[v3] x86/rtc: Remove __init for runtime functions

Message ID af226986-c6ba-5192-4544-b878fd3140c8@nokia.com
State New
Headers
Series [v3] x86/rtc: Remove __init for runtime functions |

Commit Message

Matija Glavinic Pecotic Nov. 24, 2022, 4:57 p.m. UTC
  set_rtc_noop, get_rtc_noop are used runtime, therefore need init removed.
Crash was observed on x86 platform where cmos rtc is unused and disabled
via device tree. Function was triggered from ntp: sync_hw_clock, although
CONFIG_RTC_SYSTOHC=n, however sync_cmos_clock doesn't honour that.

  Workqueue: events_power_efficient sync_hw_clock
  RIP: 0010:set_rtc_noop
  Call Trace:
   update_persistent_clock64
   sync_hw_clock

Fix by dropping __init and making set/get_rtc_noop available runtime.

Fixes: c311ed6183f4 ("x86/init: Allow DT configured systems to disable RTC at boot time")
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Matija Glavinic Pecotic <matija.glavinic-pecotic.ext@nokia.com>
---
v3: Update commit message based on Andy's comments (further trim to backtrace, proper tag)
v2: Update commit message based on Andy's comments

 arch/x86/kernel/x86_init.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
  

Comments

Andy Shevchenko Nov. 24, 2022, 5:06 p.m. UTC | #1
On Thu, Nov 24, 2022 at 05:57:30PM +0100, Matija Glavinic Pecotic wrote:
> set_rtc_noop, get_rtc_noop are used runtime, therefore need init removed.
> Crash was observed on x86 platform where cmos rtc is unused and disabled
> via device tree. Function was triggered from ntp: sync_hw_clock, although
> CONFIG_RTC_SYSTOHC=n, however sync_cmos_clock doesn't honour that.
> 
>   Workqueue: events_power_efficient sync_hw_clock
>   RIP: 0010:set_rtc_noop
>   Call Trace:
>    update_persistent_clock64
>    sync_hw_clock
> 
> Fix by dropping __init and making set/get_rtc_noop available runtime.
> 
> Fixes: c311ed6183f4 ("x86/init: Allow DT configured systems to disable RTC at boot time")
> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

Side note: seems you sent this not using `git send-email ...`
(I see my @intel.com rather than @linux.intel.com in the Cc list).

But at least the Cc list looks closer to what MAINTAINERS says.

And you forgot to make it clear that a) it's a resend, by altering
--subject-prefix, and b) mentioning this in the changelog.
  

Patch

diff --git a/arch/x86/kernel/x86_init.c b/arch/x86/kernel/x86_init.c
index 57353519bc11..0ab707bbefe2 100644
--- a/arch/x86/kernel/x86_init.c
+++ b/arch/x86/kernel/x86_init.c
@@ -32,8 +32,8 @@  static int __init iommu_init_noop(void) { return 0; }
 static void iommu_shutdown_noop(void) { }
 bool __init bool_x86_init_noop(void) { return false; }
 void x86_op_int_noop(int cpu) { }
-static __init int set_rtc_noop(const struct timespec64 *now) { return -EINVAL; }
-static __init void get_rtc_noop(struct timespec64 *now) { }
+static int set_rtc_noop(const struct timespec64 *now) { return -EINVAL; }
+static void get_rtc_noop(struct timespec64 *now) { }
 
 static __initconst const struct of_device_id of_cmos_match[] = {
 	{ .compatible = "motorola,mc146818" },