Backports the following two patches to fix the issue of IMA mishandling LSM based rule during LSM policy update, causing a file to match an unexpected rule.
Some changes were made to these patches, which was stated in the commit message of corresponding patch.
GUO Zihua (1): ima: Handle -ESTALE returned by ima_filter_rule_match()
Janne Karhunen (1): ima: use the lsm policy update notifier
security/integrity/ima/ima.h | 2 + security/integrity/ima/ima_main.c | 8 ++ security/integrity/ima/ima_policy.c | 146 +++++++++++++++++++++++----- 3 files changed, 130 insertions(+), 26 deletions(-)
From: Janne Karhunen janne.karhunen@gmail.com
[ Upstream commit b169424551930a9325f700f502802f4d515194e5 ]
This patch is backported to resolve the issue of IMA ignoreing LSM part of an LSM based rule. As the LSM notifier chain was an atomic notifier chain, we'll not be able to call synchronize_rcu() within our notifier handling function.
Original patch message is as follows:
Don't do lazy policy updates while running the rule matching, run the updates as they happen.
Depends on commit f242064c5df3 ("LSM: switch to blocking policy update notifiers")
Signed-off-by: Janne Karhunen janne.karhunen@gmail.com Signed-off-by: Mimi Zohar zohar@linux.ibm.com Cc: stable@vger.kernel.org #4.19.y Signed-off-by: GUO Zihua guozihua@huawei.com --- security/integrity/ima/ima.h | 2 + security/integrity/ima/ima_main.c | 8 ++ security/integrity/ima/ima_policy.c | 115 +++++++++++++++++++++++----- 3 files changed, 105 insertions(+), 20 deletions(-)
diff --git a/security/integrity/ima/ima.h b/security/integrity/ima/ima.h index e2916b115b93..dc564ed9a790 100644 --- a/security/integrity/ima/ima.h +++ b/security/integrity/ima/ima.h @@ -154,6 +154,8 @@ int ima_measurements_show(struct seq_file *m, void *v); unsigned long ima_get_binary_runtime_size(void); int ima_init_template(void); void ima_init_template_list(void); +int ima_lsm_policy_change(struct notifier_block *nb, unsigned long event, + void *lsm_data);
/* * used to protect h_table and sha_table diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c index 2d31921fbda4..f461b3e2de00 100644 --- a/security/integrity/ima/ima_main.c +++ b/security/integrity/ima/ima_main.c @@ -41,6 +41,10 @@ int ima_appraise; int ima_hash_algo = HASH_ALGO_SHA1; static int hash_setup_done;
+static struct notifier_block ima_lsm_policy_notifier = { + .notifier_call = ima_lsm_policy_change, +}; + static int __init hash_setup(char *str) { struct ima_template_desc *template_desc = ima_template_desc_current(); @@ -553,6 +557,10 @@ static int __init init_ima(void) error = ima_init(); }
+ error = register_lsm_notifier(&ima_lsm_policy_notifier); + if (error) + pr_warn("Couldn't register LSM notifier, error %d\n", error); + if (!error) ima_update_policy_flag();
diff --git a/security/integrity/ima/ima_policy.c b/security/integrity/ima/ima_policy.c index b2dadff3626b..1e0251e9510a 100644 --- a/security/integrity/ima/ima_policy.c +++ b/security/integrity/ima/ima_policy.c @@ -256,31 +256,112 @@ static void ima_free_rule(struct ima_rule_entry *entry) kfree(entry); }
