public inbox for linux-security-module@vger.kernel.org
 help / color / mirror / Atom feed
From: "Mickaël Salaün" <mic@digikod.net>
To: "Günther Noack" <gnoack3000@gmail.com>
Cc: John Johansen <john.johansen@canonical.com>,
	 kernel test robot <lkp@intel.com>,
	linux-security-module@vger.kernel.org,
	 Tingmao Wang <m@maowtm.org>,
	Justin Suess <utilityemal77@gmail.com>,
	 Samasth Norway Ananda <samasth.norway.ananda@oracle.com>,
	Matthieu Buffet <matthieu@buffet.re>,
	 Mikhail Ivanov <ivanov.mikhail1@huawei-partners.com>,
	konstantin.meskhidze@huawei.com,
	 Demi Marie Obenour <demiobenour@gmail.com>,
	Alyssa Ross <hi@alyssa.is>, Jann Horn <jannh@google.com>,
	 Tahera Fahimi <fahimitahera@gmail.com>,
	Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
	 Kuniyuki Iwashima <kuniyu@google.com>,
	Georgia Garcia <georgia.garcia@canonical.com>
Subject: Re: [PATCH v8 03/12] landlock: Replace union access_masks_all with helper functions
Date: Mon, 30 Mar 2026 11:56:33 +0200	[thread overview]
Message-ID: <20260330.ohsoe7vu2Nob@digikod.net> (raw)
In-Reply-To: <20260327164838.38231-4-gnoack3000@gmail.com>

On Fri, Mar 27, 2026 at 05:48:28PM +0100, Günther Noack wrote:
> * Stop using a union for access_masks_all.
> * Expose helper functions for intersection checks and union operations.
> 
> The memory layout of bitfields is only loosely defined by the C
> standard, so our static assertion that expects a fixed size was
> brittle, and it broke on some compilers when we attempted to add a
> 17th file system access right.
> 
> Reported-by: kernel test robot <lkp@intel.com>
> Closes: https://lore.kernel.org/oe-kbuild-all/202603261438.jBx2DGNe-lkp@intel.com/
> Signed-off-by: Günther Noack <gnoack3000@gmail.com>
> ---
>  security/landlock/access.h  | 21 ++++++++++++++-------
>  security/landlock/cred.h    | 10 ++--------
>  security/landlock/ruleset.h | 13 ++++---------
>  3 files changed, 20 insertions(+), 24 deletions(-)

I'd prefer this approach:

diff --git a/security/landlock/access.h b/security/landlock/access.h
index 89dc8e7b93da..bc9efbb5c900 100644
--- a/security/landlock/access.h
+++ b/security/landlock/access.h
@@ -50,7 +50,7 @@ struct access_masks {
        access_mask_t fs : LANDLOCK_NUM_ACCESS_FS;
        access_mask_t net : LANDLOCK_NUM_ACCESS_NET;
        access_mask_t scope : LANDLOCK_NUM_SCOPE;
-};
+} __packed;

 union access_masks_all {
        struct access_masks masks;
@@ -58,7 +58,7 @@ union access_masks_all {
 };

 /* Makes sure all fields are covered. */
-static_assert(sizeof(typeof_member(union access_masks_all, masks)) ==
+static_assert(sizeof(typeof_member(union access_masks_all, masks)) <=
              sizeof(typeof_member(union access_masks_all, all)));

 /**


This keep the check and make sure new field will not introduce issues, but more
importantly it save memory, which was one of the goal of struct access_masks.

If that's good with you I'll replace your patch and squash this packed
annotation with the following patch.


