2024-03-22 13:20:59

by Arnd Bergmann

[permalink] [raw]
Subject: [PATCH] irqflags: explicitly ignore lockdep_hrtimer_exit() argument

From: Arnd Bergmann <[email protected]>

When building with 'make W=1' but CONFIG_TRACE_IRQFLAGS=n, the
unused argument to lockdep_hrtimer_exit() causes a warning:

kernel/time/hrtimer.c:1655:14: error: variable 'expires_in_hardirq' set but not used [-Werror=unused-but-set-variable]

This is intentional behavior, so add a cast to void to shut up the warning.

Fixes: 73d20564e0dc ("hrtimer: Don't dereference the hrtimer pointer after the callback")
Signed-off-by: Arnd Bergmann <[email protected]>
---
include/linux/irqflags.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/linux/irqflags.h b/include/linux/irqflags.h
index 147feebd508c..3f003d5fde53 100644
--- a/include/linux/irqflags.h
+++ b/include/linux/irqflags.h
@@ -114,7 +114,7 @@ do { \
# define lockdep_softirq_enter() do { } while (0)
# define lockdep_softirq_exit() do { } while (0)
# define lockdep_hrtimer_enter(__hrtimer) false
-# define lockdep_hrtimer_exit(__context) do { } while (0)
+# define lockdep_hrtimer_exit(__context) do { (void)(__context); } while (0)
# define lockdep_posixtimer_enter() do { } while (0)
# define lockdep_posixtimer_exit() do { } while (0)
# define lockdep_irq_work_enter(__work) do { } while (0)
--
2.39.2



Subject: Re: [PATCH] irqflags: explicitly ignore lockdep_hrtimer_exit() argument

On 2024-03-22 14:20:41 [+0100], Arnd Bergmann wrote:
> From: Arnd Bergmann <[email protected]>
>
> When building with 'make W=1' but CONFIG_TRACE_IRQFLAGS=n, the
> unused argument to lockdep_hrtimer_exit() causes a warning:
>
> kernel/time/hrtimer.c:1655:14: error: variable 'expires_in_hardirq' set but not used [-Werror=unused-but-set-variable]
>
> This is intentional behavior, so add a cast to void to shut up the warning.
>
> Fixes: 73d20564e0dc ("hrtimer: Don't dereference the hrtimer pointer after the callback")
> Signed-off-by: Arnd Bergmann <[email protected]>

Hmm. Seems I forgot about it
https://lore.kernel.org/all/[email protected]/

Maybe we could the commit message from the referenced patch. Either way,
Reviewed-by: Sebastian Andrzej Siewior <[email protected]>

Sebastian