From: "Serge E. Hallyn" <serge@hallyn.com>
To: "Mickaël Salaün" <mic@digikod.net>
Cc: "Serge Hallyn" <sergeh@kernel.org>,
"Paul Moore" <paul@paul-moore.com>,
"Günther Noack" <gnoack@google.com>,
"John Johansen" <john.johansen@canonical.com>,
"KP Singh" <kpsingh@kernel.org>,
"Matt Bobrowski" <matt@bobrowski.net>,
"Stephen Smalley" <stephen.smalley.work@gmail.com>,
linux-security-module@vger.kernel.org,
kernel-team@cloudflare.com
Subject: Re: [PATCH v1] lsm: Set PF_SUPERPRIV only when all mmap_addr hooks allow a mapping
Date: Fri, 20 Nov 2026 12:49:43 -0600 [thread overview]
Message-ID: <awCWRzL9Y2uEm9pp@hallyn.com> (raw)
In-Reply-To: <20261118165603.2573289-1-mic@digikod.net>
On Wed, Nov 18, 2026 at 05:56:02PM +0100, Mickaël Salaün wrote:
> cap_mmap_addr() sets PF_SUPERPRIV when commoncap grants CAP_SYS_RAWIO
> for a mapping below vm.mmap_min_addr, before the other LSMs' mmap_addr
> hooks run. If one of them then denies the mapping (e.g. SELinux's
> mmap_zero check), process accounting still reports a use of privileges.
>
> Set the flag in security_mmap_addr() once all hooks allowed the mapping,
> as ns_capable() does after security_capable(). commoncap runs first and
> denies such a mapping without CAP_SYS_RAWIO, so a successful hook chain
> means that the capability was used.
>
> Cc: Paul Moore <paul@paul-moore.com>
> Cc: Serge Hallyn <serge@hallyn.com>
Reviewed-by: Serge Hallyn <serge@hallyn.com>
Although I have two suggestions that you can take or leave - first, I
think rc == 0 would be nicer than !rc. Second, you could keep the
PF_SUPERPRIV setting in one place by having security_mmap_addr defined
only once in security.h, and having it call do_security_mmap_addr(),
which changes based on whether CONFIG_SECURITY is set. Otherwise I
could see the two spots getting out of sync after some future update.
> Cc: Stephen Smalley <stephen.smalley.work@gmail.com>
> Reported-by: Günther Noack <gnoack@google.com>
> Closes: https://patch.msgid.link/avxjfHybQBcATFKV@google.com
> Fixes: 9c0d90103c7e ("Capabilities: move cap_file_mmap to commoncap.c")
> Signed-off-by: Mickaël Salaün <mic@digikod.net>
> ---
> include/linux/security.h | 6 +++++-
> security/commoncap.c | 6 +-----
> security/security.c | 7 ++++++-
> 3 files changed, 12 insertions(+), 7 deletions(-)
>
> diff --git a/include/linux/security.h b/include/linux/security.h
> index 1e89bc1fec17..faf118b58ade 100644
> --- a/include/linux/security.h
> +++ b/include/linux/security.h
> @@ -1209,7 +1209,11 @@ static inline int security_mmap_backing_file(struct vm_area_struct *vma,
>
> static inline int security_mmap_addr(unsigned long addr)
> {
> - return cap_mmap_addr(addr);
> + int rc = cap_mmap_addr(addr);
> +
> + if (!rc && addr < dac_mmap_min_addr)
> + current->flags |= PF_SUPERPRIV;
> + return rc;
> }
>
> static inline int security_file_mprotect(struct vm_area_struct *vma,
> diff --git a/security/commoncap.c b/security/commoncap.c
> index 9794a594bdd9..36c7b394867d 100644
> --- a/security/commoncap.c
> +++ b/security/commoncap.c
> @@ -1471,13 +1471,9 @@ int cap_mmap_addr(unsigned long addr)
> {
> int ret = 0;
>
> - if (addr < dac_mmap_min_addr) {
> + if (addr < dac_mmap_min_addr)
> ret = cap_capable(current_cred(), &init_user_ns, CAP_SYS_RAWIO,
> CAP_OPT_NONE);
> - /* set PF_SUPERPRIV if it turns out we allow the low mmap */
> - if (ret == 0)
> - current->flags |= PF_SUPERPRIV;
> - }
> return ret;
> }
>
> diff --git a/security/security.c b/security/security.c
> index 36365fac5801..fd6717de31ad 100644
> --- a/security/security.c
> +++ b/security/security.c
> @@ -2687,7 +2687,12 @@ EXPORT_SYMBOL_GPL(security_mmap_backing_file);
> */
> int security_mmap_addr(unsigned long addr)
> {
> - return call_int_hook(mmap_addr, addr);
> + int rc = call_int_hook(mmap_addr, addr);
> +
> + /* commoncap granted CAP_SYS_RAWIO and no LSM denied the mapping. */
> + if (!rc && addr < dac_mmap_min_addr)
> + current->flags |= PF_SUPERPRIV;
> + return rc;
> }
>
> /**
> --
> 2.56.0
prev parent reply other threads:[~2026-11-20 18:49 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-18 16:56 [PATCH v1] lsm: Set PF_SUPERPRIV only when all mmap_addr hooks allow a mapping Mickaël Salaün
2026-11-18 17:03 ` sashiko-bot
2026-11-20 18:49 ` Serge E. Hallyn [this message]
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=awCWRzL9Y2uEm9pp@hallyn.com \
--to=serge@hallyn.com \
--cc=gnoack@google.com \
--cc=john.johansen@canonical.com \
--cc=kernel-team@cloudflare.com \
--cc=kpsingh@kernel.org \
--cc=linux-security-module@vger.kernel.org \
--cc=matt@bobrowski.net \
--cc=mic@digikod.net \
--cc=paul@paul-moore.com \
--cc=sergeh@kernel.org \
--cc=stephen.smalley.work@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.