> 
> diff --git a/security/landlock/access.h b/security/landlock/access.h
> index 42c95747d7bd..277b6ed7f7bb 100644
> --- a/security/landlock/access.h
> +++ b/security/landlock/access.h
> @@ -52,14 +52,21 @@ struct access_masks {
>  	access_mask_t scope : LANDLOCK_NUM_SCOPE;
>  };
>  
> -union access_masks_all {
> -	struct access_masks masks;
> -	u32 all;
> -};
> +/* Checks whether two access masks have any common bit set. */
> +static inline bool access_masks_intersect(const struct access_masks a,
> +					  const struct access_masks b)
> +{
> +	return (a.fs & b.fs) || (a.net & b.net) || (a.scope & b.scope);
> +}
>  
> -/* Makes sure all fields are covered. */
> -static_assert(sizeof(typeof_member(union access_masks_all, masks)) ==
> -	      sizeof(typeof_member(union access_masks_all, all)));
> +/* ORs the bits of @src into @dst. */
> +static inline void access_masks_merge(struct access_masks *dst,
> +				      const struct access_masks src)
> +{
> +	dst->fs |= src.fs;
> +	dst->net |= src.net;
> +	dst->scope |= src.scope;
> +}
>  
>  /**
>   * struct layer_access_masks - A boolean matrix of layers and access rights
> diff --git a/security/landlock/cred.h b/security/landlock/cred.h
> index f287c56b5fd4..207a6db1c086 100644
> --- a/security/landlock/cred.h
> +++ b/security/landlock/cred.h
> @@ -123,9 +123,6 @@ landlock_get_applicable_subject(const struct cred *const cred,
>  				const struct access_masks masks,
>  				size_t *const handle_layer)
>  {
> -	const union access_masks_all masks_all = {
> -		.masks = masks,
> -	};
>  	const struct landlock_ruleset *domain;
>  	ssize_t layer_level;
>  
> @@ -138,11 +135,8 @@ landlock_get_applicable_subject(const struct cred *const cred,
>  
>  	for (layer_level = domain->num_layers - 1; layer_level >= 0;
>  	     layer_level--) {
> -		union access_masks_all layer = {
> -			.masks = domain->access_masks[layer_level],
> -		};
> -
> -		if (layer.all & masks_all.all) {
> +		if (access_masks_intersect(domain->access_masks[layer_level],
> +					   masks)) {
>  			if (handle_layer)
>  				*handle_layer = layer_level;
>  
> diff --git a/security/landlock/ruleset.h b/security/landlock/ruleset.h
> index 889f4b30301a..9f8b33815c2c 100644
> --- a/security/landlock/ruleset.h
> +++ b/security/landlock/ruleset.h
> @@ -229,18 +229,13 @@ static inline void landlock_get_ruleset(struct landlock_ruleset *const ruleset)
>  static inline struct access_masks
>  landlock_union_access_masks(const struct landlock_ruleset *const domain)
>  {
> -	union access_masks_all matches = {};
> +	struct access_masks matches = {};
>  	size_t layer_level;
>  
> -	for (layer_level = 0; layer_level < domain->num_layers; layer_level++) {
> -		union access_masks_all layer = {
> -			.masks = domain->access_masks[layer_level],
> -		};
> +	for (layer_level = 0; layer_level < domain->num_layers; layer_level++)
> +		access_masks_merge(&matches, domain->access_masks[layer_level]);
>  
> -		matches.all |= layer.all;
> -	}
> -
> -	return matches.masks;
> +	return matches;
>  }
>  
>  static inline void
> -- 
> 2.53.0
> 
> 

  reply	other threads:[~2026-03-30  9:56 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-27 16:48 [PATCH v8 00/12] landlock: UNIX connect() control by pathname and scope Günther Noack
2026-03-27 16:48 ` [PATCH v8 01/12] lsm: Add LSM hook security_unix_find Günther Noack
2026-03-27 17:55   ` Paul Moore
2026-03-27 16:48 ` [PATCH v8 02/12] landlock: Use mem_is_zero() in is_layer_masks_allowed() Günther Noack
2026-03-27 16:48 ` [PATCH v8 03/12] landlock: Replace union access_masks_all with helper functions Günther Noack
2026-03-30  9:56   ` Mickaël Salaün [this message]
2026-03-30 10:53     ` Mickaël Salaün
2026-03-27 16:48 ` [PATCH v8 04/12] landlock: Control pathname UNIX domain socket resolution by path Günther Noack
2026-03-27 16:48 ` [PATCH v8 05/12] landlock: Clarify BUILD_BUG_ON check in scoping logic Günther Noack
2026-03-27 16:48 ` [PATCH v8 06/12] samples/landlock: Add support for named UNIX domain socket restrictions Günther Noack
2026-03-27 16:48 ` [PATCH v8 07/12] selftests/landlock: Replace access_fs_16 with ACCESS_ALL in fs_test Günther Noack
2026-03-27 16:48 ` [PATCH v8 08/12] selftests/landlock: Test LANDLOCK_ACCESS_FS_RESOLVE_UNIX Günther Noack
2026-03-27 16:48 ` [PATCH v8 09/12] selftests/landlock: Audit test for LANDLOCK_ACCESS_FS_RESOLVE_UNIX Günther Noack
2026-03-27 16:48 ` [PATCH v8 10/12] selftests/landlock: Check that coredump sockets stay unrestricted Günther Noack
2026-03-27 16:48 ` [PATCH v8 11/12] selftests/landlock: fs_test: Simplify ruleset creation and enforcement Günther Noack
2026-03-27 16:48 ` [PATCH v8 12/12] landlock: Document FS access right for pathname UNIX sockets Günther Noack

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260330.ohsoe7vu2Nob@digikod.net \
    --to=mic@digikod.net \
    --cc=bigeasy@linutronix.de \
    --cc=demiobenour@gmail.com \
    --cc=fahimitahera@gmail.com \
    --cc=georgia.garcia@canonical.com \
    --cc=gnoack3000@gmail.com \
    --cc=hi@alyssa.is \
    --cc=ivanov.mikhail1@huawei-partners.com \
    --cc=jannh@google.com \
    --cc=john.johansen@canonical.com \
    --cc=konstantin.meskhidze@huawei.com \
    --cc=kuniyu@google.com \
    --cc=linux-security-module@vger.kernel.org \
    --cc=lkp@intel.com \
    --cc=m@maowtm.org \
    --cc=matthieu@buffet.re \
    --cc=samasth.norway.ananda@oracle.com \
    --cc=utilityemal77@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox