linux-integrity.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [RFC] ima: check ima-policy's path
@ 2021-06-29  6:38 Tianxing Zhang
  2021-06-29 11:55 ` Mimi Zohar
  0 siblings, 1 reply; 2+ messages in thread
From: Tianxing Zhang @ 2021-06-29  6:38 UTC (permalink / raw)
  To: zohar; +Cc: linux-integrity

Hi, I was reading the function ima_write_policy in ima/ima_fs.c when I find the issue:

> static ssize_t ima_write_policy(struct file *file, const char __user *buf,
>                                              size_t datalen, loff_t *ppos)
> {
>         ...
>
>         if (data[0] == '/') {
>                 result = ima_read_policy(data);
>         } else if (ima_appraise & IMA_APPRAISE_POLICY) {
>                 pr_err("signed policy file (specified as an absolute pathname) required\n");
>                 integrity_audit_msg(AUDIT_INTEGRITY_STATUS, NULL, NULL,
>                                              "policy_update", "signed policy required",
>                                              1, 0);
>         ...
>         return result;
> }

For the absolute path written by the user, we only check the prefix "/". Actually, we can echo an illegal path to the /sys/kernel/security/ima/policy, e.g. "/\rtest: ddddddddddddddddddd" to inject some logs into dmesg.

Then ima_read_policy is called to return error:

> static ssize_t ima_read_policy(char *path)
> {
>         ...
>         rc = kernel_read_file_from_path(path, 0, &data, INT_MAX, NULL,
>                                                       READING_POLICY);
>         if (rc < 0) {
>                 pr_err("Unable to open file: %s (%d)", path, rc);
>                 return rc;
>         }
>         ...
> }

In ima_read_policy, the illegal path would be logged into dmesg like this:

> ...
> test: ddddddddddddddddddd (-2)/
> test: ddddddddddddddddddd (-2)/
> test: ddddddddddddddddddd (-2)/
> test: ddddddddddddddddddd (-2)/

I suggest that we should check the path in ima_write_policy to make sure it's a valid one, at least literally.

thanks,

Tianxing

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2021-06-29 11:56 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-06-29  6:38 [RFC] ima: check ima-policy's path Tianxing Zhang
2021-06-29 11:55 ` Mimi Zohar

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).