kptr_restrict for hiding kernel pointers
Subsystems (6)
PATCH v4 1/1 kptr_restrict for hiding kernel pointers
Add the %pK printk format specifier and
the /proc/sys/kernel/kptr_restrict sysctl.
The %pK format specifier is designed to hide exposed kernel pointers, specifically via /proc interfaces. Exposing these pointers provides an easy target for kernel write vulnerabilities, since they reveal the locations of writable structures containing easily triggerable function pointers. The behavior of %pK depends on the kptr_restrict sysctl.
If kptr_restrict is set to 0, no deviation from the standard %p behavior occurs. If kptr_restrict is set to 1, if the current user (intended to be a reader via seq_printf(), etc.) does not have CAP_SYSLOG (which is currently in the LSM tree), kernel pointers using %pK are printed as 0's. If kptr_restrict is set to 2, kernel pointers using %pK are printed as 0's regardless of privileges. Replacing with 0's was chosen over the default "(null)", which cannot be parsed by userland %p, which expects "(nil)".
v4 incorporates Eric Paris' suggestion of using has_capability_noaudit(), since failing this capability check is not a policy violation but rather a code path choice and shouldn't generate potentially excessive log noise. Adjusted IRQ comment for clarity.
v3 adds the "2" setting, cleans up documentation, removes the CONFIG, and incorporates changes and suggestions from Andrew Morton.
v2 improves checking for inappropriate context, on suggestion by Peter Zijlstra. Thanks to Thomas Graf for suggesting use of a centralized format specifier.
---
Documentation/sysctl/kernel.txt | 14 ++++++++++++++ include/linux/kernel.h | 2 ++ kernel/sysctl.c | 9 +++++++++ lib/vsprintf.c | 24 ++++++++++++++++++++++++ 4 files changed, 49 insertions(+), 0 deletions(-)
Documentation/sysctl/kernel.txt +14 -0
@@ -34,6 +34,7 @@ show up in /proc/sys/kernel: - hotplug - java-appletviewer [ binfmt_java, obsolete ] - java-interpreter [ binfmt_java, obsolete ] +- kptr_restrict - kstack_depth_to_print [ X86 only ] - l2cr [ PPC only ] - modprobe ==> Documentation/debugging-modules.txt
@@ -261,6 +262,19 @@ This flag controls the L2 cache of G3 processor boards. If ============================================================== +kptr_restrict: + +This toggle indicates whether restrictions are placed on +exposing kernel addresses via /proc and other interfaces. When +kptr_restrict is set to (0), the default, there are no +restrictions. When kptr_restrict is set to (1), kernel pointers +printed using the %pK format specifier will be replaced with 0's +unless the user has CAP_SYSLOG. When kptr_restrict is set to +(2), kernel pointers printed using %pK will be replaced with 0's +regardless of privileges. + +============================================================== + kstack_depth_to_print: (X86 only) Controls the number of words to print when dumping the raw
include/linux/kernel.h +2 -0
@@ -201,6 +201,8 @@ extern int sscanf(const char *, const char *, ...)extern int vsscanf (const char *,const char *,va_list ) __attribute__ ((format (scanf,2 ,0 ))); +extern int kptr_restrict;/* for sysctl */ +extern int get_option (char **str,int *pint);extern char *get_options (const char *str,int nints,int *ints);extern unsigned long long memparse (const char *ptr,char **retptr);
kernel/sysctl.c +9 -0
@@ -713,6 +713,15 @@ static struct ctl_table kern_table[] = { },#endif { + .procname = "kptr_restrict", + .data = &kptr_restrict, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &zero, + .extra2 = &two, + }, + { .procname = "ngroups_max", .data = &ngroups_max, .maxlen = sizeof (int),
lib/vsprintf.c +24 -0
@@ -936,6 +936,8 @@ char *uuid_string(char *buf, char *end, const u8 *addr,return string (buf, end, uuid, spec); } +int kptr_restrict; +/* * Show a '%p' thing. A kernel extension is that the '%p' is followed * by an extra set of alphanumeric characters that are extended format
@@ -979,6 +981,7 @@ char *uuid_string(char *buf, char *end, const u8 *addr,* Implements a "recursive vsnprintf". * Do not use this feature without some mechanism to verify the * correctness of the format string and va_list arguments. +* - 'K' For a kernel pointer that should be hidden from unprivileged users * * Note: The difference between 'S' and 'F' is that on ia64 and ppc64 * function pointers are really function descriptors, which contain a
@@ -1035,6 +1038,27 @@ char *pointer(const char *fmt, char *buf, char *end, void *ptr,return buf +vsnprintf (buf, end - buf, ((struct va_format *)ptr)->fmt, *(((struct va_format *)ptr)->va)); +case 'K' : +/* +* %pK cannot be used in IRQ context because its test +* for CAP_SYSLOG would be meaningless. +*/ +if (in_irq () ||in_serving_softirq () ||in_nmi ()) +WARN_ONCE (1 ,"%%pK used in interrupt context.\n" ); + +if (!kptr_restrict) +break ;/* %pK does not obscure pointers */ + +if ((kptr_restrict !=2 ) && +has_capability_noaudit (current,CAP_SYSLOG )) +break ;/* privileged apps expose pointers, +unless kptr_restrict is 2 */ + +if (spec.field_width ==-1 ) { + spec.field_width =2 *sizeof (void *); + spec.flags |=ZEROPAD ; + } +return number (buf, end,0 , spec); } spec.flags |=SMALL ;if (spec.field_width ==-1 ) {
Documentation/sysctl/kernel.txt This flag controls the L2 cache of G3 processor boards. If · line 274
Quoted 10 lines · Dan Rosenberg
Documentation/sysctl/kernel.txt +10
-0+kptr_restrict: + +This toggle indicates whether restrictions are placed on +exposing kernel addresses via /proc and other interfaces. When +kptr_restrict is set to (0), the default, there are no +restrictions. When kptr_restrict is set to (1), kernel pointers +printed using the %pK format specifier will be replaced with 0's +unless the user has CAP_SYSLOG. When kptr_restrict is set to +(2), kernel pointers printed using %pK will be replaced with 0's +regardless of privileges.
