| Message ID | 20220209153535.818830-7-mark.rutland@arm.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | arm64 / sched/preempt: support PREEMPT_DYNAMIC with static keys | expand |
On Wed, Feb 09, 2022 at 03:35:34PM +0000, Mark Rutland wrote: > For historical reasons, the decision of whether or not to preempt is > spread across arm64_preempt_schedule_irq() and __el1_irq(), and it would > be clearer if this were all in one place. > > Also, arm64_preempt_schedule_irq() calls lockdep_assert_irqs_disabled(), > but this is redundant, as we have a subsequent identical assertion in > __exit_to_kernel_mode(), and preempt_schedule_irq() will > BUG_ON(!irqs_disabled()) anyway. > > This patch removes the redundant assertion and centralizes the > preemption decision making within arm64_preempt_schedule_irq(). > > Other than the slight change to assertion behaviour, there should be no > functional change as a result of this patch. > > Signed-off-by: Mark Rutland <mark.rutland@arm.com> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: James Morse <james.morse@arm.com> > Cc: Joey Gouly <joey.gouly@arm.com> > Cc: Valentin Schneider <valentin.schneider@arm.com> > Cc: Will Deacon <will@kernel.org> I acked this patch in v2, has anything changed? Well, here it is again: Acked-by: Catalin Marinas <catalin.marinas@arm.com> BTW, you have a typo in the subject.
On Wed, Feb 09, 2022 at 06:10:52PM +0000, Catalin Marinas wrote: > On Wed, Feb 09, 2022 at 03:35:34PM +0000, Mark Rutland wrote: > > For historical reasons, the decision of whether or not to preempt is > > spread across arm64_preempt_schedule_irq() and __el1_irq(), and it would > > be clearer if this were all in one place. > > > > Also, arm64_preempt_schedule_irq() calls lockdep_assert_irqs_disabled(), > > but this is redundant, as we have a subsequent identical assertion in > > __exit_to_kernel_mode(), and preempt_schedule_irq() will > > BUG_ON(!irqs_disabled()) anyway. > > > > This patch removes the redundant assertion and centralizes the > > preemption decision making within arm64_preempt_schedule_irq(). > > > > Other than the slight change to assertion behaviour, there should be no > > functional change as a result of this patch. > > > > Signed-off-by: Mark Rutland <mark.rutland@arm.com> > > Cc: Catalin Marinas <catalin.marinas@arm.com> > > Cc: James Morse <james.morse@arm.com> > > Cc: Joey Gouly <joey.gouly@arm.com> > > Cc: Valentin Schneider <valentin.schneider@arm.com> > > Cc: Will Deacon <will@kernel.org> > > I acked this patch in v2, has anything changed? Well, here it is again: Sorry; I had meant to add your acks. This patch is the same as in v2; the other patch has some minor changes as in the cover letter (adding includes and always exposing a couple of function prototypes). > Acked-by: Catalin Marinas <catalin.marinas@arm.com> Thanks! > BTW, you have a typo in the subject. I'll go fix that now. Mark.
diff --git a/arch/arm64/kernel/entry-common.c b/arch/arm64/kernel/entry-common.c index ef7fcefb96bd..2c639b6b676d 100644 --- a/arch/arm64/kernel/entry-common.c +++ b/arch/arm64/kernel/entry-common.c @@ -222,7 +222,16 @@ static void noinstr arm64_exit_el1_dbg(struct pt_regs *regs) static void __sched arm64_preempt_schedule_irq(void) { - lockdep_assert_irqs_disabled(); + if (!IS_ENABLED(CONFIG_PREEMPTION)) + return; + + /* + * Note: thread_info::preempt_count includes both thread_info::count + * and thread_info::need_resched, and is not equivalent to + * preempt_count(). + */ + if (READ_ONCE(current_thread_info()->preempt_count) != 0) + return; /* * DAIF.DA are cleared at the start of IRQ/FIQ handling, and when GIC @@ -438,14 +447,7 @@ static __always_inline void __el1_irq(struct pt_regs *regs, do_interrupt_handler(regs, handler); irq_exit_rcu(); - /* - * Note: thread_info::preempt_count includes both thread_info::count - * and thread_info::need_resched, and is not equivalent to - * preempt_count(). - */ - if (IS_ENABLED(CONFIG_PREEMPTION) && - READ_ONCE(current_thread_info()->preempt_count) == 0) - arm64_preempt_schedule_irq(); + arm64_preempt_schedule_irq(); exit_to_kernel_mode(regs); }
For historical reasons, the decision of whether or not to preempt is spread across arm64_preempt_schedule_irq() and __el1_irq(), and it would be clearer if this were all in one place. Also, arm64_preempt_schedule_irq() calls lockdep_assert_irqs_disabled(), but this is redundant, as we have a subsequent identical assertion in __exit_to_kernel_mode(), and preempt_schedule_irq() will BUG_ON(!irqs_disabled()) anyway. This patch removes the redundant assertion and centralizes the preemption decision making within arm64_preempt_schedule_irq(). Other than the slight change to assertion behaviour, there should be no functional change as a result of this patch. Signed-off-by: Mark Rutland <mark.rutland@arm.com> Cc: Catalin Marinas <catalin.marinas@arm.com> Cc: James Morse <james.morse@arm.com> Cc: Joey Gouly <joey.gouly@arm.com> Cc: Valentin Schneider <valentin.schneider@arm.com> Cc: Will Deacon <will@kernel.org> --- arch/arm64/kernel/entry-common.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-)