From: KP Singh <kpsingh@kernel.org>
To: Andrii Nakryiko <andrii.nakryiko@gmail.com>
Cc: Casey Schaufler <casey@schaufler-ca.com>,
linux-security-module@vger.kernel.org, bpf@vger.kernel.org,
paul@paul-moore.com, keescook@chromium.org, song@kernel.org,
daniel@iogearbox.net, ast@kernel.org, jannh@google.com,
Kui-Feng Lee <sinquersw@gmail.com>
Subject: Re: [PATCH v2 2/5] security: Count the LSMs enabled at compile time
Date: Sat, 16 Sep 2023 02:54:29 +0200 [thread overview]
Message-ID: <CACYkzJ4B=hcOgz_C3jYQzRYchXQGCbSQym6a2aQEM7OXbdho7w@mail.gmail.com> (raw)
In-Reply-To: <CAEf4BzYuurXCTfqkfLc3RvWZiUzJ2am2GwcYgZgiEb91cGGZaw@mail.gmail.com>
On Sat, Jun 17, 2023 at 12:27 AM Andrii Nakryiko
<andrii.nakryiko@gmail.com> wrote:
>
> On Thu, Jun 15, 2023 at 5:38 PM Casey Schaufler <casey@schaufler-ca.com> wrote:
> >
> > On 6/15/2023 5:04 PM, KP Singh wrote:
> > > These macros are a clever trick to determine a count of the number of
> > > LSMs that are enabled in the config to ascertain the maximum number of
> > > static calls that need to be configured per LSM hook.
> > >
> > > Without this one would need to generate static calls for (number of
> > > possible LSMs * number of LSM hooks) which ends up being quite wasteful
> > > especially when some LSMs are not compiled into the kernel.
> > >
> > > Suggested-by: Kui-Feng Lee <sinquersw@gmail.com>
> > > Signed-off-by: KP Singh <kpsingh@kernel.org>
> > > ---
> > > include/linux/lsm_count.h | 131 ++++++++++++++++++++++++++++++++++++++
> > > 1 file changed, 131 insertions(+)
> > > create mode 100644 include/linux/lsm_count.h
> > >
> > > diff --git a/include/linux/lsm_count.h b/include/linux/lsm_count.h
> > > new file mode 100644
> > > index 000000000000..818f62ffa723
> > > --- /dev/null
> > > +++ b/include/linux/lsm_count.h
> > > @@ -0,0 +1,131 @@
> > > +/* SPDX-License-Identifier: GPL-2.0 */
> > > +
> > > +/*
> > > + * Copyright (C) 2023 Google LLC.
> > > + */
> > > +
> > > +#ifndef __LINUX_LSM_COUNT_H
> > > +#define __LINUX_LSM_COUNT_H
> > > +
> > > +#include <linux/kconfig.h>
> > > +
> > > +/*
> > > + * Macros to count the number of LSMs enabled in the kernel at compile time.
> > > + */
> > > +
> > > +#define __LSM_COUNT_15(x, y...) 15
> > > +#define __LSM_COUNT_14(x, y...) 14
> > > +#define __LSM_COUNT_13(x, y...) 13
> > > +#define __LSM_COUNT_12(x, y...) 12
> > > +#define __LSM_COUNT_11(x, y...) 11
> > > +#define __LSM_COUNT_10(x, y...) 10
> > > +#define __LSM_COUNT_9(x, y...) 9
> > > +#define __LSM_COUNT_8(x, y...) 8
> > > +#define __LSM_COUNT_7(x, y...) 7
> > > +#define __LSM_COUNT_6(x, y...) 6
> > > +#define __LSM_COUNT_5(x, y...) 5
> > > +#define __LSM_COUNT_4(x, y...) 4
> > > +#define __LSM_COUNT_3(x, y...) 3
> > > +#define __LSM_COUNT_2(x, y...) 2
> > > +#define __LSM_COUNT_1(x, y...) 1
> > > +#define __LSM_COUNT_0(x, y...) 0
> > > +
> > > +#define __LSM_COUNT1_15(x, y...) __LSM_COUNT ## x ## _15(y)
> > > +#define __LSM_COUNT1_14(x, y...) __LSM_COUNT ## x ## _14(y)
> > > +#define __LSM_COUNT1_13(x, y...) __LSM_COUNT ## x ## _13(y)
> > > +#define __LSM_COUNT1_12(x, y...) __LSM_COUNT ## x ## _12(y)
