ntp: Fix ADJ_SETOFFSET being used w/ ADJ_NANO
authorJohn Stultz <john.stultz@linaro.org>
Thu, 21 Jan 2016 23:03:34 +0000 (15:03 -0800)
committerThomas Gleixner <tglx@linutronix.de>
Fri, 22 Jan 2016 11:01:42 +0000 (12:01 +0100)
Recently, in commit 37cf4dc3370f I forgot to check if the timeval being passed
was actually a timespec (as is signaled with ADJ_NANO).

This resulted in that patch breaking ADJ_SETOFFSET users who set
ADJ_NANO, by rejecting valid timespecs that were compared with
valid timeval ranges.

This patch addresses this by checking for the ADJ_NANO flag and
using the timepsec check instead in that case.

Reported-by: Harald Hoyer <harald@redhat.com>
Reported-by: Kay Sievers <kay@vrfy.org>
Fixes: 37cf4dc3370f "time: Verify time values in adjtimex ADJ_SETOFFSET to avoid overflow"
Signed-off-by: John Stultz <john.stultz@linaro.org>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: David Herrmann <dh.herrmann@gmail.com>
Link: http://lkml.kernel.org/r/1453417415-19110-2-git-send-email-john.stultz@linaro.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
kernel/time/ntp.c

index 36f2ca0..6df8927 100644 (file)
@@ -685,8 +685,18 @@ int ntp_validate_timex(struct timex *txc)
                if (!capable(CAP_SYS_TIME))
                        return -EPERM;
 
-               if (!timeval_inject_offset_valid(&txc->time))
-                       return -EINVAL;
+               if (txc->modes & ADJ_NANO) {
+                       struct timespec ts;
+
+                       ts.tv_sec = txc->time.tv_sec;
+                       ts.tv_nsec = txc->time.tv_usec;
+                       if (!timespec_inject_offset_valid(&ts))
+                               return -EINVAL;
+
+               } else {
+                       if (!timeval_inject_offset_valid(&txc->time))
+                               return -EINVAL;
+               }
        }
 
        /*