apparmor: move to per loaddata files, instead of replicating in profiles

The loaddata sets cover more than just a single profile and should
be tracked at the ns level. Move the load data files under the namespace
and reference the files from the profiles via a symlink.
Signed-off-by: John Johansen <john.johansen@canonical.com>
Reviewed-by: Seth Arnold <seth.arnold@canonical.com>
Reviewed-by: Kees Cook <keescook@chromium.org>

security/apparmor/apparmorfs.c

@@ -98,14 +98,11 @@ static struct aa_loaddata *aa_simple_write_to_buffer(const char __user *userbuf,
 		return ERR_PTR(-ESPIPE);
 
 	/* freed by caller to simple_write_to_buffer */
-	data = kvmalloc(sizeof(*data) + alloc_size, GFP_KERNEL);
-	if (data == NULL)
-		return ERR_PTR(-ENOMEM);
-	kref_init(&data->count);
-	data->size = copy_size;
-	data->hash = NULL;
-	data->abi = 0;
+	data = aa_loaddata_alloc(alloc_size);
+	if (IS_ERR(data))
+		return data;
 
+	data->size = copy_size;
 	if (copy_from_user(data->data, userbuf, copy_size)) {
 		kvfree(data);
 		return ERR_PTR(-EFAULT);
@@ -213,6 +210,11 @@ static const struct file_operations aa_fs_profile_remove = {
 	.llseek = default_llseek,
 };
 
+void __aa_bump_ns_revision(struct aa_ns *ns)
+{
+	ns->revision++;
+}
+
 /**
  * query_data - queries a policy and writes its data to buf
  * @buf: the resulting data is stored here (NOT NULL)
@@ -559,68 +561,88 @@ static const struct file_operations aa_fs_ns_name = {
 	.release	= single_release,
 };
 
-static int rawdata_release(struct inode *inode, struct file *file)
+
+/* policy/raw_data/ * file ops */
+
+#define SEQ_RAWDATA_FOPS(NAME)						      \
+static int seq_rawdata_ ##NAME ##_open(struct inode *inode, struct file *file)\
+{									      \
+	return seq_rawdata_open(inode, file, seq_rawdata_ ##NAME ##_show);    \
+}									      \
+									      \
+static const struct file_operations seq_rawdata_ ##NAME ##_fops = {	      \
+	.owner		= THIS_MODULE,					      \
+	.open		= seq_rawdata_ ##NAME ##_open,			      \
+	.read		= seq_read,					      \
+	.llseek		= seq_lseek,					      \
+	.release	= seq_rawdata_release,				      \
+}									      \
+
+static int seq_rawdata_open(struct inode *inode, struct file *file,
+			    int (*show)(struct seq_file *, void *))
 {
-	/* TODO: switch to loaddata when profile switched to symlink */
-	aa_put_loaddata(file->private_data);
+	struct aa_loaddata *data = __aa_get_loaddata(inode->i_private);
+	int error;
 
-	return 0;
+	if (!data)
+		/* lost race this ent is being reaped */
+		return -ENOENT;
+
+	error = single_open(file, show, data);
+	if (error) {
+		AA_BUG(file->private_data &&
+		       ((struct seq_file *)file->private_data)->private);
+		aa_put_loaddata(data);
+	}
+
+	return error;
 }
 
-static int aa_fs_seq_raw_abi_show(struct seq_file *seq, void *v)
+static int seq_rawdata_release(struct inode *inode, struct file *file)
 {
-	struct aa_proxy *proxy = seq->private;
-	struct aa_profile *profile = aa_get_profile_rcu(&proxy->profile);
+	struct seq_file *seq = (struct seq_file *) file->private_data;
+
+	if (seq)
+		aa_put_loaddata(seq->private);
 
-	if (profile->rawdata->abi)
-		seq_printf(seq, "v%d\n", profile->rawdata->abi);
+	return single_release(inode, file);
+}
 
-	aa_put_profile(profile);
+static int seq_rawdata_abi_show(struct seq_file *seq, void *v)
+{
+	struct aa_loaddata *data = seq->private;
+
+	seq_printf(seq, "v%d\n", data->abi);
 
 	return 0;
 }
 
-static int aa_fs_seq_raw_abi_open(struct inode *inode, struct file *file)
+static int seq_rawdata_revision_show(struct seq_file *seq, void *v)
 {
-	return aa_fs_seq_profile_open(inode, file, aa_fs_seq_raw_abi_show);
-}
+	struct aa_loaddata *data = seq->private;
 
-static const struct file_operations aa_fs_seq_raw_abi_fops = {
-	.owner		= THIS_MODULE,
-	.open		= aa_fs_seq_raw_abi_open,
-	.read		= seq_read,
-	.llseek		= seq_lseek,
-	.release	= aa_fs_seq_profile_release,
-};
+	seq_printf(seq, "%ld\n", data->revision);
 
-static int aa_fs_seq_raw_hash_show(struct seq_file *seq, void *v)
+	return 0;
+}
+
+static int seq_rawdata_hash_show(struct seq_file *seq, void *v)
 {
-	struct aa_proxy *proxy = seq->private;
-	struct aa_profile *profile = aa_get_profile_rcu(&proxy->profile);
+	struct aa_loaddata *data = seq->private;
 	unsigned int i, size = aa_hash_size();
 
-	if (profile->rawdata->hash) {
+	if (data->hash) {
 		for (i = 0; i < size; i++)
-			seq_printf(seq, "%.2x", profile->rawdata->hash[i]);
+			seq_printf(seq, "%.2x", data->hash[i]);
 		seq_putc(seq, '\n');
 	}
-	aa_put_profile(profile);
 
 	return 0;
 }
 
-static int aa_fs_seq_raw_hash_open(struct inode *inode, struct file *file)
-{
-	return aa_fs_seq_profile_open(inode, file, aa_fs_seq_raw_hash_show);
-}
-
-static const struct file_operations aa_fs_seq_raw_hash_fops = {
-	.owner		= THIS_MODULE,
-	.open		= aa_fs_seq_raw_hash_open,
-	.read		= seq_read,
-	.llseek		= seq_lseek,
-	.release	= aa_fs_seq_profile_release,
-};
+SEQ_RAWDATA_FOPS(abi);
+SEQ_RAWDATA_FOPS(revision);
+SEQ_RAWDATA_FOPS(hash);
 
 static ssize_t rawdata_read(struct file *file, char __user *buf, size_t size,
 			    loff_t *ppos)
@@ -631,27 +653,120 @@ static ssize_t rawdata_read(struct file *file, char __user *buf, size_t size,
 				       rawdata->size);
 }
 
-static int rawdata_open(struct inode *inode, struct file *file)
+static int rawdata_release(struct inode *inode, struct file *file)
 {
-	struct aa_proxy *proxy = inode->i_private;
-	struct aa_profile *profile;
+	aa_put_loaddata(file->private_data);
+
+	return 0;
+}
 
+static int rawdata_open(struct inode *inode, struct file *file)
+{
 	if (!policy_view_capable(NULL))
 		return -EACCES;
-	profile = aa_get_profile_rcu(&proxy->profile);
-	file->private_data = aa_get_loaddata(profile->rawdata);
-	aa_put_profile(profile);
+	file->private_data = __aa_get_loaddata(inode->i_private);
+	if (!file->private_data)
+		/* lost race: this entry is being reaped */
+		return -ENOENT;
 
 	return 0;
 }
 
-static const struct file_operations aa_fs_rawdata_fops = {
+static const struct file_operations rawdata_fops = {
 	.open = rawdata_open,
 	.read = rawdata_read,
 	.llseek = generic_file_llseek,
 	.release = rawdata_release,
 };
 