> > > +#define __LSM_COUNT1_10(x, y...) __LSM_COUNT ## x ## _11(y)
> > > +#define __LSM_COUNT1_9(x, y...) __LSM_COUNT ## x ## _10(y)
> > > +#define __LSM_COUNT1_8(x, y...) __LSM_COUNT ## x ## _9(y)
> > > +#define __LSM_COUNT1_7(x, y...) __LSM_COUNT ## x ## _8(y)
> > > +#define __LSM_COUNT1_6(x, y...) __LSM_COUNT ## x ## _7(y)
> > > +#define __LSM_COUNT1_5(x, y...) __LSM_COUNT ## x ## _6(y)
> > > +#define __LSM_COUNT1_4(x, y...) __LSM_COUNT ## x ## _5(y)
> > > +#define __LSM_COUNT1_3(x, y...) __LSM_COUNT ## x ## _4(y)
> > > +#define __LSM_COUNT1_2(x, y...) __LSM_COUNT ## x ## _3(y)
> > > +#define __LSM_COUNT1_1(x, y...) __LSM_COUNT ## x ## _2(y)
> > > +#define __LSM_COUNT1_0(x, y...) __LSM_COUNT ## x ## _1(y)
> > > +#define __LSM_COUNT(x, y...) __LSM_COUNT ## x ## _0(y)
> > > +
> > > +#define __LSM_COUNT_EXPAND(x...) __LSM_COUNT(x)
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY)
> > > +#define CAPABILITIES_ENABLED 1,
> > > +#else
> > > +#define CAPABILITIES_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_SELINUX)
> > > +#define SELINUX_ENABLED 1,
> > > +#else
> > > +#define SELINUX_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_SMACK)
> > > +#define SMACK_ENABLED 1,
> > > +#else
> > > +#define SMACK_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_APPARMOR)
> > > +#define APPARMOR_ENABLED 1,
> > > +#else
> > > +#define APPARMOR_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_TOMOYO)
> > > +#define TOMOYO_ENABLED 1,
> > > +#else
> > > +#define TOMOYO_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_YAMA)
> > > +#define YAMA_ENABLED 1,
> > > +#else
> > > +#define YAMA_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_LOADPIN)
> > > +#define LOADPIN_ENABLED 1,
> > > +#else
> > > +#define LOADPIN_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_LOCKDOWN_LSM)
> > > +#define LOCKDOWN_ENABLED 1,
> > > +#else
> > > +#define LOCKDOWN_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_BPF_LSM)
> > > +#define BPF_LSM_ENABLED 1,
> > > +#else
> > > +#define BPF_LSM_ENABLED
> > > +#endif
> > > +
> > > +#if IS_ENABLED(CONFIG_BPF_LSM)
> > > +#define BPF_LSM_ENABLED 1,
> > > +#else
> > > +#define BPF_LSM_ENABLED
> > > +#endif
>
> duplicate that redefined BPF_LSM_ENABLED unnecessarily
>
> > > +
> > > +#if IS_ENABLED(CONFIG_SECURITY_LANDLOCK)
> > > +#define LANDLOCK_ENABLED 1,
> > > +#else
> > > +#define LANDLOCK_ENABLED
> > > +#endif
> > > +
> > > +#define MAX_LSM_COUNT \
> > > + __LSM_COUNT_EXPAND( \
> > > + CAPABILITIES_ENABLED \
> > > + SELINUX_ENABLED \
> > > + SMACK_ENABLED \
> > > + APPARMOR_ENABLED \
> > > + TOMOYO_ENABLED \
> > > + YAMA_ENABLED \
> > > + LOADPIN_ENABLED \
> > > + LOCKDOWN_ENABLED \
> > > + BPF_LSM_ENABLED \
> > > + LANDLOCK_ENABLED)
> > > +
> >
> > Wouldn't the following be simpler? It's from my LSM syscall patchset.
>
> Of course it would be, but unfortunately it doesn't work with the
> UNROLL() macro. This MAX_LSM_COUNT has to evaluate a compile-time
> integer *literal* (not any sort of expression), so that UNROLL(N,...)
> can do its magic.
