Lock down ftrace

RFC PATCH In review David Howells 2 participants 3 messages Nov 9, 2017

Subsystems (3)

Tracing Security (LSM) ftrace

lore

Share & tools

lore

David Howells

author

2017-11-09 16:42 UTC #0 lore raw

RFC PATCH 1/1 Lock down ftrace

Hi,

I (may) need to lock down ftrace under secure boot conditions as part of the patch series that can be found here:

https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/log/?h=efi-lock-down

Can you tell me that if the attached patch is sufficient to the cause?

Thanks,
David
---
commit 3fb4590c0ad0a751b2090c489df79193510e6aaa

Author: David Howells <[email protected]>
Date:   Wed Nov 8 15:41:02 2017 +0000

Lock down ftrace

Disallow the use of ftrace when the kernel is locked down. This patch turns off ftrace_enabled late in the kernel boot so that the selftest can still be potentially be run.

The sysctl that controls ftrace_enables is also disallowed when the kernel is locked down. If the lockdown is lifted, then the sysctl can be used to reenable ftrace - if ftrace was compiled with CONFIG_DYNAMIC_FTRACE, that is; if it wasn't then it won't be possible to reenable it.

This prevents crypto data theft by analysis of execution patterns, and, if in future ftrace also logs the register contents at the time, will prevent data theft by that mechanism also.

Reported-by Alexei Starovoitov <[email protected]>

Signed-off-by David Howells <[email protected]>

kernel/trace/ftrace.c +22 -0

index 6abfafd7f173..9c7135963d80 100644
@@ -6508,6 +6508,9 @@ ftrace_enable_sysctl(struct ctl_table *table, int write,
 {
 	int ret = -ENODEV;
 
+	if (kernel_is_locked_down("Use of ftrace"))
+		return -EPERM;
+
 	mutex_lock(&ftrace_lock);
 
 	if (unlikely(ftrace_disabled))

@@ -6896,3 +6899,22 @@ void ftrace_graph_exit_task(struct task_struct *t)
 	kfree(ret_stack);
 }
 #endif
+
+#ifdef CONFIG_LOCK_DOWN_KERNEL
+static int __init ftrace_lock_down(void)
+{
+	mutex_lock(&ftrace_lock);
+
+	if (!ftrace_disabled && ftrace_enabled &&
+	    kernel_is_locked_down("Use of ftrace")) {
+		ftrace_enabled = false;
+		last_ftrace_enabled = false;
+		ftrace_trace_function = ftrace_stub;
+		ftrace_shutdown_sysctl();
+	}
+
+	mutex_unlock(&ftrace_lock);
+	return 0;
+}
+late_initcall(ftrace_lock_down);
+#endif

--
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

Steven Rostedt

maintainer

2017-11-09 18:03 UTC #1 lore raw

Quoted 18 lines · On Thu, 09 Nov 2017 16:42:12 +0000 David Howells <[email protected]> wrote:

Hi,
I (may) need to lock down ftrace under secure boot conditions as part of the patch series that can be found here:
https://git.kernel.org/pub/scm/linux/kernel/git/dhowells/linux-fs.git/log/?h=efi-lock-down
Can you tell me that if the attached patch is sufficient to the cause?
Thanks,
David
---
commit 3fb4590c0ad0a751b2090c489df79193510e6aaa
Author: David Howells <[email protected]>
Date:   Wed Nov 8 15:41:02 2017 +0000
Lock down ftrace

Quoted 8 lines

Disallow the use of ftrace when the kernel is locked down. This patch turns off ftrace_enabled late in the kernel boot so that the selftest can still be potentially be run.
The sysctl that controls ftrace_enables is also disallowed when the kernel is locked down. If the lockdown is lifted, then the sysctl can be used to reenable ftrace - if ftrace was compiled with CONFIG_DYNAMIC_FTRACE, that is; if it wasn't then it won't be possible to reenable it.

Actually, I see it being enabled with DYNAMIC_FTRACE not set. Calling into sysctl and enabling ftrace_enable, will allow the ftrace_trace_function to be set to something other than ftrace_stub again, allowing for static function tracing to run too.

Quoted 35 lines

This prevents crypto data theft by analysis of execution patterns, and, if in future ftrace also logs the register contents at the time, will prevent data theft by that mechanism also.

Reported-by Alexei Starovoitov <[email protected]>

Signed-off-by David Howells <[email protected]>

kernel/trace/ftrace.c +12 -0

index 6abfafd7f173..9c7135963d80 100644
@@ -6508,6 +6508,9 @@ ftrace_enable_sysctl(struct ctl_table *table, int write,
 {
 	int ret = -ENODEV;
 
+	if (kernel_is_locked_down("Use of ftrace"))
+		return -EPERM;
+
 	mutex_lock(&ftrace_lock);
 
 	if (unlikely(ftrace_disabled))

@@ -6896,3 +6899,22 @@ void ftrace_graph_exit_task(struct task_struct *t)
 	kfree(ret_stack);
 }
 #endif
+
+#ifdef CONFIG_LOCK_DOWN_KERNEL
+static int __init ftrace_lock_down(void)
+{
+	mutex_lock(&ftrace_lock);
+
+	if (!ftrace_disabled && ftrace_enabled &&
+	    kernel_is_locked_down("Use of ftrace")) {
+		ftrace_enabled = false;

kernel/trace/ftrace.c void ftrace_graph_exit_task(struct task_struct *t) · line 6920

Quoted 10 lines

kernel/trace/ftrace.c +10 -0

+		last_ftrace_enabled = false;
+		ftrace_trace_function = ftrace_stub;
+		ftrace_shutdown_sysctl();
+	}
+
+	mutex_unlock(&ftrace_lock);
+	return 0;
+}
+late_initcall(ftrace_lock_down);
+#endif

Looks fine to me. We discussed this offline, and this appears to implement what we finally decided would be sufficient.

-- Steve

David Howells

author

2017-11-09 21:54 UTC #2 lore raw

Quoted 4 lines · Steven Rostedt

Actually, I see it being enabled with DYNAMIC_FTRACE not set. Calling into sysctl and enabling ftrace_enable, will allow the ftrace_trace_function to be set to something other than ftrace_stub again, allowing for static function tracing to run too.

Hmmm... Okay, I'm not sure what the sysctl achieves in non-dynamic mode. Some of the functions used by ftrace_enable_sysctl() are stubbed out in that case. I was thinking that was stubbed out also, but apparently not.

Anyway, ftrace_enable_sysctl() is also prohibited in lockdown mode.

David
--
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