: add attr_flags_to_mnt_flags helper
fs: split out functions to hold writers
namespace: only take read lock in do_reconfigure_mnt()
mount: make {lock,unlock}_mount_hash() static
namespace: take lock_mount_hash() directly when changing flags
nfs: do not export idmapped mounts
overlayfs: do not mount on top of idmapped mounts
ecryptfs: do not mount on top of idmapped mounts
ima: handle idmapped mounts
apparmor: handle idmapped mounts
fs: make helpers idmap mount aware
exec: handle idmapped mounts
would_dump: handle idmapped mounts
...
2021-02-23 21:39:45 +00:00
|
|
|
func_data);
|
2019-10-31 03:31:30 +00:00
|
|
|
if (!(action & IMA_MEASURE))
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!pcr)
|
|
|
|
pcr = CONFIG_IMA_MEASURE_PCR_IDX;
|
|
|
|
|
2019-06-24 06:23:29 +00:00
|
|
|
iint.ima_hash = &hash.hdr;
|
|
|
|
iint.ima_hash->algo = ima_hash_algo;
|
|
|
|
iint.ima_hash->length = hash_digest_size[ima_hash_algo];
|
|
|
|
|
|
|
|
ret = ima_calc_buffer_hash(buf, size, iint.ima_hash);
|
2020-06-18 21:10:12 +00:00
|
|
|
if (ret < 0) {
|
|
|
|
audit_cause = "hashing_error";
|
2019-06-24 06:23:29 +00:00
|
|
|
goto out;
|
2020-06-18 21:10:12 +00:00
|
|
|
}
|
2019-06-24 06:23:29 +00:00
|
|
|
|
2021-01-08 04:07:02 +00:00
|
|
|
if (buf_hash) {
|
|
|
|
memcpy(digest_hash, hash.hdr.digest, digest_hash_len);
|
|
|
|
|
|
|
|
ret = ima_calc_buffer_hash(digest_hash, digest_hash_len,
|
|
|
|
iint.ima_hash);
|
|
|
|
if (ret < 0) {
|
|
|
|
audit_cause = "hashing_error";
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
event_data.buf = digest_hash;
|
|
|
|
event_data.buf_len = digest_hash_len;
|
|
|
|
}
|
|
|
|
|
2019-10-31 03:31:30 +00:00
|
|
|
ret = ima_alloc_init_template(&event_data, &entry, template);
|
2020-06-18 21:10:12 +00:00
|
|
|
if (ret < 0) {
|
|
|
|
audit_cause = "alloc_entry";
|
2019-06-24 06:23:29 +00:00
|
|
|
goto out;
|
2020-06-18 21:10:12 +00:00
|
|
|
}
|
2019-06-24 06:23:29 +00:00
|
|
|
|
2021-01-08 04:07:02 +00:00
|
|
|
ret = ima_store_template(entry, violation, NULL, event_data.buf, pcr);
|
2020-06-18 21:10:12 +00:00
|
|
|
if (ret < 0) {
|
|
|
|
audit_cause = "store_entry";
|
2019-06-24 06:23:29 +00:00
|
|
|
ima_free_template_entry(entry);
|
2020-06-18 21:10:12 +00:00
|
|
|
}
|
2019-06-24 06:23:29 +00:00
|
|
|
|
|
|
|
out:
|
2020-02-19 00:06:10 +00:00
|
|
|
if (ret < 0)
|
2020-06-18 21:10:12 +00:00
|
|
|
integrity_audit_message(AUDIT_INTEGRITY_PCR, NULL, eventname,
|
|
|
|
func_measure_str(func),
|
|
|
|
audit_cause, ret, 0, ret);
|
2020-02-19 00:06:10 +00:00
|
|
|
|
2019-06-24 06:23:29 +00:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* ima_kexec_cmdline - measure kexec cmdline boot args
|
2020-07-09 06:19:11 +00:00
|
|
|
* @kernel_fd: file descriptor of the kexec kernel being loaded
|
2019-06-24 06:23:29 +00:00
|
|
|
* @buf: pointer to buffer
|
|
|
|
* @size: size of buffer
|
|
|
|
*
|
|
|
|
* Buffers can only be measured, not appraised.
