Skip to contentSkip to conversation

security: Constify sysctl tables

PATCH In review Ricardo B. Marliere 4 participants 11 messages Jan 23, 2025 – Feb 14, 2025

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

cover security: Constify sysctl tables

Hello,

Please consider pulling these small cleanup patches. They simply add "const" to a couple of ctl_table structs within your subsystem.

Thank you,
- Ricardo.

---
Ricardo B. Marliere (2):
yama: Make sysctl table const
LoadPin: Make sysctl table const

 security/loadpin/loadpin.c | 2 +-
 security/yama/yama_lsm.c   | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

---
base-commit: a9a5e0bdc5a77a7c662ad4be0ad661f0b0d5e99d
change-id: 20250123-sysctl-kees-d7c805ee39cf

Best regards,

Signature · 1 line
Ricardo B. Marliere <[email protected]>
#1 lore raw

PATCH 1/2 yama: Make sysctl table const

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move yama_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

---
 security/yama/yama_lsm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/yama/yama_lsm.c +1 -1
index e1a5e13ea269d4917fe2c78cb79f1d44881653c2..54bd5f535ac1fef9409ea96ad60825f565318daf 100644
@@ -454,7 +454,7 @@ static int yama_dointvec_minmax(const struct ctl_table *table, int write,
 
 static int max_scope = YAMA_SCOPE_NO_ATTACH;
 
-static struct ctl_table yama_sysctl_table[] = {
+static const struct ctl_table yama_sysctl_table[] = {
 	{
 		.procname       = "ptrace_scope",
 		.data           = &ptrace_scope,
Signature · 1 line
2.48.1
maintainer Reviewed-by on v1 1/2
#2 lore raw

Re: [PATCH 1/2] yama: Make sysctl table const

Quoted 8 lines · Ricardo B. Marliere

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move yama_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

Signature · 1 line
Kees Cook
maintainer
#3 lore raw

Re: [PATCH 1/2] yama: Make sysctl table const

In reply to Ricardo B. Marliere
Quoted 17 lines · Ricardo B. Marliere

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move yama_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

---
 security/yama/yama_lsm.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/yama/yama_lsm.c +0 -0
index e1a5e13ea269d4917fe2c78cb79f1d44881653c2..54bd5f535ac1fef9409ea96ad60825f565318daf 100644

security/yama/yama_lsm.c static int yama_dointvec_minmax(const struct ctl_table *table, int write, · line 457

Quoted 5 lines
security/yama/yama_lsm.c +1 -1
@@ -454,7 +454,7 @@ static int yama_dointvec_minmax(const struct ctl_table *table, int write,
 
 static int max_scope = YAMA_SCOPE_NO_ATTACH;
 
-static struct ctl_table yama_sysctl_table[] = {
+static const struct ctl_table yama_sysctl_table[] = {

This was done via commit 1751f872cc97 ("treewide: const qualify ctl_tables where applicable").

Signature · 1 line
Kees Cook
#4 lore raw

PATCH 2/2 LoadPin: Make sysctl table const

In reply to Ricardo B. Marliere

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move loadpin_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

---
 security/loadpin/loadpin.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/loadpin/loadpin.c +1 -1
index 68252452b66cb913638abbca2adea26219e77d37..e2d664b7602629c08c86d6d02158a4e9dd189b1a 100644
@@ -53,7 +53,7 @@ static bool deny_reading_verity_digests;
 #endif
 
 #ifdef CONFIG_SYSCTL
-static struct ctl_table loadpin_sysctl_table[] = {
+static const struct ctl_table loadpin_sysctl_table[] = {
 	{
 		.procname       = "enforce",
 		.data           = &enforce,
Signature · 1 line
2.48.1
maintainer Reviewed-by on v1 2/2
#5 lore raw

Re: [PATCH 2/2] LoadPin: Make sysctl table const

Quoted 8 lines · Ricardo B. Marliere

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move loadpin_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

Signature · 1 line
Kees Cook
#6 lore raw

Re: [PATCH 2/2] LoadPin: Make sysctl table const

In reply to Ricardo B. Marliere
kernel test robot noticed the following build errors:

Hi Ricardo,

kernel test robot noticed the following build errors:

[auto build test ERROR on a9a5e0bdc5a77a7c662ad4be0ad661f0b0d5e99d]

url:    https://github.com/intel-lab-lkp/linux/commits/Ricardo-B-Marliere/yama-Make-sysctl-table-const/20250124-033808
base:   a9a5e0bdc5a77a7c662ad4be0ad661f0b0d5e99d
patch link:    https://lore.kernel.org/r/20250123-sysctl-kees-v1-2-533359e74d66%40suse.com
patch subject: [PATCH 2/2] LoadPin: Make sysctl table const
config: arm64-allmodconfig (https://download.01.org/0day-ci/archive/20250202/[email protected]/config)
compiler: clang version 18.1.8 (https://github.com/llvm/llvm-project 3b5b5c1ec4a3095ab096dd780e84d7ab81f3d7ff)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250202/[email protected]/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/