+static void ima_lsm_free_rule(struct ima_rule_entry *entry) +{ + int i; + + for (i = 0; i < MAX_LSM_RULES; i++) { + kfree(entry->lsm[i].rule); + kfree(entry->lsm[i].args_p); + } + kfree(entry); +} + +static struct ima_rule_entry *ima_lsm_copy_rule(struct ima_rule_entry *entry) +{ + struct ima_rule_entry *nentry; + int i, result; + + nentry = kmalloc(sizeof(*nentry), GFP_KERNEL); + if (!nentry) + return NULL; + + /* + * Immutable elements are copied over as pointers and data; only + * lsm rules can change + */ + memcpy(nentry, entry, sizeof(*nentry)); + memset(nentry->lsm, 0, FIELD_SIZEOF(struct ima_rule_entry, lsm)); + + for (i = 0; i < MAX_LSM_RULES; i++) { + if (!entry->lsm[i].rule) + continue; + + nentry->lsm[i].type = entry->lsm[i].type; + nentry->lsm[i].args_p = kstrdup(entry->lsm[i].args_p, + GFP_KERNEL); + if (!nentry->lsm[i].args_p) + goto out_err; + + result = security_filter_rule_init(nentry->lsm[i].type, + Audit_equal, + nentry->lsm[i].args_p, + &nentry->lsm[i].rule); + if (result == -EINVAL) + pr_warn("ima: rule for LSM '%d' is undefined\n", + entry->lsm[i].type); + } + return nentry; + +out_err: + ima_lsm_free_rule(nentry); + return NULL; +} + +static int ima_lsm_update_rule(struct ima_rule_entry *entry) +{ + struct ima_rule_entry *nentry; + + nentry = ima_lsm_copy_rule(entry); + if (!nentry) + return -ENOMEM; + + list_replace_rcu(&entry->list, &nentry->list); + ima_lsm_free_rule(entry); + + return 0; +} + /* * The LSM policy can be reloaded, leaving the IMA LSM based rules referring * to the old, stale LSM policy. Update the IMA LSM based rules to reflect - * the reloaded LSM policy. We assume the rules still exist; and BUG_ON() if - * they don't. + * the reloaded LSM policy. */ static void ima_lsm_update_rules(void) { - struct ima_rule_entry *entry; - int result; - int i; + struct ima_rule_entry *entry, *e; + int i, result, needs_update;
- list_for_each_entry(entry, &ima_policy_rules, list) { + list_for_each_entry_safe(entry, e, &ima_policy_rules, list) { + needs_update = 0; for (i = 0; i < MAX_LSM_RULES; i++) { - if (!entry->lsm[i].rule) - continue; - result = security_filter_rule_init(entry->lsm[i].type, - Audit_equal, - entry->lsm[i].args_p, - &entry->lsm[i].rule); - BUG_ON(!entry->lsm[i].rule); + if (entry->lsm[i].rule) { + needs_update = 1; + break; + } + } + if (!needs_update) + continue; + + result = ima_lsm_update_rule(entry); + if (result) { + pr_err("ima: lsm rule update error %d\n", + result); + return; } } }
+int ima_lsm_policy_change(struct notifier_block *nb, unsigned long event, + void *lsm_data) +{ + if (event != LSM_POLICY_CHANGE) + return NOTIFY_DONE; + + ima_lsm_update_rules(); + return NOTIFY_OK; +} + /** * ima_match_rules - determine whether an inode matches the measure rule. * @rule: a pointer to a rule @@ -334,11 +415,10 @@ static bool ima_match_rules(struct ima_rule_entry *rule, struct inode *inode, for (i = 0; i < MAX_LSM_RULES; i++) { int rc = 0; u32 osid; - int retried = 0;
if (!rule->lsm[i].rule) continue; -retry: + switch (i) { case LSM_OBJ_USER: case LSM_OBJ_ROLE: @@ -361,11 +441,6 @@ static bool ima_match_rules(struct ima_rule_entry *rule, struct inode *inode, default: break; } - if ((rc < 0) && (!retried)) { - retried = 1; - ima_lsm_update_rules(); - goto retry; - } if (!rc) return false; }
[ Upstream commit c7423dbdbc9ecef7fff5239d144cad4b9887f4de ]
IMA relies on the blocking LSM policy notifier callback to update the LSM based IMA policy rules.
When SELinux update its policies, IMA would be notified and starts updating all its lsm rules one-by-one. During this time, -ESTALE would be returned by ima_filter_rule_match() if it is called with a LSM rule that has not yet been updated. In ima_match_rules(), -ESTALE is not handled, and the LSM rule is considered a match, causing extra files to be measured by IMA.
Fix it by re-initializing a temporary rule if -ESTALE is returned by ima_filter_rule_match(). The origin rule in the rule list would be updated by the LSM policy notifier callback.