|
|
|
|
*/
|
2020-07-09 06:19:11 +00:00
|
|
|
void ima_kexec_cmdline(int kernel_fd, const void *buf, int size)
|
2019-06-24 06:23:29 +00:00
|
|
|
{
|
2020-07-09 06:19:11 +00:00
|
|
|
struct fd f;
|
|
|
|
|
|
|
|
if (!buf || !size)
|
|
|
|
return;
|
|
|
|
|
|
|
|
f = fdget(kernel_fd);
|
|
|
|
if (!f.file)
|
|
|
|
return;
|
|
|
|
|
2021-01-21 13:19:45 +00:00
|
|
|
process_buffer_measurement(file_mnt_user_ns(f.file), file_inode(f.file),
|
|
|
|
buf, size, "kexec-cmdline", KEXEC_CMDLINE, 0,
|
idmapped-mounts-v5.12
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYCegywAKCRCRxhvAZXjc
ouJ6AQDlf+7jCQlQdeKKoN9QDFfMzG1ooemat36EpRRTONaGuAD8D9A4sUsG4+5f
4IU5Lj9oY4DEmF8HenbWK2ZHsesL2Qg=
=yPaw
-----END PGP SIGNATURE-----
Merge tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux
Pull idmapped mounts from Christian Brauner:
"This introduces idmapped mounts which has been in the making for some
time. Simply put, different mounts can expose the same file or
directory with different ownership. This initial implementation comes
with ports for fat, ext4 and with Christoph's port for xfs with more
filesystems being actively worked on by independent people and
maintainers.
Idmapping mounts handle a wide range of long standing use-cases. Here
are just a few:
- Idmapped mounts make it possible to easily share files between
multiple users or multiple machines especially in complex
scenarios. For example, idmapped mounts will be used in the
implementation of portable home directories in
systemd-homed.service(8) where they allow users to move their home
directory to an external storage device and use it on multiple
computers where they are assigned different uids and gids. This
effectively makes it possible to assign random uids and gids at
login time.
- It is possible to share files from the host with unprivileged
containers without having to change ownership permanently through
chown(2).
- It is possible to idmap a container's rootfs and without having to
mangle every file. For example, Chromebooks use it to share the
user's Download folder with their unprivileged containers in their
Linux subsystem.
- It is possible to share files between containers with
non-overlapping idmappings.
- Filesystem that lack a proper concept of ownership such as fat can
use idmapped mounts to implement discretionary access (DAC)
permission checking.
- They allow users to efficiently changing ownership on a per-mount
basis without having to (recursively) chown(2) all files. In
contrast to chown (2) changing ownership of large sets of files is
instantenous with idmapped mounts. This is especially useful when
ownership of a whole root filesystem of a virtual machine or
container is changed. With idmapped mounts a single syscall
mount_setattr syscall will be sufficient to change the ownership of
all files.
- Idmapped mounts always take the current ownership into account as
idmappings specify what a given uid or gid is supposed to be mapped
to. This contrasts with the chown(2) syscall which cannot by itself
take the current ownership of the files it changes into account. It
simply changes the ownership to the specified uid and gid. This is
especially problematic when recursively chown(2)ing a large set of
files which is commong with the aforementioned portable home
directory and container and vm scenario.
- Idmapped mounts allow to change ownership locally, restricting it
to specific mounts, and temporarily as the ownership changes only
apply as long as the mount exists.
