| Message ID | 1469469547-8336-2-git-send-email-john.johansen@canonical.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Mon, 25 Jul 2016, John Johansen wrote: > From: Arnd Bergmann <arnd@arndb.de> > > The newly added Kconfig option could never work and just causes a build error > when disabled: > > security/apparmor/lsm.c:675:25: error: 'CONFIG_SECURITY_APPARMOR_HASH_DEFAULT' undeclared here (not in a function) > bool aa_g_hash_policy = CONFIG_SECURITY_APPARMOR_HASH_DEFAULT; > > The problem is that the macro undefined in this case, and we need to use the IS_ENABLED() > helper to turn it into a boolean constant. > > Another minor problem with the original patch is that the option is even offered > in sysfs when SECURITY_APPARMOR_HASH is not enabled, so this also hides the option > in that case. > > Signed-off-by: Arnd Bergmann <arnd@arndb.de> > Fixes: 6059f71f1e94 ("apparmor: add parameter to control whether policy hashing is used") > Signed-off-by: John Johansen <john.johansen@canonical.com> Please rebase this against current Linus.
On 07/26/2016 04:38 AM, James Morris wrote: > On Mon, 25 Jul 2016, John Johansen wrote: > >> From: Arnd Bergmann <arnd@arndb.de> >> >> The newly added Kconfig option could never work and just causes a build error >> when disabled: >> >> security/apparmor/lsm.c:675:25: error: 'CONFIG_SECURITY_APPARMOR_HASH_DEFAULT' undeclared here (not in a function) >> bool aa_g_hash_policy = CONFIG_SECURITY_APPARMOR_HASH_DEFAULT; >> >> The problem is that the macro undefined in this case, and we need to use the IS_ENABLED() >> helper to turn it into a boolean constant. >> >> Another minor problem with the original patch is that the option is even offered >> in sysfs when SECURITY_APPARMOR_HASH is not enabled, so this also hides the option >> in that case. >> >> Signed-off-by: Arnd Bergmann <arnd@arndb.de> >> Fixes: 6059f71f1e94 ("apparmor: add parameter to control whether policy hashing is used") >> Signed-off-by: John Johansen <john.johansen@canonical.com> > > Please rebase this against current Linus. > Well this is a fix for a patch in the security next queue, so linus doesn't have it yet. I can wait for your 4.8 pull request and rebase then -- To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/security/apparmor/crypto.c b/security/apparmor/crypto.c index 532471d..b75dab0 100644 --- a/security/apparmor/crypto.c +++ b/security/apparmor/crypto.c @@ -39,6 +39,9 @@ int aa_calc_profile_hash(struct aa_profile *profile, u32 version, void *start, int error = -ENOMEM; u32 le32_version = cpu_to_le32(version); + if (!aa_g_hash_policy) + return 0; + if (!apparmor_tfm) return 0; diff --git a/security/apparmor/lsm.c b/security/apparmor/lsm.c index 3be30c7..41b8cb1 100644 --- a/security/apparmor/lsm.c +++ b/security/apparmor/lsm.c @@ -671,9 +671,11 @@ enum profile_mode aa_g_profile_mode = APPARMOR_ENFORCE; module_param_call(mode, param_set_mode, param_get_mode, &aa_g_profile_mode, S_IRUSR | S_IWUSR); +#ifdef CONFIG_SECURITY_APPARMOR_HASH /* whether policy verification hashing is enabled */ -bool aa_g_hash_policy = CONFIG_SECURITY_APPARMOR_HASH_DEFAULT; +bool aa_g_hash_policy = IS_ENABLED(CONFIG_SECURITY_APPARMOR_HASH_DEFAULT); module_param_named(hash_policy, aa_g_hash_policy, aabool, S_IRUSR | S_IWUSR); +#endif /* Debug mode */ bool aa_g_debug; diff --git a/security/apparmor/policy_unpack.c b/security/apparmor/policy_unpack.c index b9b1c66..1381206 100644 --- a/security/apparmor/policy_unpack.c +++ b/security/apparmor/policy_unpack.c @@ -778,8 +778,7 @@ int aa_unpack(void *udata, size_t size, struct list_head *lh, const char **ns) if (error) goto fail_profile; - if (aa_g_hash_policy) - error = aa_calc_profile_hash(profile, e.version, start, + error = aa_calc_profile_hash(profile, e.version, start, e.pos - start); if (error) goto fail_profile;