From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751186AbdEaNyX (ORCPT ); Wed, 31 May 2017 09:54:23 -0400 Received: from foss.arm.com ([217.140.101.70]:44484 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751146AbdEaNyW (ORCPT ); Wed, 31 May 2017 09:54:22 -0400 Subject: Re: [patch 1/2] genirq: Handle NOAUTOEN interrupt setup proper To: Thomas Gleixner , LKML References: <20170531095831.291595474@linutronix.de> <20170531100212.130986205@linutronix.de> Cc: jeffy , Brian Norris , dianders@chromium.org, tfiga@chromium.org From: Marc Zyngier Organization: ARM Ltd Message-ID: <333511ca-9c5d-1a78-1c29-ebd172d6fe2f@arm.com> Date: Wed, 31 May 2017 14:54:19 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101 Thunderbird/45.8.0 MIME-Version: 1.0 In-Reply-To: <20170531100212.130986205@linutronix.de> Content-Type: text/plain; charset=iso-8859-15 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Thomas, On 31/05/17 10:58, Thomas Gleixner wrote: > If an interrupt is marked NOAUTOEN then request_irq() installs the action, > but does not enable the interrupt via startup_irq(). The interrupt is > enabled via enable_irq() later from the driver. enable_irq() calls > irq_enable(). > > That means that for interrupts which have a irq_startup() callback this > callback is never invoked. Neither is irq_domain_activate_irq() invoked for > such interrupts. > > If an interrupt depends on irq_startup() or irq_domain_activate_irq() then > the enable via irq_enable() is not enough. > > Add a status flag IRQD_IRQ_STARTED_UP and use this to select the proper > mechanism in enable_irq(). Use the flag also to avoid pointless calls into > the low level functions. > > Signed-off-by: Thomas Gleixner > --- > include/linux/irq.h | 6 ++++ > kernel/irq/chip.c | 73 +++++++++++++++++++++++++++++++++++++------------ > kernel/irq/internals.h | 1 > kernel/irq/manage.c | 2 - > 4 files changed, 64 insertions(+), 18 deletions(-) > > --- a/include/linux/irq.h > +++ b/include/linux/irq.h > @@ -216,6 +216,7 @@ enum { > IRQD_WAKEUP_ARMED = (1 << 19), > IRQD_FORWARDED_TO_VCPU = (1 << 20), > IRQD_AFFINITY_MANAGED = (1 << 21), > + IRQD_IRQ_STARTED_UP = (1 << 22), > }; > > #define __irqd_to_state(d) ACCESS_PRIVATE((d)->common, state_use_accessors) > @@ -329,6 +330,11 @@ static inline void irqd_clr_activated(st > __irqd_to_state(d) &= ~IRQD_ACTIVATED; > } > > +static inline bool irqd_is_started_up(struct irq_data *d) nit: since we have set/clr_started, consider making this is_started (without the up)? Or add the _up to clr/set? > +{ > + return __irqd_to_state(d) & IRQD_IRQ_STARTED_UP; > +} > + > #undef __irqd_to_state > > static inline irq_hw_number_t irqd_to_hwirq(struct irq_data *d) > --- a/kernel/irq/chip.c > +++ b/kernel/irq/chip.c > @@ -185,37 +185,71 @@ static void irq_state_set_masked(struct > irqd_set(&desc->irq_data, IRQD_IRQ_MASKED); > } > > +static void irq_state_clr_started(struct irq_desc *desc) > +{ > + irqd_clear(&desc->irq_data, IRQD_IRQ_STARTED_UP); > +} > + > +static void irq_state_set_started(struct irq_desc *desc) > +{ > + irqd_set(&desc->irq_data, IRQD_IRQ_STARTED_UP); > +} > + > int irq_startup(struct irq_desc *desc, bool resend) > { > int ret = 0; > > - irq_state_clr_disabled(desc); > + if (irqd_is_started_up(&desc->irq_data)) > + return 0; > + How about: if (irqd_is_started_up(&desc->irq_data)) { irq_enable(desc); return 0; } > desc->depth = 0; > > irq_domain_activate_irq(&desc->irq_data); > if (desc->irq_data.chip->irq_startup) { > ret = desc->irq_data.chip->irq_startup(&desc->irq_data); > + irq_state_clr_disabled(desc); > irq_state_clr_masked(desc); > } else { > irq_enable(desc); > } > + irq_state_set_started(desc); > + > if (resend) > check_irq_resend(desc); > + > return ret; > } > > +static void __irq_disable(struct irq_desc *desc, bool mask); > + > void irq_shutdown(struct irq_desc *desc) > { > - irq_state_set_disabled(desc); > - desc->depth = 1; > - if (desc->irq_data.chip->irq_shutdown) > - desc->irq_data.chip->irq_shutdown(&desc->irq_data); > - else if (desc->irq_data.chip->irq_disable) > - desc->irq_data.chip->irq_disable(&desc->irq_data); > - else > - desc->irq_data.chip->irq_mask(&desc->irq_data); > + if (irqd_is_started_up(&desc->irq_data)) { > + desc->depth = 1; > + if (desc->irq_data.chip->irq_shutdown) { > + desc->irq_data.chip->irq_shutdown(&desc->irq_data); > + irq_state_set_disabled(desc); > + irq_state_set_masked(desc); > + } else { > + __irq_disable(desc, true); > + } > + irq_state_clr_started(desc); > + } > + /* > + * This must be called even if the interrupt was never started up, > + * because the activation can happen before the interrupt is > + * available for request/startup. It has it's own state tracking so > + * it's safe to call it unconditonally. unconditionally? > + */ > irq_domain_deactivate_irq(&desc->irq_data); > - irq_state_set_masked(desc); > +} > + > +void irq_enable_or_startup(struct irq_desc *desc) > +{ > + if (!irqd_is_started_up(&desc->irq_data)) > + irq_startup(desc, false); > + else > + irq_enable(desc); get rid of this new function... > } > > void irq_enable(struct irq_desc *desc) > @@ -228,6 +262,17 @@ void irq_enable(struct irq_desc *desc) > irq_state_clr_masked(desc); > } > > +static void __irq_disable(struct irq_desc *desc, bool mask) > +{ > + irq_state_set_disabled(desc); > + if (desc->irq_data.chip->irq_disable) { > + desc->irq_data.chip->irq_disable(&desc->irq_data); > + irq_state_set_masked(desc); > + } else if (mask) { > + mask_irq(desc); > + } > +} > + > /** > * irq_disable - Mark interrupt disabled > * @desc: irq descriptor which should be disabled > @@ -250,13 +295,7 @@ void irq_enable(struct irq_desc *desc) > */ > void irq_disable(struct irq_desc *desc) > { > - irq_state_set_disabled(desc); > - if (desc->irq_data.chip->irq_disable) { > - desc->irq_data.chip->irq_disable(&desc->irq_data); > - irq_state_set_masked(desc); > - } else if (irq_settings_disable_unlazy(desc)) { > - mask_irq(desc); > - } > + __irq_disable(desc, irq_settings_disable_unlazy(desc)); > } > > void irq_percpu_enable(struct irq_desc *desc, unsigned int cpu) > --- a/kernel/irq/internals.h > +++ b/kernel/irq/internals.h > @@ -68,6 +68,7 @@ extern void __enable_irq(struct irq_desc > > extern int irq_startup(struct irq_desc *desc, bool resend); > extern void irq_shutdown(struct irq_desc *desc); > +extern void irq_enable_or_startup(struct irq_desc *desc); > extern void irq_enable(struct irq_desc *desc); > extern void irq_disable(struct irq_desc *desc); > extern void irq_percpu_enable(struct irq_desc *desc, unsigned int cpu); > --- a/kernel/irq/manage.c > +++ b/kernel/irq/manage.c > @@ -533,7 +533,7 @@ void __enable_irq(struct irq_desc *desc) > goto err_out; > /* Prevent probing on this irq: */ > irq_settings_set_noprobe(desc); > - irq_enable(desc); > + irq_enable_or_startup(desc); and make this irq_startup(desc, false)? > check_irq_resend(desc); > /* fall-through */ > } > > Otherwise: Acked-by: Marc Zyngier M. -- Jazz is not dead. It just smells funny...