Several userspace projects have either already put up patches and
pull-requests for this feature or will do so should you decide to pull
this:
- systemd: In a wide variety of scenarios but especially right away
in their implementation of portable home directories.
https://systemd.io/HOME_DIRECTORY/
- container runtimes: containerd, runC, LXD:To share data between
host and unprivileged containers, unprivileged and privileged
containers, etc. The pull request for idmapped mounts support in
containerd, the default Kubernetes runtime is already up for quite
a while now: https://github.com/containerd/containerd/pull/4734
- The virtio-fs developers and several users have expressed interest
in using this feature with virtual machines once virtio-fs is
ported.
- ChromeOS: Sharing host-directories with unprivileged containers.
I've tightly synced with all those projects and all of those listed
here have also expressed their need/desire for this feature on the
mailing list. For more info on how people use this there's a bunch of
talks about this too. Here's just two recent ones:
https://www.cncf.io/wp-content/uploads/2020/12/Rootless-Containers-in-Gitpod.pdf
https://fosdem.org/2021/schedule/event/containers_idmap/
This comes with an extensive xfstests suite covering both ext4 and
xfs:
https://git.kernel.org/brauner/xfstests-dev/h/idmapped_mounts
It covers truncation, creation, opening, xattrs, vfscaps, setid
execution, setgid inheritance and more both with idmapped and
non-idmapped mounts. It already helped to discover an unrelated xfs
setgid inheritance bug which has since been fixed in mainline. It will
be sent for inclusion with the xfstests project should you decide to
merge this.
In order to support per-mount idmappings vfsmounts are marked with
user namespaces. The idmapping of the user namespace will be used to
map the ids of vfs objects when they are accessed through that mount.
By default all vfsmounts are marked with the initial user namespace.
The initial user namespace is used to indicate that a mount is not
idmapped. All operations behave as before and this is verified in the
testsuite.
Based on prior discussions we want to attach the whole user namespace
and not just a dedicated idmapping struct. This allows us to reuse all
the helpers that already exist for dealing with idmappings instead of
introducing a whole new range of helpers. In addition, if we decide in
the future that we are confident enough to enable unprivileged users
to setup idmapped mounts the permission checking can take into account
whether the caller is privileged in the user namespace the mount is
currently marked with.
The user namespace the mount will be marked with can be specified by
passing a file descriptor refering to the user namespace as an
argument to the new mount_setattr() syscall together with the new
MOUNT_ATTR_IDMAP flag. The system call follows the openat2() pattern
of extensibility.
The following conditions must be met in order to create an idmapped
mount:
- The caller must currently have the CAP_SYS_ADMIN capability in the
user namespace the underlying filesystem has been mounted in.
- The underlying filesystem must support idmapped mounts.
- The mount must not already be idmapped. This also implies that the
idmapping of a mount cannot be altered once it has been idmapped.
- The mount must be a detached/anonymous mount, i.e. it must have
been created by calling open_tree() with the OPEN_TREE_CLONE flag
and it must not already have been visible in the filesystem.
The last two points guarantee easier semantics for userspace and the
kernel and make the implementation significantly simpler.
By default vfsmounts are marked with the initial user namespace and no
behavioral or performance changes are observed.
The manpage with a detailed description can be found here:
https://git.kernel.org/brauner/man-pages/c/1d7b902e2875a1ff342e036a9f866a995640aea8
In order to support idmapped mounts, filesystems need to be changed
and mark themselves with the FS_ALLOW_IDMAP flag in fs_flags. The
patches to convert individual filesystem are not very large or
complicated overall as can be seen from the included fat, ext4, and
xfs ports. Patches for other filesystems are actively worked on and
will be sent out separately. The xfstestsuite can be used to verify
that port has been done correctly.
The mount_setattr() syscall is motivated independent of the idmapped
mounts patches and it's been around since July 2019. One of the most
valuable features of the new mount api is the ability to perform
mounts based on file descriptors only.
Together with the lookup restrictions available in the openat2()
RESOLVE_* flag namespace which we added in v5.6 this is the first time
we are close to hardened and race-free (e.g. symlinks) mounting and
path resolution.