+static void remove_rawdata_dents(struct aa_loaddata *rawdata)
+{
+	int i;
+
+	for (i = 0; i < AAFS_LOADDATA_NDENTS; i++) {
+		if (!IS_ERR_OR_NULL(rawdata->dents[i])) {
+			/* no refcounts on i_private */
+			securityfs_remove(rawdata->dents[i]);
+			rawdata->dents[i] = NULL;
+		}
+	}
+}
+
+void __aa_fs_remove_rawdata(struct aa_loaddata *rawdata)
+{
+	AA_BUG(rawdata->ns && !mutex_is_locked(&rawdata->ns->lock));
+
+	if (rawdata->ns) {
+		remove_rawdata_dents(rawdata);
+		list_del_init(&rawdata->list);
+		aa_put_ns(rawdata->ns);
+		rawdata->ns = NULL;
+	}
+}
+
+int __aa_fs_create_rawdata(struct aa_ns *ns, struct aa_loaddata *rawdata)
+{
+	struct dentry *dent, *dir;
+
+	AA_BUG(!ns);
+	AA_BUG(!rawdata);
+	AA_BUG(!mutex_is_locked(&ns->lock));
+	AA_BUG(!ns_subdata_dir(ns));
+
+	/*
+	 * just use ns revision dir was originally created at. This is
+	 * under ns->lock and if load is successful revision will be
+	 * bumped and is guaranteed to be unique
+	 */
+	rawdata->name = kasprintf(GFP_KERNEL, "%ld", ns->revision);
+	if (!rawdata->name)
+		return -ENOMEM;
+
+	dir = securityfs_create_dir(rawdata->name, ns_subdata_dir(ns));
+	if (IS_ERR(dir))
+		/* ->name freed when rawdata freed */
+		return PTR_ERR(dir);
+	rawdata->dents[AAFS_LOADDATA_DIR] = dir;
+
+	dent = securityfs_create_file("abi", S_IFREG | 0444, dir, rawdata,
+				      &seq_rawdata_abi_fops);
+	if (IS_ERR(dent))
+		goto fail;
+	rawdata->dents[AAFS_LOADDATA_ABI] = dent;
+
+	dent = securityfs_create_file("revision", S_IFREG | 0444, dir, rawdata,
+				      &seq_rawdata_revision_fops);
+	if (IS_ERR(dent))
+		goto fail;
+	rawdata->dents[AAFS_LOADDATA_REVISION] = dent;
+
+	if (aa_g_hash_policy) {
+		dent = securityfs_create_file("sha1", S_IFREG | 0444, dir,
+					      rawdata, &seq_rawdata_hash_fops);
+		if (IS_ERR(dent))
+			goto fail;
+		rawdata->dents[AAFS_LOADDATA_HASH] = dent;
+	}
+
+	dent = securityfs_create_file("raw_data", S_IFREG | 0444,
+				      dir, rawdata, &rawdata_fops);
+	if (IS_ERR(dent))
+		goto fail;
+	rawdata->dents[AAFS_LOADDATA_DATA] = dent;
+	d_inode(dent)->i_size = rawdata->size;
+
+	rawdata->ns = aa_get_ns(ns);
+	list_add(&rawdata->list, &ns->rawdata_list);
+	/* no refcount on inode rawdata */
+
+	return 0;
+
+fail:
+	remove_rawdata_dents(rawdata);
+
+	return PTR_ERR(dent);
+}
+
 /** fns to setup dynamic per profile/namespace files **/
 void __aa_fs_profile_rmdir(struct aa_profile *profile)
 {
@@ -703,7 +818,41 @@ static struct dentry *create_profile_file(struct dentry *dir, const char *name,
 	return dent;
 }
 
-/* requires lock be held */
+static int profile_depth(struct aa_profile *profile)
+{
+	int depth = 0;
+
+	rcu_read_lock();
+	for (depth = 0; profile; profile = rcu_access_pointer(profile->parent))
+		depth++;
+	rcu_read_unlock();
+
+	return depth;
+}
+
+static int gen_symlink_name(char *buffer, size_t bsize, int depth,
+			    const char *dirname, const char *fname)
+{
+	int error;
+
+	for (; depth > 0; depth--) {
+		if (bsize < 7)
+			return -ENAMETOOLONG;
+		strcpy(buffer, "../../");
+		buffer += 6;
+		bsize -= 6;
+	}
+
+	error = snprintf(buffer, bsize, "raw_data/%s/%s", dirname, fname);
+	if (error >= bsize || error < 0)
+		return -ENAMETOOLONG;
+
+	return 0;
+}
+
+/*
+ * Requires: @profile->ns->lock held
+ */
 int __aa_fs_profile_mkdir(struct aa_profile *profile, struct dentry *parent)
 {
 	struct aa_profile *child;
@@ -766,26 +915,35 @@ int __aa_fs_profile_mkdir(struct aa_profile *profile, struct dentry *parent)
 	}
 