Fixes: b16942455193 ("ima: use the lsm policy update notifier") Signed-off-by: GUO Zihua guozihua@huawei.com Reviewed-by: Roberto Sassu roberto.sassu@huawei.com Signed-off-by: Mimi Zohar zohar@linux.ibm.com Cc: stable@vger.kernel.org # 4.19.y --- security/integrity/ima/ima_policy.c | 34 ++++++++++++++++++++++------- 1 file changed, 26 insertions(+), 8 deletions(-)
diff --git a/security/integrity/ima/ima_policy.c b/security/integrity/ima/ima_policy.c index 1e0251e9510a..dd146a34a53a 100644 --- a/security/integrity/ima/ima_policy.c +++ b/security/integrity/ima/ima_policy.c @@ -378,6 +378,9 @@ static bool ima_match_rules(struct ima_rule_entry *rule, struct inode *inode, enum ima_hooks func, int mask) { int i; + bool result = false; + struct ima_rule_entry *lsm_rule = rule; + bool rule_reinitialized = false;
if ((rule->flags & IMA_FUNC) && (rule->func != func && func != POST_SETATTR)) @@ -416,35 +419,50 @@ static bool ima_match_rules(struct ima_rule_entry *rule, struct inode *inode, int rc = 0; u32 osid;
- if (!rule->lsm[i].rule) + if (!lsm_rule->lsm[i].rule) continue;
+retry: switch (i) { case LSM_OBJ_USER: case LSM_OBJ_ROLE: case LSM_OBJ_TYPE: security_inode_getsecid(inode, &osid); rc = security_filter_rule_match(osid, - rule->lsm[i].type, + lsm_rule->lsm[i].type, Audit_equal, - rule->lsm[i].rule, + lsm_rule->lsm[i].rule, NULL); break; case LSM_SUBJ_USER: case LSM_SUBJ_ROLE: case LSM_SUBJ_TYPE: rc = security_filter_rule_match(secid, - rule->lsm[i].type, + lsm_rule->lsm[i].type, Audit_equal, - rule->lsm[i].rule, + lsm_rule->lsm[i].rule, NULL); default: break; } - if (!rc) - return false; + if (rc == -ESTALE && !rule_reinitialized) { + lsm_rule = ima_lsm_copy_rule(rule); + if (lsm_rule) { + rule_reinitialized = true; + goto retry; + } + } + if (!rc) { + result = false; + goto out; + } } - return true; + result = true; + +out: + if (rule_reinitialized) + ima_lsm_free_rule(lsm_rule); + return result; }
/*
On Tue, Dec 27, 2022 at 09:47:29AM +0800, GUO Zihua wrote:
[ Upstream commit c7423dbdbc9ecef7fff5239d144cad4b9887f4de ]
For obvious reasons we can not only take this patch (from 6.2-rc1) into 4.19.y as that would cause people who upgrade from 4.19.y to a newer stable kernel to have a regression.
Please submit backports for all stable kernels if you wish to see this in older ones to prevent problems like this from happening.
But also, why are you still on 4.19.y? What is wrong with 5.4.y at this point in time? If we dropped support for 4.19.y in January, what would that cause to happen for your systems?
thanks,
greg k-h
On Tue, 2022-12-27 at 08:37 +0100, Greg KH wrote:
On Tue, Dec 27, 2022 at 09:47:29AM +0800, GUO Zihua wrote:
[ Upstream commit c7423dbdbc9ecef7fff5239d144cad4b9887f4de ]
For obvious reasons we can not only take this patch (from 6.2-rc1) into 4.19.y as that would cause people who upgrade from 4.19.y to a newer stable kernel to have a regression.
Please submit backports for all stable kernels if you wish to see this in older ones to prevent problems like this from happening.
Sasha has already queued the original commit and the dependencies for the 6.1, 6.0, and 5.15 stable kernels. Those kernels all had the call_lsm_notifier() to call_blocking_lsm_notifier() change. Prior to 5.3, the change to the blocking notifier would need to be backported as well. This version of the backport still needs to be reviewed.
thanks,
Mimi
But also, why are you still on 4.19.y? What is wrong with 5.4.y at this point in time? If we dropped support for 4.19.y in January, what would that cause to happen for your systems?
thanks,
greg k-h
Hi Greg and Mimi.
Fall sick for a couple days.
On 2022/12/27 19:56, Mimi Zohar wrote:
On Tue, 2022-12-27 at 08:37 +0100, Greg KH wrote:
On Tue, Dec 27, 2022 at 09:47:29AM +0800, GUO Zihua wrote:
[ Upstream commit c7423dbdbc9ecef7fff5239d144cad4b9887f4de ]
For obvious reasons we can not only take this patch (from 6.2-rc1) into 4.19.y as that would cause people who upgrade from 4.19.y to a newer stable kernel to have a regression.
Please submit backports for all stable kernels if you wish to see this in older ones to prevent problems like this from happening.
Sasha has already queued the original commit and the dependencies for the 6.1, 6.0, and 5.15 stable kernels. Those kernels all had the call_lsm_notifier() to call_blocking_lsm_notifier() change. Prior to 5.3, the change to the blocking notifier would need to be backported as well. This version of the backport still needs to be reviewed.
Indeed the current solution needs further testing and review. One of the concern raised by Huaxin is a possible UAF caused by the call to free rule in update_rule. Will it be possible to backport also the change which turn call_lsm_notifier() into call_blocking_lsm_notifier()?
thanks,
Mimi
But also, why are you still on 4.19.y? What is wrong with 5.4.y at this point in time? If we dropped support for 4.19.y in January, what would that cause to happen for your systems?
Well it's all about backward compatibility. We still got some products using the 4.19.y LTS kernel and we would still needs to provide support for this version of the kernel. If 4.19.y got EOL or EOS in January next year, our company surely would develop corresponding plans to handle that change.
thanks,
greg k-h
linux-stable-mirror@lists.linaro.org