While userspace has started porting to the new mount api to mount
proper filesystems and create new bind-mounts it is currently not
possible to change mount options of an already existing bind mount in
the new mount api since the mount_setattr() syscall is missing.
With the addition of the mount_setattr() syscall we remove this last
restriction and userspace can now fully port to the new mount api,
covering every use-case the old mount api could. We also add the
crucial ability to recursively change mount options for a whole mount
tree, both removing and adding mount options at the same time. This
syscall has been requested multiple times by various people and
projects.
There is a simple tool available at
https://github.com/brauner/mount-idmapped
that allows to create idmapped mounts so people can play with this
patch series. I'll add support for the regular mount binary should you
decide to pull this in the following weeks:
Here's an example to a simple idmapped mount of another user's home
directory:
u1001@f2-vm:/$ sudo ./mount --idmap both:1000:1001:1 /home/ubuntu/ /mnt
u1001@f2-vm:/$ ls -al /home/ubuntu/
total 28
drwxr-xr-x 2 ubuntu ubuntu 4096 Oct 28 22:07 .
drwxr-xr-x 4 root root 4096 Oct 28 04:00 ..
-rw------- 1 ubuntu ubuntu 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 ubuntu ubuntu 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 ubuntu ubuntu 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 ubuntu ubuntu 807 Feb 25 2020 .profile
-rw-r--r-- 1 ubuntu ubuntu 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 ubuntu ubuntu 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ ls -al /mnt/
total 28
drwxr-xr-x 2 u1001 u1001 4096 Oct 28 22:07 .
drwxr-xr-x 29 root root 4096 Oct 28 22:01 ..
-rw------- 1 u1001 u1001 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 u1001 u1001 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 u1001 u1001 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 u1001 u1001 807 Feb 25 2020 .profile
-rw-r--r-- 1 u1001 u1001 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 u1001 u1001 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ touch /mnt/my-file
u1001@f2-vm:/$ setfacl -m u:1001:rwx /mnt/my-file
u1001@f2-vm:/$ sudo setcap -n 1001 cap_net_raw+ep /mnt/my-file
u1001@f2-vm:/$ ls -al /mnt/my-file
-rw-rwxr--+ 1 u1001 u1001 0 Oct 28 22:14 /mnt/my-file
u1001@f2-vm:/$ ls -al /home/ubuntu/my-file
-rw-rwxr--+ 1 ubuntu ubuntu 0 Oct 28 22:14 /home/ubuntu/my-file
u1001@f2-vm:/$ getfacl /mnt/my-file
getfacl: Removing leading '/' from absolute path names
# file: mnt/my-file
# owner: u1001
# group: u1001
user::rw-
user:u1001:rwx
group::rw-
mask::rwx
other::r--
u1001@f2-vm:/$ getfacl /home/ubuntu/my-file
getfacl: Removing leading '/' from absolute path names
# file: home/ubuntu/my-file
# owner: ubuntu
# group: ubuntu
user::rw-
user:ubuntu:rwx
group::rw-
mask::rwx
other::r--"
* tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux: (41 commits)
xfs: remove the possibly unused mp variable in xfs_file_compat_ioctl
xfs: support idmapped mounts
ext4: support idmapped mounts
fat: handle idmapped mounts
tests: add mount_setattr() selftests
fs: introduce MOUNT_ATTR_IDMAP
fs: add mount_setattr()
fs: add attr_flags_to_mnt_flags helper
fs: split out functions to hold writers
namespace: only take read lock in do_reconfigure_mnt()
mount: make {lock,unlock}_mount_hash() static
namespace: take lock_mount_hash() directly when changing flags
nfs: do not export idmapped mounts
overlayfs: do not mount on top of idmapped mounts
ecryptfs: do not mount on top of idmapped mounts
ima: handle idmapped mounts
apparmor: handle idmapped mounts
fs: make helpers idmap mount aware
exec: handle idmapped mounts
would_dump: handle idmapped mounts
...
