Re: [RFC PATCH v2 1/2] ima: limit the number of open-writers integrity violations

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On 2/4/25 10:32 PM, Mimi Zohar wrote:
Limit the number of open-writers integrity violation audit messages
and records in the IMA measurement list emitted when re-opening a
file for read.

.. to only show one violation. (?)


Signed-off-by: Mimi Zohar <zohar@xxxxxxxxxxxxx>
---
  security/integrity/ima/ima.h      |  1 +
  security/integrity/ima/ima_main.c | 14 ++++++++++++--
  2 files changed, 13 insertions(+), 2 deletions(-)

diff --git a/security/integrity/ima/ima.h b/security/integrity/ima/ima.h
index a4f284bd846c..7f21568544dd 100644
--- a/security/integrity/ima/ima.h
+++ b/security/integrity/ima/ima.h
@@ -182,6 +182,7 @@ struct ima_kexec_hdr {
  #define IMA_CHANGE_ATTR		2
  #define IMA_DIGSIG		3
  #define IMA_MUST_MEASURE	4
+#define IMA_LIMIT_VIOLATIONS	5
/* IMA integrity metadata associated with an inode */
  struct ima_iint_cache {
diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c
index 28b8b0db6f9b..5091ad931677 100644
--- a/security/integrity/ima/ima_main.c
+++ b/security/integrity/ima/ima_main.c
@@ -137,8 +137,16 @@ static void ima_rdwr_violation_check(struct file *file,
  	} else {
  		if (must_measure)
  			set_bit(IMA_MUST_MEASURE, &iint->atomic_flags);

iint is assumed to be here under, which is correct when must_measure is set since how the single caller of this function calls it.

-		if (inode_is_open_for_write(inode) && must_measure)
-			send_writers = true;
+
+		if (inode_is_open_for_write(inode) && must_measure) {
+			if (!iint)
+				iint = ima_iint_find(inode);

When must_measure is set, the case of iint == NULL should never occur here.

+
+			/* Limit number of open_writers violations */
+			if (iint && !test_and_set_bit(IMA_LIMIT_VIOLATIONS,
+						      &iint->atomic_flags))
if (!test_and_set_bit(...))


+				send_writers = true;
+		}
  	}
if (!send_tomtou && !send_writers)
@@ -167,6 +175,8 @@ static void ima_check_last_writer(struct ima_iint_cache *iint,
  	if (atomic_read(&inode->i_writecount) == 1) {
  		struct kstat stat;
+ clear_bit(IMA_LIMIT_VIOLATIONS, &iint->atomic_flags);
+
  		update = test_and_clear_bit(IMA_UPDATE_XATTR,
  					    &iint->atomic_flags);
  		if ((iint->flags & IMA_NEW_FILE) ||

I tested the two patches. For the file /usr/bin/foobar that's being measured on FILE_CHECK I get two violations for this sequence here:

# exec 3>/usr/bin/foobar
# exec 4</usr/bin/foobar
# exec 3>/usr/bin/foobar

This here now only emits one violation now.

# exec 3>/usr/bin/foobar
# exec 4</usr/bin/foobar
# exec 4</usr/bin/foobar
# exec 5</usr/bin/foobar
# exec 6</usr/bin/foobar






[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Index of Archives]     [Linux Kernel]     [Linux Kernel Hardening]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux SCSI]

  Powered by Linux