Messages in this thread | | | Date | Fri, 01 Oct 2021 10:21:06 +0100 | From | Marc Zyngier <> | Subject | Re: [PATCHv3 2/3] arm64: entry: refactor EL1 interrupt entry logic |
| |
On Thu, 30 Sep 2021 14:17:07 +0100, Pingfan Liu <kernelfans@gmail.com> wrote: > > From: Mark Rutland <mark.rutland@arm.com> > > Currently we distinguish IRQ and definitely-PNMI at entry/exit time > via the enter_el1_irq_or_nmi() and enter_el1_irq_or_nmi() helpers. In > subsequent patches we'll need to handle the two cases more distinctly > in the body of the exception handler. > > To make this possible, this patch refactors el1_interrupt to be a > top-level dispatcher to separate handlers for the IRQ and PNMI cases, > removing the need for the enter_el1_irq_or_nmi() and > exit_el1_irq_or_nmi() helpers. > > Note that since arm64_enter_nmi() calls __nmi_enter(), which > increments the preemt_count, we could never preempt when handling a > PNMI. We now only check for preemption in the IRQ case, which makes > this clearer. > > There should be no functional change as a result of this patch. > > Signed-off-by: Mark Rutland <mark.rutland@arm.com> > Cc: "Paul E. McKenney" <paulmck@kernel.org> > Cc: Catalin Marinas <catalin.marinas@arm.com> > Cc: Will Deacon <will@kernel.org> > Cc: Marc Zyngier <maz@kernel.org> > Cc: Joey Gouly <joey.gouly@arm.com> > Cc: Sami Tolvanen <samitolvanen@google.com> > Cc: Julien Thierry <julien.thierry@arm.com> > Cc: Thomas Gleixner <tglx@linutronix.de> > Cc: Yuichi Ito <ito-yuichi@fujitsu.com> > Cc: Pingfan Liu <kernelfans@gmail.com> > Cc: linux-kernel@vger.kernel.org > To: linux-arm-kernel@lists.infradead.org > --- > arch/arm64/kernel/entry-common.c | 44 ++++++++++++++++---------------- > 1 file changed, 22 insertions(+), 22 deletions(-) > > diff --git a/arch/arm64/kernel/entry-common.c b/arch/arm64/kernel/entry-common.c > index 32f9796c4ffe..5f1473319fb0 100644 > --- a/arch/arm64/kernel/entry-common.c > +++ b/arch/arm64/kernel/entry-common.c > @@ -219,22 +219,6 @@ static void noinstr arm64_exit_el1_dbg(struct pt_regs *regs) > lockdep_hardirqs_on(CALLER_ADDR0); > } > > -static void noinstr enter_el1_irq_or_nmi(struct pt_regs *regs) > -{ > - if (IS_ENABLED(CONFIG_ARM64_PSEUDO_NMI) && !interrupts_enabled(regs)) > - arm64_enter_nmi(regs); > - else > - enter_from_kernel_mode(regs); > -} > - > -static void noinstr exit_el1_irq_or_nmi(struct pt_regs *regs) > -{ > - if (IS_ENABLED(CONFIG_ARM64_PSEUDO_NMI) && !interrupts_enabled(regs)) > - arm64_exit_nmi(regs); > - else > - exit_to_kernel_mode(regs); > -} > - > static void __sched arm64_preempt_schedule_irq(void) > { > lockdep_assert_irqs_disabled(); > @@ -432,14 +416,19 @@ asmlinkage void noinstr el1h_64_sync_handler(struct pt_regs *regs) > } > } > > -static void noinstr el1_interrupt(struct pt_regs *regs, > - void (*handler)(struct pt_regs *)) > +static __always_inline void > +__el1_pnmi(struct pt_regs *regs, void (*handler)(struct pt_regs *)) > { > - write_sysreg(DAIF_PROCCTX_NOIRQ, daif); > - > - enter_el1_irq_or_nmi(regs); > + arm64_enter_nmi(regs); > do_interrupt_handler(regs, handler); > + arm64_exit_nmi(regs); > +} > > +static __always_inline void > +__el1_interrupt(struct pt_regs *regs, void (*handler)(struct pt_regs *)) > +{ > + enter_from_kernel_mode(regs); > + do_interrupt_handler(regs, handler); > /* > * Note: thread_info::preempt_count includes both thread_info::count > * and thread_info::need_resched, and is not equivalent to > @@ -448,8 +437,19 @@ static void noinstr el1_interrupt(struct pt_regs *regs, > if (IS_ENABLED(CONFIG_PREEMPTION) && > READ_ONCE(current_thread_info()->preempt_count) == 0) > arm64_preempt_schedule_irq(); > + exit_to_kernel_mode(regs); > +} > + > +static void noinstr el1_interrupt(struct pt_regs *regs, > + void (*handler)(struct pt_regs *)) > +{ > + write_sysreg(DAIF_PROCCTX_NOIRQ, daif); > + > + if (IS_ENABLED(CONFIG_ARM64_PSEUDO_NMI) && !interrupts_enabled(regs)) > + __el1_pnmi(regs, handler); > + else > + __el1_interrupt(regs, handler); > > - exit_el1_irq_or_nmi(regs);
nit: spurious blank line.
> } > > asmlinkage void noinstr el1h_64_irq_handler(struct pt_regs *regs)
With Mark's remark addressed,
Reviewed-by: Marc Zyngier <maz@kernel.org>
M.
-- Without deviation from the norm, progress is not possible.
| |