>
>
> KP, this __LSM_COUNT_EXPAND() is actually doing exactly what already
> existing COUNT_ARGS() macro from linux/kernel.h does, which is
> implemented way more succinctly:
>
> #define __COUNT_ARGS(_0, _1, _2, _3, _4, _5, _6, _7, _8, _9, _10, _11,
> _12, _n, X...) _n
> #define COUNT_ARGS(X...) __COUNT_ARGS(, ##X, 12, 11, 10, 9, 8, 7, 6,
> 5, 4, 3, 2, 1, 0)
>
>
> The only problem is that:
>
> #define ___COUNT_ARGS(args...) COUNT_ARGS(args)
> #define MAX_LSM_COUNT \
> ___COUNT_ARGS( \
> CAPABILITIES_ENABLED \
> SELINUX_ENABLED \
> SMACK_ENABLED \
> APPARMOR_ENABLED \
> TOMOYO_ENABLED \
> YAMA_ENABLED \
> LOADPIN_ENABLED \
> LOCKDOWN_ENABLED \
> BPF_LSM_ENABLED \
> LANDLOCK_ENABLED)
>
> overcounts by one, because of that trailing command within each
> XXX_ENABLED definition.
>
>
> But still, instead of a multi-line __LSM_COUNT{,1}_N set of macros, it
> might be better to use the COUNT_ARGS trick, but just account for
> those trailing commas? E.g., maybe just do a COUNT_COMMAS() macro
> which will adjust all the return values by 1 down, except when there
> is no comma (still 0).
>
> It's pretty minor in the grand scheme of things, but just something
> for you to be aware of.
I am back and revving this up again (after a hiatus due to health
stuff and then ramping back at work). Apologies for the radio silence
here.
I agree, Also if you notice CAPABILITIES_ENABLED is kinda bogus, and
CONFIG_SECURITY is used as a proxy, overcounting by 1 is actually what
I need. So, thanks, this makes it much simpler.
^^
(I realized I had replied the above to Andrii and not replied back to the list).
>
>
> > It certainly takes up fewer lines and would be easier to maintain
> > than the set of macros you've proposed.
> >
> > +#define LSM_COUNT ( \
> > + (IS_ENABLED(CONFIG_SECURITY) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_SELINUX) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_SMACK) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_TOMOYO) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_IMA) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_APPARMOR) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_YAMA) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_LOADPIN) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_SAFESETID) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_LOCKDOWN_LSM) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_BPF_LSM) ? 1 : 0) + \
> > + (IS_ENABLED(CONFIG_SECURITY_LANDLOCK) ? 1 : 0))
> >
> >
> > > +#endif /* __LINUX_LSM_COUNT_H */
> >
next prev parent reply other threads:[~2023-09-16 0:56 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-16 0:04 [PATCH v2 0/5] Reduce overhead of LSMs with static calls KP Singh
2023-06-16 0:04 ` [PATCH v2 1/5] kernel: Add helper macros for loop unrolling KP Singh
2023-06-16 0:04 ` [PATCH v2 2/5] security: Count the LSMs enabled at compile time KP Singh
2023-06-16 0:38 ` Casey Schaufler
2023-06-16 22:27 ` Andrii Nakryiko
2023-09-16 0:54 ` KP Singh [this message]
2023-06-16 0:04 ` [PATCH v2 3/5] security: Replace indirect LSM hook calls with static calls KP Singh
2023-06-16 1:05 ` Casey Schaufler
2023-06-17 15:09 ` KP Singh
2023-06-16 3:41 ` kernel test robot
2023-06-16 21:09 ` kernel test robot
2023-06-17 15:39 ` KP Singh
2023-06-20 21:53 ` Kees Cook
2023-06-16 0:04 ` [PATCH v2 4/5] bpf: Only enable BPF LSM hooks when an LSM program is attached KP Singh
2023-06-16 0:04 ` [PATCH v2 5/5] security: Add CONFIG_SECURITY_HOOK_LIKELY KP Singh
2023-06-16 1:14 ` Casey Schaufler
2023-06-17 15:11 ` KP Singh
2023-06-20 20:58 ` Kees Cook
2023-09-18 13:27 ` KP Singh
2023-09-18 13:55 ` Paul Moore
2023-09-18 16:28 ` KP Singh
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='CACYkzJ4B=hcOgz_C3jYQzRYchXQGCbSQym6a2aQEM7OXbdho7w@mail.gmail.com' \
--to=kpsingh@kernel.org \
--cc=andrii.nakryiko@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=casey@schaufler-ca.com \
--cc=daniel@iogearbox.net \
--cc=jannh@google.com \
--cc=keescook@chromium.org \
--cc=linux-security-module@vger.kernel.org \
--cc=paul@paul-moore.com \
--cc=sinquersw@gmail.com \
--cc=song@kernel.org \
/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;
as well as URLs for NNTP newsgroup(s).