 	if (profile->rawdata) {
-		dent = create_profile_file(dir, "raw_sha1", profile,
-					   &aa_fs_seq_raw_hash_fops);
+		char target[64];
+		int depth = profile_depth(profile);
+
+		error = gen_symlink_name(target, sizeof(target), depth,
+					 profile->rawdata->name, "sha1");
+		if (error < 0)
+			goto fail2;
+		dent = securityfs_create_symlink("raw_sha1", dir, target, NULL);
 		if (IS_ERR(dent))
 			goto fail;
 		profile->dents[AAFS_PROF_RAW_HASH] = dent;
 
-		dent = create_profile_file(dir, "raw_abi", profile,
-					   &aa_fs_seq_raw_abi_fops);
+		error = gen_symlink_name(target, sizeof(target), depth,
+					 profile->rawdata->name, "abi");
+		if (error < 0)
+			goto fail2;
+		dent = securityfs_create_symlink("raw_abi", dir, target, NULL);
 		if (IS_ERR(dent))
 			goto fail;
 		profile->dents[AAFS_PROF_RAW_ABI] = dent;
 
-		dent = securityfs_create_file("raw_data", S_IFREG | 0444, dir,
-					      profile->proxy,
-					      &aa_fs_rawdata_fops);
+		error = gen_symlink_name(target, sizeof(target), depth,
+					 profile->rawdata->name, "raw_data");
+		if (error < 0)
+			goto fail2;
+		dent = securityfs_create_symlink("raw_data", dir, target, NULL);
 		if (IS_ERR(dent))
 			goto fail;
 		profile->dents[AAFS_PROF_RAW_DATA] = dent;
-		d_inode(dent)->i_size = profile->rawdata->size;
-		aa_get_proxy(profile->proxy);
 	}
 
 	list_for_each_entry(child, &profile->base.profiles, base.list) {
@@ -805,6 +963,16 @@ int __aa_fs_profile_mkdir(struct aa_profile *profile, struct dentry *parent)
 	return error;
 }
 
+static void __aa_fs_list_remove_rawdata(struct aa_ns *ns)
+{
+	struct aa_loaddata *ent, *tmp;
+
+	AA_BUG(!mutex_is_locked(&ns->lock));
+
+	list_for_each_entry_safe(ent, tmp, &ns->rawdata_list, list)
+		__aa_fs_remove_rawdata(ent);
+}
+
 void __aa_fs_ns_rmdir(struct aa_ns *ns)
 {
 	struct aa_ns *sub;
@@ -823,6 +991,8 @@ void __aa_fs_ns_rmdir(struct aa_ns *ns)
 		mutex_unlock(&sub->lock);
 	}
 
+	__aa_fs_list_remove_rawdata(ns);
+
 	if (ns_subns_dir(ns)) {
 		sub = d_inode(ns_subns_dir(ns))->i_private;
 		aa_put_ns(sub);

security/apparmor/include/apparmorfs.h

@@ -106,6 +106,7 @@ enum aafs_prof_type {
 #define prof_dir(X) ((X)->dents[AAFS_PROF_DIR])
 #define prof_child_dir(X) ((X)->dents[AAFS_PROF_PROFS])
 
+void __aa_bump_ns_revision(struct aa_ns *ns);
 void __aa_fs_profile_rmdir(struct aa_profile *profile);
 void __aa_fs_profile_migrate_dents(struct aa_profile *old,
 				   struct aa_profile *new);
@@ -114,4 +115,8 @@ void __aa_fs_ns_rmdir(struct aa_ns *ns);
 int __aa_fs_ns_mkdir(struct aa_ns *ns, struct dentry *parent,
 		     const char *name);
 