2021-02-23 21:39:45 +00:00
|
|
|
NULL, false);
|
2020-07-09 06:19:11 +00:00
|
|
|
fdput(f);
|
2019-06-24 06:23:29 +00:00
|
|
|
}
|
|
|
|
|
2021-01-08 04:07:03 +00:00
|
|
|
/**
|
|
|
|
* ima_measure_critical_data - measure kernel integrity critical data
|
2021-01-08 04:07:06 +00:00
|
|
|
* @event_label: unique event label for grouping and limiting critical data
|
2021-01-08 04:07:03 +00:00
|
|
|
* @event_name: event name for the record in the IMA measurement list
|
|
|
|
* @buf: pointer to buffer data
|
|
|
|
* @buf_len: length of buffer data (in bytes)
|
|
|
|
* @hash: measure buffer data hash
|
|
|
|
*
|
|
|
|
* Measure data critical to the integrity of the kernel into the IMA log
|
|
|
|
* and extend the pcr. Examples of critical data could be various data
|
|
|
|
* structures, policies, and states stored in kernel memory that can
|
|
|
|
* impact the integrity of the system.
|
|
|
|
*/
|
2021-01-08 04:07:06 +00:00
|
|
|
void ima_measure_critical_data(const char *event_label,
|
|
|
|
const char *event_name,
|
2021-01-08 04:07:03 +00:00
|
|
|
const void *buf, size_t buf_len,
|
|
|
|
bool hash)
|
|
|
|
{
|
2021-01-08 04:07:06 +00:00
|
|
|
if (!event_name || !event_label || !buf || !buf_len)
|
2021-01-08 04:07:03 +00:00
|
|
|
return;
|
|
|
|
|
idmapped-mounts-v5.12
-----BEGIN PGP SIGNATURE-----
iHUEABYKAB0WIQRAhzRXHqcMeLMyaSiRxhvAZXjcogUCYCegywAKCRCRxhvAZXjc
ouJ6AQDlf+7jCQlQdeKKoN9QDFfMzG1ooemat36EpRRTONaGuAD8D9A4sUsG4+5f
4IU5Lj9oY4DEmF8HenbWK2ZHsesL2Qg=
=yPaw
-----END PGP SIGNATURE-----
Merge tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux
Pull idmapped mounts from Christian Brauner:
"This introduces idmapped mounts which has been in the making for some
time. Simply put, different mounts can expose the same file or
directory with different ownership. This initial implementation comes
with ports for fat, ext4 and with Christoph's port for xfs with more
filesystems being actively worked on by independent people and
maintainers.
Idmapping mounts handle a wide range of long standing use-cases. Here
are just a few:
- Idmapped mounts make it possible to easily share files between
multiple users or multiple machines especially in complex
scenarios. For example, idmapped mounts will be used in the
implementation of portable home directories in
systemd-homed.service(8) where they allow users to move their home
directory to an external storage device and use it on multiple
computers where they are assigned different uids and gids. This
effectively makes it possible to assign random uids and gids at
login time.
- It is possible to share files from the host with unprivileged
containers without having to change ownership permanently through
chown(2).
- It is possible to idmap a container's rootfs and without having to
mangle every file. For example, Chromebooks use it to share the
user's Download folder with their unprivileged containers in their
Linux subsystem.
- It is possible to share files between containers with
non-overlapping idmappings.
- Filesystem that lack a proper concept of ownership such as fat can
use idmapped mounts to implement discretionary access (DAC)
permission checking.
- They allow users to efficiently changing ownership on a per-mount
basis without having to (recursively) chown(2) all files. In
contrast to chown (2) changing ownership of large sets of files is
instantenous with idmapped mounts. This is especially useful when
ownership of a whole root filesystem of a virtual machine or
container is changed. With idmapped mounts a single syscall
mount_setattr syscall will be sufficient to change the ownership of
all files.
