ima: change how MODULE_SIG_FORCE is checked on modules checking policy
Subsystems (5)
cover ima: change how MODULE_SIG_FORCE is checked on modules checking policy
This patchset ensure that IMA's modules checking policy:
measure func=MODULE_CHECK uid=0
rely on the correct value of CONFIG_MODULE_SIG_FORCE, since the way it is today the code completely ignores the module.sig_enforce cmdline param, which behaves in a OR logic with the CONFIG value (CONFIG_MODULE_SIG_FORCE || module.sig_enforce). That said, everytime a module would load, in the current checking code, when the kernel was not compiled with the CONFIG set the call to init_module syscall fails with -EACCES:
# strace -f -v modprobe <any-module> | grep init_module init_module(0x55b9bcc9bba0, 17763, "") = -1 EACCES (Permission denied)
With this patchset the result would rely on the module.sig_enforce cmdline as well. Once the CONFIG is not set, but the param is, the result would be 'success', as it should be:
# strace -f -v modprobe <any-module> | grep init_module init_module(0x7f9602d6e010, 386646, "") = 0
The patchset was tested in two different kernels: 4.13.6 (Fedora 27) and 4.14.0-rc4 (integrity-next tree)
Bruno E. O. Meneguele (2):
module: export module signature enforcement status
ima: check signature enforcement against cmdline param instead of
CONFIG
include/linux/module.h | 2 ++ kernel/module.c | 8 ++++++++ security/integrity/ima/ima_main.c | 6 +++--- 3 files changed, 13 insertions(+), 3 deletions(-)
Signature · 6 lines
2.13.6 -- To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo at vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
PATCH 1/2 module: export module signature enforcement status
A static variable sig_enforce is used as status var to indicate the real value of CONFIG_MODULE_SIG_FORCE, once this one is set the var will hold true, but if the CONFIG is not set the status var will hold whatever value is present in the module.sig_enforce kernel cmdline param: true when =1 and false when =0 or not present.
Considering this cmdline param take place over the CONFIG value when it's not set, other places in the kernel could missbehave since they would have only the CONFIG_MODULE_SIG_FORCE value to rely on. Exporting this status var allows the kernel to rely in the effective value of module signature enforcement, being it from CONFIG value or cmdline param.
--- include/linux/module.h | 2 ++ kernel/module.c | 8 ++++++++ 2 files changed, 10 insertions(+)
include/linux/module.h +2 -0
@@ -806,4 +806,6 @@ static inline bool module_sig_ok(struct module *module) }#endif /* CONFIG_MODULE_SIG */ +bool is_module_sig_enforced (void ); +#endif /* _LINUX_MODULE_H */
kernel/module.c +8 -0
@@ -278,6 +278,14 @@ static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE);module_param (sig_enforce, bool_enable_only,0644 );#endif /* !CONFIG_MODULE_SIG_FORCE */ +/* Export sig_enforce kernel cmdline parameter to allow other subsystems rely +* on that instead of directly to CONFIG_MODULE_SIG_FORCE config. */ +bool is_module_sig_enforced (void ) +{ +return sig_enforce; +} +EXPORT_SYMBOL (is_module_sig_enforced); +/* Block module loading/unloading? */ int modules_disabled =0 ;core_param (nomodule, modules_disabled, bint,0 );
Signature · 6 lines
2.13.6 -- To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo at vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] module: export module signature enforcement status
Quoted 8 lines · Bruno E. O. Meneguele
A static variable
sig_enforceis used as status var to indicate the real value ofCONFIG_MODULE_SIG_FORCE, once this one is set the var will hold true, but if the CONFIG is not set the status var will hold whatever value is present in the module.sig_enforce kernel cmdline param: true when =1 and false when =0 or not present.Considering this cmdline param take place over the CONFIG value when it's not set, other places in the kernel could missbehave since they
^misbehave
Quoted 4 lines
would have only the
CONFIG_MODULE_SIG_FORCEvalue to rely on. Exporting this status var allows the kernel to rely in the effective value of module signature enforcement, being it from CONFIG value or cmdline param.
Thanks! There's a minor checkpatch warning below.
Quoted 22 lines
Signed-off-by Bruno E. O. Meneguele <[email protected]>--- include/linux/module.h | 2 ++ kernel/module.c | 8 ++++++++ 2 files changed, 10 insertions(+)include/linux/module.h +2
-0@@ -806,4 +806,6 @@ static inline bool module_sig_ok(struct module *module) }#endif /* CONFIG_MODULE_SIG */ +bool is_module_sig_enforced (void ); +#endif /* _LINUX_MODULE_H */ kernel/module.c +0
-0
kernel/module.c static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE); · line 282
Quoted 5 lines
kernel/module.c +2
-0@@ -278,6 +278,14 @@ static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE);module_param (sig_enforce, bool_enable_only,0644 );#endif /* !CONFIG_MODULE_SIG_FORCE */ +/* Export sig_enforce kernel cmdline parameter to allow other subsystems rely +* on that instead of directly to CONFIG_MODULE_SIG_FORCE config. */
Checkpatch complains. ?Please use the normal format:
/*
?*
?*/
kernel/module.c static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE); · line 291
Quoted 9 lines
kernel/module.c +6
-0+bool is_module_sig_enforced (void ) +{ +return sig_enforce; +} +EXPORT_SYMBOL (is_module_sig_enforced); +/* Block module loading/unloading? */ int modules_disabled =0 ;core_param (nomodule, modules_disabled, bint,0 );
--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo at vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 1/2] module: export module signature enforcement status
Quoted 7 lines · Mimi Zohar
Quoted 6 lines · Bruno E. O. Meneguele
A static variable
sig_enforceis used as status var to indicate the real value ofCONFIG_MODULE_SIG_FORCE, once this one is set the var will hold true, but if the CONFIG is not set the status var will hold whatever value is present in the module.sig_enforce kernel cmdline param: true when =1 and false when =0 or not present.
Quoted 5 lines
Quoted 2 lines
Considering this cmdline param take place over the CONFIG value when it's not set, other places in the kernel could missbehave since they
^misbehave
right.
Quoted 32 lines
Quoted 4 lines
would have only the
CONFIG_MODULE_SIG_FORCEvalue to rely on. Exporting this status var allows the kernel to rely in the effective value of module signature enforcement, being it from CONFIG value or cmdline param.Thanks! There's a minor checkpatch warning below.
Quoted 25 lines
Signed-off-by Bruno E. O. Meneguele <[email protected]>--- include/linux/module.h | 2 ++ kernel/module.c | 8 ++++++++ 2 files changed, 10 insertions(+)include/linux/module.h +2
-0@@ -806,4 +806,6 @@ static inline bool module_sig_ok(struct module *module) }#endif /* CONFIG_MODULE_SIG */ +bool is_module_sig_enforced (void ); +#endif /* _LINUX_MODULE_H */ kernel/module.c +0
-0@@ -278,6 +278,14 @@ static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE);module_param (sig_enforce, bool_enable_only,0644 );#endif /* !CONFIG_MODULE_SIG_FORCE */
Quoted 8 lines
kernel/module.c static bool sig_enforce = IS_ENABLED(CONFIG_MODULE_SIG_FORCE); · line 282
Quoted 2 lines
kernel/module.c +2
-0+/* Export sig_enforce kernel cmdline parameter to allow other subsystems rely +* on that instead of directly to CONFIG_MODULE_SIG_FORCE config. */ Checkpatch complains. Please use the normal format:
/*
*
*/
Oh, sure, sorry for that! I'll resend a v2 of the patchset in a minute.
Thanks.