All errors (new ones prefixed by >>):

>> security/loadpin/loadpin.c:75:34: error: cannot assign to variable 'loadpin_sysctl_table' with const-qualified type 'const struct ctl_table[1]'
      75 |                 loadpin_sysctl_table[0].extra1 = SYSCTL_ZERO;
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^
   security/loadpin/loadpin.c:56:31: note: variable 'loadpin_sysctl_table' declared const here
      56 | static const struct ctl_table loadpin_sysctl_table[] = {
         | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
      57 |         {
         |         ~
      58 |                 .procname       = "enforce",
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      59 |                 .data           = &enforce,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
      60 |                 .maxlen         = sizeof(int),
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      61 |                 .mode           = 0644,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~
      62 |                 .proc_handler   = proc_dointvec_minmax,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      63 |                 .extra1         = SYSCTL_ONE,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      64 |                 .extra2         = SYSCTL_ONE,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      65 |         },
         |         ~~
      66 | };
         | ~
   security/loadpin/loadpin.c:77:34: error: cannot assign to variable 'loadpin_sysctl_table' with const-qualified type 'const struct ctl_table[1]'
      77 |                 loadpin_sysctl_table[0].extra1 = SYSCTL_ONE;
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^
   security/loadpin/loadpin.c:56:31: note: variable 'loadpin_sysctl_table' declared const here
      56 | static const struct ctl_table loadpin_sysctl_table[] = {
         | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
      57 |         {
         |         ~
      58 |                 .procname       = "enforce",
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      59 |                 .data           = &enforce,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~
      60 |                 .maxlen         = sizeof(int),
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      61 |                 .mode           = 0644,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~
      62 |                 .proc_handler   = proc_dointvec_minmax,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      63 |                 .extra1         = SYSCTL_ONE,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      64 |                 .extra2         = SYSCTL_ONE,
         |                 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
      65 |         },
         |         ~~
      66 | };
         | ~

2 errors generated.

vim +75 security/loadpin/loadpin.c

9b091556a073a9 Kees Cook 2016-04-20  67  
60ba1028fc7b73 Kees Cook 2022-12-09  68  static void set_sysctl(bool is_writable)
9b091556a073a9 Kees Cook 2016-04-20  69  {
9b091556a073a9 Kees Cook 2016-04-20  70  	/*
9b091556a073a9 Kees Cook 2016-04-20  71  	 * If load pinning is not enforced via a read-only block
9b091556a073a9 Kees Cook 2016-04-20  72  	 * device, allow sysctl to change modes for testing.
9b091556a073a9 Kees Cook 2016-04-20  73  	 */
60ba1028fc7b73 Kees Cook 2022-12-09  74  	if (is_writable)
60ba1028fc7b73 Kees Cook 2022-12-09 @75  		loadpin_sysctl_table[0].extra1 = SYSCTL_ZERO;
60ba1028fc7b73 Kees Cook 2022-12-09  76  	else
60ba1028fc7b73 Kees Cook 2022-12-09  77  		loadpin_sysctl_table[0].extra1 = SYSCTL_ONE;
60ba1028fc7b73 Kees Cook 2022-12-09  78  }
60ba1028fc7b73 Kees Cook 2022-12-09  79  #else
60ba1028fc7b73 Kees Cook 2022-12-09  80  static inline void set_sysctl(bool is_writable) { }
60ba1028fc7b73 Kees Cook 2022-12-09  81  #endif

60ba1028fc7b73 Kees Cook 2022-12-09 82

Signature · 2 lines
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
maintainer
#7 lore raw

Re: [PATCH 2/2] LoadPin: Make sysctl table const

In reply to Ricardo B. Marliere
Quoted 17 lines · Ricardo B. Marliere

Since commit 7abc9b53bd51 ("sysctl: allow registration of const struct ctl_table"), the sysctl registration API allows for struct ctl_table to be in read-only memory. Move loadpin_sysctl_table to be declared at build time, instead of having to be dynamically allocated at boot time.

