On 2/7/2025 9:48 AM, Stefan Berger wrote:

On 2/7/25 12:06 PM, Mimi Zohar wrote:

quoted

On Fri, 2025-02-07 at 10:16 -0500, Mimi Zohar wrote:

quoted

On Mon, 2025-02-03 at 15:20 -0800, steven chen wrote:

quoted

The amount of memory allocated at kexec load, even with the extra 
memory
allocated, might not be large enough for the entire measurement 
list.  The
indeterminate interval between kexec 'load' and 'execute' could 
exacerbate
this problem.

Define two new IMA events, 'kexec_load' and 'kexec_execute', to be
measured as critical data at kexec 'load' and 'execute' respectively.
Report the allocated kexec segment size, IMA binary log size and the
runtime measurements count as part of those events.

These events, and the values reported through them, serve as 
markers in
the IMA log to verify the IMA events are captured during kexec soft
reboot.  The presence of a 'kexec_load' event in between the last two
'boot_aggregate' events in the IMA log implies this is a kexec soft
reboot, and not a cold-boot. And the absence of 'kexec_execute' event
after kexec soft reboot implies missing events in that window which
results in inconsistency with TPM PCR quotes, necessitating a cold 
boot
for a successful remote attestation.

As a reminder, please include directions for verifying the buffer 
data hash against
the buffer data.  The directions would be similar to those in commit 
6b4da8c0e7f
("IMA: Define a new template field buf").

quoted

Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Author: Tushar Sugandhi [off-list ref]
Signed-off-by: Tushar Sugandhi <redacted>
Signed-off-by: steven chen <redacted>
---
  security/integrity/ima/ima_kexec.c | 23 +++++++++++++++++++++++
  1 file changed, 23 insertions(+)

diff --git a/security/integrity/ima/ima_kexec.c

b/security/integrity/ima/ima_kexec.c
index c9c916f69ca7..0342ddfa9342 100644

--- a/security/integrity/ima/ima_kexec.c
+++ b/security/integrity/ima/ima_kexec.c

@@ -17,6 +17,8 @@

  #include "ima.h"
    #ifdef CONFIG_IMA_KEXEC
+#define IMA_KEXEC_EVENT_LEN 256
+
  static struct seq_file ima_kexec_file;
  static void *ima_kexec_buffer;
  static size_t kexec_segment_size;

@@ -36,6 +38,24 @@ static void ima_free_kexec_file_buf(struct 

seq_file *sf)
      ima_reset_kexec_file(sf);
  }
  +static void ima_measure_kexec_event(const char *event_name)
+{
+    char ima_kexec_event[IMA_KEXEC_EVENT_LEN];
+    size_t buf_size = 0;
+    long len;
+
+    buf_size = ima_get_binary_runtime_size();
+    len = atomic_long_read(&ima_htable.len);
+
+    scnprintf(ima_kexec_event, IMA_KEXEC_EVENT_LEN,
+ "kexec_segment_size=%lu;ima_binary_runtime_size=%lu;"
+            "ima_runtime_measurements_count=%ld;",
+                kexec_segment_size, buf_size, len);

 From scripts/checkpatch.pl, "Alignment should match open parenthesis".

quoted

+
+    ima_measure_critical_data("ima_kexec", event_name, 
ima_kexec_event,
+                    strlen(ima_kexec_event), false, NULL,
0);

 From the kernel-doc scnprintf(), returns the number of bytes.  
There should be no
need to calculate it using strlen().

quoted

+}
+
  static int ima_alloc_kexec_file_buf(size_t segment_size)
  {
      /*

@@ -60,6 +80,7 @@ static int ima_alloc_kexec_file_buf(size_t 

segment_size)
  out:
      ima_kexec_file.read_pos = 0;
      ima_kexec_file.count = sizeof(struct ima_kexec_hdr);    /* 
reserved
space
*/
+    ima_measure_kexec_event("kexec_load");
        return 0;
  }

@@ -201,6 +222,8 @@ static int ima_update_kexec_buffer(struct 

notifier_block
*self,
          return ret;
      }
  +    ima_measure_kexec_event("kexec_execute");
+
      ret = ima_dump_measurement_list(&buf_size, &buf,
                      kexec_segment_size);

After fixing up and applying this patch set to 6.14.0-rc1, I'm not 
seeing the
"kexec_execute".  Even after changing the default extra memory, I'm 
still not
seeing
the measurement.

FYI, after reverting commit 254ef9541d68 ("ima: Suspend PCR extends 
and log appends
when rebooting"), I'm seeing the "kexec_execute" measurement.

I would try sth. like this:
static int ima_reboot_notifier(struct notifier_block *nb,
                   unsigned long action,
                   void *data)
{
    if (action == SYS_RESTART && data && !strcmp(data, "kexec reboot"))
        ima_measure_kexec_event("kexec_execute");

quoted

Mimi

Thanks, will fix this in 6.14.0-rc1 in next version