- Idmapped mounts always take the current ownership into account as
idmappings specify what a given uid or gid is supposed to be mapped
to. This contrasts with the chown(2) syscall which cannot by itself
take the current ownership of the files it changes into account. It
simply changes the ownership to the specified uid and gid. This is
especially problematic when recursively chown(2)ing a large set of
files which is commong with the aforementioned portable home
directory and container and vm scenario.
- Idmapped mounts allow to change ownership locally, restricting it
to specific mounts, and temporarily as the ownership changes only
apply as long as the mount exists.
Several userspace projects have either already put up patches and
pull-requests for this feature or will do so should you decide to pull
this:
- systemd: In a wide variety of scenarios but especially right away
in their implementation of portable home directories.
https://systemd.io/HOME_DIRECTORY/
- container runtimes: containerd, runC, LXD:To share data between
host and unprivileged containers, unprivileged and privileged
containers, etc. The pull request for idmapped mounts support in
containerd, the default Kubernetes runtime is already up for quite
a while now: https://github.com/containerd/containerd/pull/4734
- The virtio-fs developers and several users have expressed interest
in using this feature with virtual machines once virtio-fs is
ported.
- ChromeOS: Sharing host-directories with unprivileged containers.
I've tightly synced with all those projects and all of those listed
here have also expressed their need/desire for this feature on the
mailing list. For more info on how people use this there's a bunch of
talks about this too. Here's just two recent ones:
https://www.cncf.io/wp-content/uploads/2020/12/Rootless-Containers-in-Gitpod.pdf
https://fosdem.org/2021/schedule/event/containers_idmap/
This comes with an extensive xfstests suite covering both ext4 and
xfs:
https://git.kernel.org/brauner/xfstests-dev/h/idmapped_mounts
It covers truncation, creation, opening, xattrs, vfscaps, setid
execution, setgid inheritance and more both with idmapped and
non-idmapped mounts. It already helped to discover an unrelated xfs
setgid inheritance bug which has since been fixed in mainline. It will
be sent for inclusion with the xfstests project should you decide to
merge this.
In order to support per-mount idmappings vfsmounts are marked with
user namespaces. The idmapping of the user namespace will be used to
map the ids of vfs objects when they are accessed through that mount.
By default all vfsmounts are marked with the initial user namespace.
The initial user namespace is used to indicate that a mount is not
idmapped. All operations behave as before and this is verified in the
testsuite.
Based on prior discussions we want to attach the whole user namespace
and not just a dedicated idmapping struct. This allows us to reuse all
the helpers that already exist for dealing with idmappings instead of
introducing a whole new range of helpers. In addition, if we decide in
the future that we are confident enough to enable unprivileged users
to setup idmapped mounts the permission checking can take into account
whether the caller is privileged in the user namespace the mount is
currently marked with.
The user namespace the mount will be marked with can be specified by
passing a file descriptor refering to the user namespace as an
argument to the new mount_setattr() syscall together with the new
MOUNT_ATTR_IDMAP flag. The system call follows the openat2() pattern
of extensibility.
The following conditions must be met in order to create an idmapped
mount:
- The caller must currently have the CAP_SYS_ADMIN capability in the
user namespace the underlying filesystem has been mounted in.
- The underlying filesystem must support idmapped mounts.
- The mount must not already be idmapped. This also implies that the
idmapping of a mount cannot be altered once it has been idmapped.
- The mount must be a detached/anonymous mount, i.e. it must have
been created by calling open_tree() with the OPEN_TREE_CLONE flag
and it must not already have been visible in the filesystem.
The last two points guarantee easier semantics for userspace and the
kernel and make the implementation significantly simpler.
By default vfsmounts are marked with the initial user namespace and no
behavioral or performance changes are observed.
