[PATCH] audit: fix two bugs in the new execve audit code
On Fri, 2007-07-27 at 16:13 -0400, Steve Grubb wrote:
Hi,
I was testing our rawhide kernel and I'm scrolling these errors:
WARNING: at kernel/auditsc.c:859 audit_log_execve_info() (Not tainted)
Call Trace:
[<ffffffff8106b06f>] audit_log_exit+0x5d7/0x964
[<ffffffff81050805>] trace_hardirqs_on+0x12e/0x151
[<ffffffff8106b60b>] audit_syscall_exit+0x9b/0x300
[<ffffffff8100ee62>] syscall_trace_leave+0x2c/0x87
[<ffffffff8100beb1>] int_very_careful+0x3a/0x43
--
copy_from_user() returns the number of bytes not copied, hence 0 is the
expected output.
axi->mm might not be valid anymore when not equal to current->mm, do not
dereference before checking that - thanks to Al for spotting that.
Signed-off-by: Peter Zijlstra <a.p.zijlstra(a)chello.nl>
---
kernel/auditsc.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
Index: linux-2.6/kernel/auditsc.c
===================================================================
--- linux-2.6.orig/kernel/auditsc.c
+++ linux-2.6/kernel/auditsc.c
@@ -824,12 +824,14 @@ static void audit_log_execve_info(struct
{
int i;
long len, ret;
- const char __user *p = (const char __user *)axi->mm->arg_start;
+ const char __user *p;
char *buf;
if (axi->mm != current->mm)
return; /* execve failed, no additional info */
+ p = (const char __user *)axi->mm->arg_start;
+
for (i = 0; i < axi->argc; i++, p += len) {
len = strnlen_user(p, MAX_ARG_STRLEN);
/*
@@ -855,7 +857,7 @@ static void audit_log_execve_info(struct
* copied them here, and the mm hasn't been exposed to user-
* space yet.
*/
- if (!ret) {
+ if (ret) {
WARN_ON(1);
send_sig(SIGKILL, current, 0);
}