+struct aa_loaddata;
+void __aa_fs_remove_rawdata(struct aa_loaddata *rawdata);
+int __aa_fs_create_rawdata(struct aa_ns *ns, struct aa_loaddata *rawdata);
+
 #endif /* __AA_APPARMORFS_H */

security/apparmor/include/policy_ns.h

@@ -68,6 +68,9 @@ struct aa_ns {
 	atomic_t uniq_null;
 	long uniq_id;
 	int level;
+	long revision;
+
+	struct list_head rawdata_list;
 
 	struct dentry *dents[AAFS_NS_SIZEOF];
 };

security/apparmor/include/policy_unpack.h

@@ -17,6 +17,8 @@
 
 #include <linux/list.h>
 #include <linux/kref.h>
+#include <linux/dcache.h>
+#include <linux/workqueue.h>
 
 struct aa_load_ent {
 	struct list_head list;
@@ -36,26 +38,84 @@ struct aa_load_ent *aa_load_ent_alloc(void);
 #define PACKED_MODE_KILL	2
 #define PACKED_MODE_UNCONFINED	3
 
-/* struct aa_loaddata - buffer of policy load data set */
+struct aa_ns;
+
+enum {
+	AAFS_LOADDATA_ABI = 0,
+	AAFS_LOADDATA_REVISION,
+	AAFS_LOADDATA_HASH,
+	AAFS_LOADDATA_DATA,
+	AAFS_LOADDATA_DIR,		/* must be last actual entry */
+	AAFS_LOADDATA_NDENTS		/* count of entries */
+};
+
+/*
+ * struct aa_loaddata - buffer of policy raw_data set
+ *
+ * there is no loaddata ref for being on ns list, nor a ref from
+ * d_inode(@dentry) when grab a ref from these, @ns->lock must be held
+ * && __aa_get_loaddata() needs to be used, and the return value
+ * checked, if NULL the loaddata is already being reaped and should be
+ * considered dead.
+ */
 struct aa_loaddata {
 	struct kref count;
+	struct list_head list;
+	struct work_struct work;
+	struct dentry *dents[AAFS_LOADDATA_NDENTS];
+	struct aa_ns *ns;
+	char *name;
 	size_t size;
+	long revision;			/* the ns policy revision this caused */
 	int abi;
 	unsigned char *hash;
+
 	char data[];
 };
 
 int aa_unpack(struct aa_loaddata *udata, struct list_head *lh, const char **ns);
 
+/**
+ * __aa_get_loaddata - get a reference count to uncounted data reference
+ * @data: reference to get a count on
+ *
+ * Returns: pointer to reference OR NULL if race is lost and reference is
+ *          being repeated.
+ * Requires: @data->ns->lock held, and the return code MUST be checked
+ *
+ * Use only from inode->i_private and @data->list found references
+ */
+static inline struct aa_loaddata *
+__aa_get_loaddata(struct aa_loaddata *data)
+{
+	if (data && kref_get_unless_zero(&(data->count)))
+		return data;
+
+	return NULL;
+}
+
+/**
+ * aa_get_loaddata - get a reference count from a counted data reference
+ * @data: reference to get a count on
+ *
+ * Returns: point to reference
+ * Requires: @data to have a valid reference count on it. It is a bug
+ *           if the race to reap can be encountered when it is used.
+ */
 static inline struct aa_loaddata *
 aa_get_loaddata(struct aa_loaddata *data)
 {
-	if (data)
-		kref_get(&(data->count));
-	return data;
+	struct aa_loaddata *tmp = __aa_get_loaddata(data);
+
+	AA_BUG(data && !tmp);
+
+	return tmp;
 }
 