The manpage with a detailed description can be found here:
https://git.kernel.org/brauner/man-pages/c/1d7b902e2875a1ff342e036a9f866a995640aea8
In order to support idmapped mounts, filesystems need to be changed
and mark themselves with the FS_ALLOW_IDMAP flag in fs_flags. The
patches to convert individual filesystem are not very large or
complicated overall as can be seen from the included fat, ext4, and
xfs ports. Patches for other filesystems are actively worked on and
will be sent out separately. The xfstestsuite can be used to verify
that port has been done correctly.
The mount_setattr() syscall is motivated independent of the idmapped
mounts patches and it's been around since July 2019. One of the most
valuable features of the new mount api is the ability to perform
mounts based on file descriptors only.
Together with the lookup restrictions available in the openat2()
RESOLVE_* flag namespace which we added in v5.6 this is the first time
we are close to hardened and race-free (e.g. symlinks) mounting and
path resolution.
While userspace has started porting to the new mount api to mount
proper filesystems and create new bind-mounts it is currently not
possible to change mount options of an already existing bind mount in
the new mount api since the mount_setattr() syscall is missing.
With the addition of the mount_setattr() syscall we remove this last
restriction and userspace can now fully port to the new mount api,
covering every use-case the old mount api could. We also add the
crucial ability to recursively change mount options for a whole mount
tree, both removing and adding mount options at the same time. This
syscall has been requested multiple times by various people and
projects.
There is a simple tool available at
https://github.com/brauner/mount-idmapped
that allows to create idmapped mounts so people can play with this
patch series. I'll add support for the regular mount binary should you
decide to pull this in the following weeks:
Here's an example to a simple idmapped mount of another user's home
directory:
u1001@f2-vm:/$ sudo ./mount --idmap both:1000:1001:1 /home/ubuntu/ /mnt
u1001@f2-vm:/$ ls -al /home/ubuntu/
total 28
drwxr-xr-x 2 ubuntu ubuntu 4096 Oct 28 22:07 .
drwxr-xr-x 4 root root 4096 Oct 28 04:00 ..
-rw------- 1 ubuntu ubuntu 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 ubuntu ubuntu 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 ubuntu ubuntu 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 ubuntu ubuntu 807 Feb 25 2020 .profile
-rw-r--r-- 1 ubuntu ubuntu 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 ubuntu ubuntu 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ ls -al /mnt/
total 28
drwxr-xr-x 2 u1001 u1001 4096 Oct 28 22:07 .
drwxr-xr-x 29 root root 4096 Oct 28 22:01 ..
-rw------- 1 u1001 u1001 3154 Oct 28 22:12 .bash_history
-rw-r--r-- 1 u1001 u1001 220 Feb 25 2020 .bash_logout
-rw-r--r-- 1 u1001 u1001 3771 Feb 25 2020 .bashrc
-rw-r--r-- 1 u1001 u1001 807 Feb 25 2020 .profile
-rw-r--r-- 1 u1001 u1001 0 Oct 16 16:11 .sudo_as_admin_successful
-rw------- 1 u1001 u1001 1144 Oct 28 00:43 .viminfo
u1001@f2-vm:/$ touch /mnt/my-file
