As the front patch 6 ~ 13 did, the arm64_preempt_schedule_irq() is
same with the irq preempt schedule code of generic entry besides those
architecture-related logic called arm64_irqentry_exit_need_resched().

So add arch irqentry_exit_need_resched() to support architecture-related
need_resched() check logic, which do not affect existing architectures
that use generic entry, but support arm64 to use generic irq entry.

Suggested-by: Mark Rutland <mark.rutland@arm.com>
Suggested-by: Kevin Brodsky <kevin.brodsky@arm.com>
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com>
---
 kernel/entry/common.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)

diff --git a/kernel/entry/common.c b/kernel/entry/common.c
index 2ad132c7be05..0cc117b658b8 100644
--- a/kernel/entry/common.c
+++ b/kernel/entry/common.c
@@ -143,6 +143,20 @@ noinstr irqentry_state_t irqentry_enter(struct pt_regs *regs)
 	return ret;
 }
 
+/**
+ * arch_irqentry_exit_need_resched - Architecture specific need resched function
+ *
+ * Invoked from raw_irqentry_exit_cond_resched() to check if need resched.
+ * Defaults return true.
+ *
+ * The main purpose is to permit arch to skip preempt a task from an IRQ.
+ */
+static inline bool arch_irqentry_exit_need_resched(void);
+
+#ifndef arch_irqentry_exit_need_resched
+static inline bool arch_irqentry_exit_need_resched(void) { return true; }
+#endif
+
 void raw_irqentry_exit_cond_resched(void)
 {
 	if (!preempt_count()) {
@@ -150,7 +164,7 @@ void raw_irqentry_exit_cond_resched(void)
 		rcu_irq_exit_check_preempt();
 		if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
 			WARN_ON_ONCE(!on_thread_stack());
-		if (need_resched())
+		if (need_resched() && arch_irqentry_exit_need_resched())
 			preempt_schedule_irq();
 	}
 }
-- 
2.34.1

On Fri, Oct 25 2024 at 18:06, Jinjie Ruan wrote:

> As the front patch 6 ~ 13 did, the arm64_preempt_schedule_irq() is

Once this series is applied nobody knows what 'front patch 6 ~ 13' did.

> same with the irq preempt schedule code of generic entry besides those
> architecture-related logic called arm64_irqentry_exit_need_resched().
>
> So add arch irqentry_exit_need_resched() to support architecture-related
> need_resched() check logic, which do not affect existing architectures
> that use generic entry, but support arm64 to use generic irq entry.

Simply say:

 ARM64 requires an additional whether to reschedule on return from
 interrupt.

 Add arch_irqentry_exit_need_resched() as the default NOOP
 implementation and hook it up into the need_resched() condition in
 raw_irqentry_exit_cond_resched().

 This allows ARM64 to implement the architecture specific version for
 switchting over to the generic entry code.

That explains things completely independently. Hmm?

Thanks,

        tglx

On 2024/10/29 2:05, Thomas Gleixner wrote:
> On Fri, Oct 25 2024 at 18:06, Jinjie Ruan wrote:
> 
>> As the front patch 6 ~ 13 did, the arm64_preempt_schedule_irq() is
> 
> Once this series is applied nobody knows what 'front patch 6 ~ 13' did.

Yes, if some of the previous patches are applied, the description will
immediately become difficult to understand, the other patch's similar
commit message will be updated too.

> 
>> same with the irq preempt schedule code of generic entry besides those
>> architecture-related logic called arm64_irqentry_exit_need_resched().
>>
>> So add arch irqentry_exit_need_resched() to support architecture-related
>> need_resched() check logic, which do not affect existing architectures
>> that use generic entry, but support arm64 to use generic irq entry.
> 
> Simply say:
> 
>  ARM64 requires an additional whether to reschedule on return from
>  interrupt.
> 
>  Add arch_irqentry_exit_need_resched() as the default NOOP
>  implementation and hook it up into the need_resched() condition in
>  raw_irqentry_exit_cond_resched().
> 
>  This allows ARM64 to implement the architecture specific version for
>  switchting over to the generic entry code.
> 
> That explains things completely independently. Hmm?

Of course, this is clearer and not as coupled as other patches and
describes how to implement it.
> 
> Thanks,
> 
>         tglx
>

On Mon, Oct 28 2024 at 19:05, Thomas Gleixner wrote:

> On Fri, Oct 25 2024 at 18:06, Jinjie Ruan wrote:
>
>> As the front patch 6 ~ 13 did, the arm64_preempt_schedule_irq() is
>
> Once this series is applied nobody knows what 'front patch 6 ~ 13' did.
>
>> same with the irq preempt schedule code of generic entry besides those
>> architecture-related logic called arm64_irqentry_exit_need_resched().
>>
>> So add arch irqentry_exit_need_resched() to support architecture-related
>> need_resched() check logic, which do not affect existing architectures
>> that use generic entry, but support arm64 to use generic irq entry.
>
> Simply say:
>
>  ARM64 requires an additional whether to reschedule on return from

  ARM64 requires an additional check whether to reschedule on return from

obviously...

>  interrupt.
>
>  Add arch_irqentry_exit_need_resched() as the default NOOP
>  implementation and hook it up into the need_resched() condition in
>  raw_irqentry_exit_cond_resched().
>
>  This allows ARM64 to implement the architecture specific version for
>  switchting over to the generic entry code.
>
> That explains things completely independently. Hmm?
>
> Thanks,
>
>         tglx