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
Thanks,
Ingo
Quoted 7 lines
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
I agree entirely, but I've received a lot of resistance to these types of changes in net. I'm afraid that if it's enabled by default, no one will actually allow use of the %pK specifier where it should be used.
As far as I know, there's no actual breakage of anything in userspace, but there's a general "it might make it harder to debug things in certain limited circumstances" sentiment among some. I never understood why it is necessary for unprivileged users to be able to debug the kernel.
Does anyone else have thoughts on this?
Quoted 3 lines
Thanks,
Ingo
Quoted 11 lines · Dan Rosenberg
Quoted 6 lines
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
I agree entirely, but I've received a lot of resistance to these types of changes in net. I'm afraid that if it's enabled by default, no one will actually allow use of the %pK specifier where it should be used.
Actually, "net resistance" was against your first patches, using quick and dirty techniques (Should I remind you some of them ?)
Now you have a helper, it should be easier to integrate the changes.
At least, if a mission critical legacy app want to see real pointers values and a 2.6.38 kernel, it is a matter of sysadmin tweaks.
--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Quoted 10 lines · Dan Rosenberg
Quoted 6 lines
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
I agree entirely, but I've received a lot of resistance to these types of changes in net. I'm afraid that if it's enabled by default, no one will actually allow use of the %pK specifier where it should be used.
Some specific objections would be needed - which might arrive if the default is changed to on.
Quoted 7 lines
As far as I know, there's no actual breakage of anything in userspace, but there's a general "it might make it harder to debug things in certain limited circumstances" sentiment among some. I never understood why it is necessary for unprivileged users to be able to debug the kernel.
Does anyone else have thoughts on this?
Well, lets just enable it by default and let others argue for less security, hm?
Thanks,
Ingo
Quoted 13 lines · Eric Dumazet
Quoted 11 lines · Dan Rosenberg
Quoted 6 lines
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
I agree entirely, but I've received a lot of resistance to these types of changes in net. I'm afraid that if it's enabled by default, no one will actually allow use of the %pK specifier where it should be used.
Quoted 4 lines
Actually, "net resistance" was against your first patches, using quick and dirty techniques (Should I remind you some of them ?)
Now you have a helper, it should be easier to integrate the changes.
Great - if the concept itself wasnt objected to then i think we should flip the default to on.
Thanks,
Ingo
Andrew Morton
maintainerQuoted 16 lines · On Wed, 22 Dec 2010 17:21:18 +0100 Ingo Molnar <[email protected]> wrote:
Quoted 13 lines · Eric Dumazet
Le mercredi 22
d__cembre2010 __ 08:13 -0500, Dan Rosenberg a__crit:Quoted 11 lines
Quoted 6 lines
Hm, why is it off by default? Is there some user-space regression that is caused by this?
We really want good security measures to be active by default (and to work by default) - they are not worth much if they are not.
I agree entirely, but I've received a lot of resistance to these types of changes in net. I'm afraid that if it's enabled by default, no one will actually allow use of the %pK specifier where it should be used.
Quoted 8 lines
Quoted 4 lines
Actually, "net resistance" was against your first patches, using quick and dirty techniques (Should I remind you some of them ?)
Now you have a helper, it should be easier to integrate the changes.
Great - if the concept itself wasnt objected to then i think we should flip the default to on.
Yes, I'll make that change. If we get reports of breakage during 2.6.38-rcX then we can reconsider. But if we leave the feature disabled, we will never know...