u1001@f2-vm:/$ setfacl -m u:1001:rwx /mnt/my-file
u1001@f2-vm:/$ sudo setcap -n 1001 cap_net_raw+ep /mnt/my-file
u1001@f2-vm:/$ ls -al /mnt/my-file
-rw-rwxr--+ 1 u1001 u1001 0 Oct 28 22:14 /mnt/my-file
u1001@f2-vm:/$ ls -al /home/ubuntu/my-file
-rw-rwxr--+ 1 ubuntu ubuntu 0 Oct 28 22:14 /home/ubuntu/my-file
u1001@f2-vm:/$ getfacl /mnt/my-file
getfacl: Removing leading '/' from absolute path names
# file: mnt/my-file
# owner: u1001
# group: u1001
user::rw-
user:u1001:rwx
group::rw-
mask::rwx
other::r--
u1001@f2-vm:/$ getfacl /home/ubuntu/my-file
getfacl: Removing leading '/' from absolute path names
# file: home/ubuntu/my-file
# owner: ubuntu
# group: ubuntu
user::rw-
user:ubuntu:rwx
group::rw-
mask::rwx
other::r--"
* tag 'idmapped-mounts-v5.12' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux: (41 commits)
xfs: remove the possibly unused mp variable in xfs_file_compat_ioctl
xfs: support idmapped mounts
ext4: support idmapped mounts
fat: handle idmapped mounts
tests: add mount_setattr() selftests
fs: introduce MOUNT_ATTR_IDMAP
fs: add mount_setattr()
fs: add attr_flags_to_mnt_flags helper
fs: split out functions to hold writers
namespace: only take read lock in do_reconfigure_mnt()
mount: make {lock,unlock}_mount_hash() static
namespace: take lock_mount_hash() directly when changing flags
nfs: do not export idmapped mounts
overlayfs: do not mount on top of idmapped mounts
ecryptfs: do not mount on top of idmapped mounts
ima: handle idmapped mounts
apparmor: handle idmapped mounts
fs: make helpers idmap mount aware
exec: handle idmapped mounts
would_dump: handle idmapped mounts
...
2021-02-23 21:39:45 +00:00
|
|
|
process_buffer_measurement(&init_user_ns, NULL, buf, buf_len, event_name,
|
2021-01-08 04:07:06 +00:00
|
|
|
CRITICAL_DATA, 0, event_label,
|
2021-01-08 04:07:03 +00:00
|
|
|
hash);
|
|
|
|
}
|
|
|
|
|
2009-02-04 14:06:58 +00:00
|
|
|
static int __init init_ima(void)
|
|
|
|
{
|
|
|
|
int error;
|
|
|
|
|
2020-10-13 08:18:04 +00:00
|
|
|
ima_appraise_parse_cmdline();
|
2016-12-20 00:22:51 +00:00
|
|
|
ima_init_template_list();
|
2013-06-07 10:16:37 +00:00
|
|
|
hash_setup(CONFIG_IMA_DEFAULT_HASH);
|
2009-02-04 14:06:58 +00:00
|
|
|
error = ima_init();
|
2018-03-23 13:41:08 +00:00
|
|
|
|
|
|
|
if (error && strcmp(hash_algo_name[ima_hash_algo],
|
|
|
|
CONFIG_IMA_DEFAULT_HASH) != 0) {
|
|
|
|
pr_info("Allocating %s failed, going to use default hash algorithm %s\n",
|
|
|
|
hash_algo_name[ima_hash_algo], CONFIG_IMA_DEFAULT_HASH);
|
|
|
|
hash_setup_done = 0;
|
|
|
|
hash_setup(CONFIG_IMA_DEFAULT_HASH);
|
|
|
|
error = ima_init();
|
|
|
|
}
|
|
|
|
|
2020-03-25 10:47:07 +00:00
|
|
|
if (error)
|
|
|
|
return error;
|
|
|
|
|
2019-06-14 12:20:15 +00:00
|
|
|
error = register_blocking_lsm_notifier(&ima_lsm_policy_notifier);
|
|
|
|
if (error)
|
|
|
|
pr_warn("Couldn't register LSM notifier, error %d\n", error);
|
|
|
|
|
2018-05-10 15:15:48 +00:00
|
|
|
if (!error)
|
2014-09-12 17:35:54 +00:00
|
|
|
ima_update_policy_flag();
|
2018-05-10 15:15:48 +00:00
|
|
|
|
2009-02-04 14:06:58 +00:00
|
|
|
return error;
|
|
|
|
}
|
|
|
|
|
|
|
|
late_initcall(init_ima); /* Start IMA after the TPM is available */
|