Quoted 10 lines
Quoted 9 lines
kernel/module.c +6
-0+bool is_module_sig_enforced (void ) +{ +return sig_enforce; +} +EXPORT_SYMBOL (is_module_sig_enforced); +/* Block module loading/unloading? */ int modules_disabled =0 ;core_param (nomodule, modules_disabled, bint,0 );
Signature · 6 lines
bmeneg
PGP Key: http://bmeneg.com/pubkey.txt
[ Part 2, Application/PGP-SIGNATURE (Name: "signature.asc") 499 bytes. ]
[ Unable to print this part. ]PATCH 2/2 ima: check signature enforcement against cmdline param instead of CONFIG
In reply to Bruno E. O. MenegueleWhen the user requests MODULE_CHECK policy and its kernel is compiled with CONFIG_MODULE_SIG_FORCE not set, all modules would not load, just those loaded in initram time. One option the user would have would be set a kernel cmdline param (module.sig_enforce) to true, but the IMA module check code doesn't rely on this value, it checks just CONFIG_MODULE_SIG_FORCE.
This patch solves this problem checking for the exported value of module.sig_enforce cmdline param intead of CONFIG_MODULE_SIG_FORCE, which holds the effective value (CONFIG || param).
--- security/integrity/ima/ima_main.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
security/integrity/ima/ima_main.c +3 -3
@@ -356,12 +356,12 @@ void ima_post_path_mknod(struct dentry *dentry)*/ int ima_read_file (struct file *file,enum kernel_read_file_id read_id) { +bool sig_enforce =is_module_sig_enforced (); +if (!file && read_id ==READING_MODULE ) { -#ifndef CONFIG_MODULE_SIG_FORCE -if ((ima_appraise &IMA_APPRAISE_MODULES ) && +if (!sig_enforce && (ima_appraise &IMA_APPRAISE_MODULES ) && (ima_appraise &IMA_APPRAISE_ENFORCE ))return -EACCES ;/* INTEGRITY_UNKNOWN */ -#endif return 0 ;/* We rely on module signature checking */ }return 0 ;
Signature · 6 lines
2.13.6 -- To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo at vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH 2/2] ima: check signature enforcement against cmdline param instead of CONFIG
Quoted 7 lines · Bruno E. O. Meneguele
When the user requests
MODULE_CHECKpolicy and its kernel is compiled withCONFIG_MODULE_SIG_FORCEnot set, all modules would not load, just those loaded in initram time. One option the user would have would be set a kernel cmdline param (module.sig_enforce) to true, but the IMA module check code doesn't rely on this value, it checks justCONFIG_MODULE_SIG_FORCE.
Quoted 5 lines
This patch solves this problem checking for the exported value of module.sig_enforce cmdline param intead of
CONFIG_MODULE_SIG_FORCE, which holds the effective value (CONFIG || param).Signed-off-by Bruno E. O. Meneguele <[email protected]>
Thanks!
Quoted 14 lines
--- security/integrity/ima/ima_main.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)security/integrity/ima/ima_main.c +2
-0@@ -356,12 +356,12 @@ void ima_post_path_mknod(struct dentry *dentry)*/ int ima_read_file (struct file *file,enum kernel_read_file_id read_id) { +bool sig_enforce =is_module_sig_enforced (); +
security/integrity/ima/ima_main.c void ima_post_path_mknod(struct dentry *dentry) · line 367
Quoted 10 lines
security/integrity/ima/ima_main.c +1
-3if (!file && read_id ==READING_MODULE ) { -#ifndef CONFIG_MODULE_SIG_FORCE -if ((ima_appraise &IMA_APPRAISE_MODULES ) && +if (!sig_enforce && (ima_appraise &IMA_APPRAISE_MODULES ) && (ima_appraise &IMA_APPRAISE_ENFORCE ))return -EACCES ;/* INTEGRITY_UNKNOWN */ -#endif return 0 ;/* We rely on module signature checking */ }return 0 ;
--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo at vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html