timekeeping: Fix lost updates to tai adjustment
authorJohn Stultz <john.stultz@linaro.org>
Thu, 12 Dec 2013 02:50:25 +0000 (18:50 -0800)
committerJohn Stultz <john.stultz@linaro.org>
Mon, 23 Dec 2013 19:47:35 +0000 (11:47 -0800)
Since 48cdc135d4840 (Implement a shadow timekeeper), we have to
call timekeeping_update() after any adjustment to the timekeeping
structure in order to make sure that any adjustments to the structure
persist.

Unfortunately, the updates to the tai offset via adjtimex do not
trigger this update, causing adjustments to the tai offset to be
made and then over-written by the previous value at the next
update_wall_time() call.

This patch resovles the issue by calling timekeeping_update()
right after setting the tai offset.

Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Prarit Bhargava <prarit@redhat.com>
Cc: Richard Cochran <richardcochran@gmail.com>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: stable <stable@vger.kernel.org> #3.10+
Signed-off-by: John Stultz <john.stultz@linaro.org>
kernel/time/timekeeping.c

index 3abf53418b67c6ee92c16522299a8821856b749d..7488f0b97dee7a0ff695a3a1e3b347b98e5dc75e 100644 (file)
@@ -610,6 +610,7 @@ void timekeeping_set_tai_offset(s32 tai_offset)
        raw_spin_lock_irqsave(&timekeeper_lock, flags);
        write_seqcount_begin(&timekeeper_seq);
        __timekeeping_set_tai_offset(tk, tai_offset);
+       timekeeping_update(tk, TK_MIRROR | TK_CLOCK_WAS_SET);
        write_seqcount_end(&timekeeper_seq);
        raw_spin_unlock_irqrestore(&timekeeper_lock, flags);
        clock_was_set();
@@ -1698,7 +1699,7 @@ int do_adjtimex(struct timex *txc)
 
        if (tai != orig_tai) {
                __timekeeping_set_tai_offset(tk, tai);
-               update_pvclock_gtod(tk, true);
+               timekeeping_update(tk, TK_MIRROR | TK_CLOCK_WAS_SET);
                clock_was_set_delayed();
        }
        write_seqcount_end(&timekeeper_seq);