+void __aa_loaddata_update(struct aa_loaddata *data, long revision);
+bool aa_rawdata_eq(struct aa_loaddata *l, struct aa_loaddata *r);
 void aa_loaddata_kref(struct kref *kref);
+struct aa_loaddata *aa_loaddata_alloc(size_t size);
 static inline void aa_put_loaddata(struct aa_loaddata *data)
 {
 	if (data)

security/apparmor/policy.c

@@ -840,10 +840,13 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 	const char *ns_name, *info = NULL;
 	struct aa_ns *ns = NULL;
 	struct aa_load_ent *ent, *tmp;
+	struct aa_loaddata *rawdata_ent;
 	const char *op = OP_PROF_REPL;
 	ssize_t count, error;
+
 	LIST_HEAD(lh);
 
+	aa_get_loaddata(udata);
 	/* released below */
 	error = aa_unpack(udata, &lh, &ns_name);
 	if (error)
@@ -887,9 +890,24 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 		ns = aa_get_ns(view);
 
 	mutex_lock(&ns->lock);
+	/* check for duplicate rawdata blobs: space and file dedup */
+	list_for_each_entry(rawdata_ent, &ns->rawdata_list, list) {
+		if (aa_rawdata_eq(rawdata_ent, udata)) {
+			struct aa_loaddata *tmp;
+
+			tmp = __aa_get_loaddata(rawdata_ent);
+			/* check we didn't fail the race */
+			if (tmp) {
+				aa_put_loaddata(udata);
+				udata = tmp;
+				break;
+			}
+		}
+	}
 	/* setup parent and ns info */
 	list_for_each_entry(ent, &lh, list) {
 		struct aa_policy *policy;
+
 		ent->new->rawdata = aa_get_loaddata(udata);
 		error = __lookup_replace(ns, ent->new->base.hname, noreplace,
 					 &ent->old, &info);
@@ -929,6 +947,14 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 	}
 
 	/* create new fs entries for introspection if needed */
+	if (!udata->dents[AAFS_LOADDATA_DIR]) {
+		error = __aa_fs_create_rawdata(ns, udata);
+		if (error) {
+			info = "failed to create raw_data dir and files";
+			ent = NULL;
+			goto fail_lock;
+		}
+	}
 	list_for_each_entry(ent, &lh, list) {
 		if (ent->old) {
 			/* inherit old interface files */
@@ -955,10 +981,24 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 	}
 
 	/* Done with checks that may fail - do actual replacement */
+	__aa_bump_ns_revision(ns);
+	__aa_loaddata_update(udata, ns->revision);
 	list_for_each_entry_safe(ent, tmp, &lh, list) {
 		list_del_init(&ent->list);
 		op = (!ent->old && !ent->rename) ? OP_PROF_LOAD : OP_PROF_REPL;
 
+		if (ent->old && ent->old->rawdata == ent->new->rawdata) {
+			/* dedup actual profile replacement */
+			audit_policy(profile, op, ns_name, ent->new->base.hname,
+				     "same as current profile, skipping",
+				     error);
+			goto skip;
+		}
+
+		/*
+		 * TODO: finer dedup based on profile range in data. Load set
+		 * can differ but profile may remain unchanged
+		 */
 		audit_policy(profile, op, NULL, ent->new->base.hname,
 			     NULL, error);
 
@@ -998,12 +1038,14 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 					   aa_get_profile(ent->new));
 			__list_add_profile(&ns->base.profiles, ent->new);
 		}
+	skip:
 		aa_load_ent_free(ent);
 	}
 	mutex_unlock(&ns->lock);
 
 out:
 	aa_put_ns(ns);
+	aa_put_loaddata(udata);
 
 	if (error)
 		return error;
@@ -1013,7 +1055,7 @@ ssize_t aa_replace_profiles(struct aa_ns *view, struct aa_profile *profile,
 	mutex_unlock(&ns->lock);
 
