[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20221018150213.7n4sv7rtsh6lshd5@wittgenstein>
Date: Tue, 18 Oct 2022 17:02:13 +0200
From: Christian Brauner <brauner@...nel.org>
To: Kees Cook <keescook@...omium.org>
Cc: Mimi Zohar <zohar@...ux.ibm.com>,
Dmitry Kasatkin <dmitry.kasatkin@...il.com>,
Paul Moore <paul@...l-moore.com>,
James Morris <jmorris@...ei.org>,
"Serge E. Hallyn" <serge@...lyn.com>, Petr Vorel <pvorel@...e.cz>,
Jonathan McDowell <noodles@...com>,
Borislav Petkov <bp@...e.de>, Takashi Iwai <tiwai@...e.de>,
linux-integrity@...r.kernel.org,
linux-security-module@...r.kernel.org,
Mickaël Salaün <mic@...ikod.net>,
KP Singh <kpsingh@...nel.org>,
Casey Schaufler <casey@...aufler-ca.com>,
John Johansen <john.johansen@...onical.com>,
linux-kernel@...r.kernel.org, linux-hardening@...r.kernel.org
Subject: Re: [PATCH 4/9] ima: Move ima_file_free() into LSM
On Thu, Oct 13, 2022 at 03:36:49PM -0700, Kees Cook wrote:
> The file_free_security hook already exists for managing notification of
> released files. Use the LSM hook instead of open-coded stacking.
>
> Cc: Mimi Zohar <zohar@...ux.ibm.com>
> Cc: Dmitry Kasatkin <dmitry.kasatkin@...il.com>
> Cc: Paul Moore <paul@...l-moore.com>
> Cc: James Morris <jmorris@...ei.org>
> Cc: "Serge E. Hallyn" <serge@...lyn.com>
> Cc: Petr Vorel <pvorel@...e.cz>
> Cc: Jonathan McDowell <noodles@...com>
> Cc: Borislav Petkov <bp@...e.de>
> Cc: Takashi Iwai <tiwai@...e.de>
> Cc: linux-integrity@...r.kernel.org
> Cc: linux-security-module@...r.kernel.org
> Signed-off-by: Kees Cook <keescook@...omium.org>
> ---
> fs/file_table.c | 1 -
> include/linux/ima.h | 6 ------
> security/integrity/ima/ima_main.c | 3 ++-
> 3 files changed, 2 insertions(+), 8 deletions(-)
>
> diff --git a/fs/file_table.c b/fs/file_table.c
> index 99c6796c9f28..fa707d221a43 100644
> --- a/fs/file_table.c
> +++ b/fs/file_table.c
> @@ -311,7 +311,6 @@ static void __fput(struct file *file)
> eventpoll_release(file);
> locks_remove_file(file);
>
> - ima_file_free(file);
> if (unlikely(file->f_flags & FASYNC)) {
> if (file->f_op->fasync)
> file->f_op->fasync(-1, file, 0);
> diff --git a/include/linux/ima.h b/include/linux/ima.h
> index 6dc5143f89f2..9f18df366064 100644
> --- a/include/linux/ima.h
> +++ b/include/linux/ima.h
> @@ -19,7 +19,6 @@ extern enum hash_algo ima_get_current_hash_algo(void);
> extern int ima_file_check(struct file *file, int mask);
> extern void ima_post_create_tmpfile(struct user_namespace *mnt_userns,
> struct inode *inode);
> -extern void ima_file_free(struct file *file);
> extern void ima_post_path_mknod(struct user_namespace *mnt_userns,
> struct dentry *dentry);
> extern int ima_file_hash(struct file *file, char *buf, size_t buf_size);
> @@ -56,11 +55,6 @@ static inline void ima_post_create_tmpfile(struct user_namespace *mnt_userns,
> {
> }
>
> -static inline void ima_file_free(struct file *file)
> -{
> - return;
> -}
> -
> static inline void ima_post_path_mknod(struct user_namespace *mnt_userns,
> struct dentry *dentry)
> {
> diff --git a/security/integrity/ima/ima_main.c b/security/integrity/ima/ima_main.c
> index b3b79d030a67..94379ba40b58 100644
> --- a/security/integrity/ima/ima_main.c
> +++ b/security/integrity/ima/ima_main.c
> @@ -183,7 +183,7 @@ static void ima_check_last_writer(struct integrity_iint_cache *iint,
> *
> * Flag files that changed, based on i_version
> */
> -void ima_file_free(struct file *file)
> +static void ima_file_free(struct file *file)
> {
> struct inode *inode = file_inode(file);
> struct integrity_iint_cache *iint;
> @@ -1085,6 +1085,7 @@ static struct security_hook_list ima_hooks[] __lsm_ro_after_init = {
> LSM_HOOK_INIT(bprm_check_security, ima_bprm_check),
> LSM_HOOK_INIT(mmap_file, ima_file_mmap),
> LSM_HOOK_INIT(file_mprotect, ima_file_mprotect),
> + LSM_HOOK_INIT(file_free_security, ima_file_free),
This doesn't work afaict. If the file is opened for writing ima may
update xattrs. But by the time security_file_free() is called
put_file_access() has already been called which will have given up write
access to the file's mount.
So you would have to - just one of the possibilities - have to move
security_file_free() out of file_free() and into the old ima_file_free()
location. But that might cause semantic changes for other LSMs.
Powered by blists - more mailing lists