---
 security/loadpin/loadpin.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
security/loadpin/loadpin.c +0 -0
index 68252452b66cb913638abbca2adea26219e77d37..e2d664b7602629c08c86d6d02158a4e9dd189b1a 100644

security/loadpin/loadpin.c static bool deny_reading_verity_digests; · line 59

Quoted 8 lines
security/loadpin/loadpin.c +1 -1
@@ -53,7 +53,7 @@ static bool deny_reading_verity_digests;
 #endif
 
 #ifdef CONFIG_SYSCTL
-static struct ctl_table loadpin_sysctl_table[] = {
+static const struct ctl_table loadpin_sysctl_table[] = {
 	{
 		.procname       = "enforce",
 		.data           = &enforce,

This can't be done since the table is modified during set_sysctl().

-Kees

Signature · 1 line
Kees Cook
#8 lore raw

Re: [PATCH 2/2] LoadPin: Make sysctl table const

In reply to Ricardo B. Marliere
kernel test robot noticed the following build errors:

Hi Ricardo,

kernel test robot noticed the following build errors:

[auto build test ERROR on a9a5e0bdc5a77a7c662ad4be0ad661f0b0d5e99d]

url:    https://github.com/intel-lab-lkp/linux/commits/Ricardo-B-Marliere/yama-Make-sysctl-table-const/20250124-033808
base:   a9a5e0bdc5a77a7c662ad4be0ad661f0b0d5e99d
patch link:    https://lore.kernel.org/r/20250123-sysctl-kees-v1-2-533359e74d66%40suse.com
patch subject: [PATCH 2/2] LoadPin: Make sysctl table const
config: nios2-allmodconfig (https://download.01.org/0day-ci/archive/20250214/[email protected]/config)
compiler: nios2-linux-gcc (GCC) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250214/[email protected]/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/

All errors (new ones prefixed by >>):

security/loadpin/loadpin.c: In function 'set_sysctl':

>> security/loadpin/loadpin.c:75:48: error: assignment of member 'extra1' in read-only object
      75 |                 loadpin_sysctl_table[0].extra1 = SYSCTL_ZERO;
         |                                                ^
   security/loadpin/loadpin.c:77:48: error: assignment of member 'extra1' in read-only object
      77 |                 loadpin_sysctl_table[0].extra1 = SYSCTL_ONE;
         |                                                ^

vim +/extra1 +75 security/loadpin/loadpin.c

9b091556a073a9 Kees Cook 2016-04-20  67  
60ba1028fc7b73 Kees Cook 2022-12-09  68  static void set_sysctl(bool is_writable)
9b091556a073a9 Kees Cook 2016-04-20  69  {
9b091556a073a9 Kees Cook 2016-04-20  70  	/*
9b091556a073a9 Kees Cook 2016-04-20  71  	 * If load pinning is not enforced via a read-only block
9b091556a073a9 Kees Cook 2016-04-20  72  	 * device, allow sysctl to change modes for testing.
9b091556a073a9 Kees Cook 2016-04-20  73  	 */
60ba1028fc7b73 Kees Cook 2022-12-09  74  	if (is_writable)
60ba1028fc7b73 Kees Cook 2022-12-09 @75  		loadpin_sysctl_table[0].extra1 = SYSCTL_ZERO;
60ba1028fc7b73 Kees Cook 2022-12-09  76  	else
60ba1028fc7b73 Kees Cook 2022-12-09  77  		loadpin_sysctl_table[0].extra1 = SYSCTL_ONE;
60ba1028fc7b73 Kees Cook 2022-12-09  78  }
60ba1028fc7b73 Kees Cook 2022-12-09  79  #else
60ba1028fc7b73 Kees Cook 2022-12-09  80  static inline void set_sysctl(bool is_writable) { }
60ba1028fc7b73 Kees Cook 2022-12-09  81  #endif

60ba1028fc7b73 Kees Cook 2022-12-09 82

Signature · 2 lines
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
maintainer
#9 lore raw
In reply to Ricardo B. Marliere
Quoted 2 lines · Ricardo B. Marliere

Please consider pulling these small cleanup patches. They simply add "const" to a couple of ctl_table structs within your subsystem.

Paul, do you want to take these, or should I put them in my tree?

-Kees

Signature · 1 line
Kees Cook
maintainer Reviewed-by
#10 lore raw
Quoted 5 lines · Kees Cook
Quoted 2 lines · Ricardo B. Marliere

Please consider pulling these small cleanup patches. They simply add "const" to a couple of ctl_table structs within your subsystem.

Paul, do you want to take these, or should I put them in my tree?

Since you maintain both Yama and LoadPin and these patches are fully contained within those LSMs I kinda figured you would grab them for their respective trees. However, if you would prefer for me to take them via the LSM tree I can do that, just let me know.

You can also feel free to add my review tag to both patches if you want.

Signature · 1 line
paul-moore.com