 	/* audit cause of failure */
-	op = (!ent->old) ? OP_PROF_LOAD : OP_PROF_REPL;
+	op = (ent && !ent->old) ? OP_PROF_LOAD : OP_PROF_REPL;
 fail:
 	audit_policy(profile, op, ns_name, ent ? ent->new->base.hname : NULL,
 		     info, error);
@@ -1085,6 +1127,7 @@ ssize_t aa_remove_profiles(struct aa_ns *view, struct aa_profile *subj,
 		/* remove namespace - can only happen if fqname[0] == ':' */
 		mutex_lock(&ns->parent->lock);
 		__aa_remove_ns(ns);
+		__aa_bump_ns_revision(ns);
 		mutex_unlock(&ns->parent->lock);
 	} else {
 		/* remove profile */
@@ -1097,6 +1140,7 @@ ssize_t aa_remove_profiles(struct aa_ns *view, struct aa_profile *subj,
 		}
 		name = profile->base.hname;
 		__remove_profile(profile);
+		__aa_bump_ns_revision(ns);
 		mutex_unlock(&ns->lock);
 	}
 

security/apparmor/policy_ns.c

@@ -99,6 +99,7 @@ static struct aa_ns *alloc_ns(const char *prefix, const char *name)
 		goto fail_ns;
 
 	INIT_LIST_HEAD(&ns->sub_ns);
+	INIT_LIST_HEAD(&ns->rawdata_list);
 	mutex_init(&ns->lock);
 
 	/* released by aa_free_ns() */

security/apparmor/policy_unpack.c

@@ -122,16 +122,73 @@ static int audit_iface(struct aa_profile *new, const char *ns_name,
 	return aa_audit(AUDIT_APPARMOR_STATUS, profile, &sa, audit_cb);
 }
 
+void __aa_loaddata_update(struct aa_loaddata *data, long revision)
+{
+	AA_BUG(!data);
+	AA_BUG(!data->ns);
+	AA_BUG(!data->dents[AAFS_LOADDATA_REVISION]);
+	AA_BUG(!mutex_is_locked(&data->ns->lock));
+	AA_BUG(data->revision > revision);
+
+	data->revision = revision;
+	d_inode(data->dents[AAFS_LOADDATA_DIR])->i_mtime =
+		current_time(d_inode(data->dents[AAFS_LOADDATA_DIR]));
+	d_inode(data->dents[AAFS_LOADDATA_REVISION])->i_mtime =
+		current_time(d_inode(data->dents[AAFS_LOADDATA_REVISION]));
+}
+
+bool aa_rawdata_eq(struct aa_loaddata *l, struct aa_loaddata *r)
+{
+	if (l->size != r->size)
+		return false;
+	if (aa_g_hash_policy && memcmp(l->hash, r->hash, aa_hash_size()) != 0)
+		return false;
+	return memcmp(l->data, r->data, r->size) == 0;
+}
+
+/*
+ * need to take the ns mutex lock which is NOT safe most places that
+ * put_loaddata is called, so we have to delay freeing it
+ */
+static void do_loaddata_free(struct work_struct *work)
+{
+	struct aa_loaddata *d = container_of(work, struct aa_loaddata, work);
+	struct aa_ns *ns = aa_get_ns(d->ns);
+
+	if (ns) {
+		mutex_lock(&ns->lock);
+		__aa_fs_remove_rawdata(d);
+		mutex_unlock(&ns->lock);
+		aa_put_ns(ns);
+	}
+
+	kzfree(d->hash);
+	kfree(d->name);
+	kvfree(d);
+}
+
 void aa_loaddata_kref(struct kref *kref)
 {
 	struct aa_loaddata *d = container_of(kref, struct aa_loaddata, count);
 
 	if (d) {
-		kzfree(d->hash);
-		kvfree(d);
+		INIT_WORK(&d->work, do_loaddata_free);
+		schedule_work(&d->work);
 	}
 }
 
+struct aa_loaddata *aa_loaddata_alloc(size_t size)
+{
+	struct aa_loaddata *d = kvzalloc(sizeof(*d) + size, GFP_KERNEL);
+
+	if (d == NULL)
+		return ERR_PTR(-ENOMEM);
+	kref_init(&d->count);
+	INIT_LIST_HEAD(&d->list);
+
+	return d;
+}
+
 /* test if read will be in packed data bounds */
 static bool inbounds(struct aa_ext *e, size_t size)
 {