From 55567976629e58fde28fb70612ca73228271eef2 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 6 Oct 2020 10:10:20 +0100 Subject: [PATCH 1/4] genirq/irqdomain: Allow partial trimming of irq_data hierarchy It appears that some HW is ugly enough that not all the interrupts connected to a particular interrupt controller end up with the same hierarchy depth (some of them are terminated early). This leaves the irqchip hacker with only two choices, both equally bad: - create discrete domain chains, one for each "hierarchy depth", which is very hard to maintain - create fake hierarchy levels for the shallow paths, leading to all kind of problems (what are the safe hwirq values for these fake levels?) Implement the ability to cut short a single interrupt hierarchy from a level marked as being disconnected by using the new irq_domain_disconnect_hierarchy() helper. The irqdomain allocation code will then perform the trimming Signed-off-by: Marc Zyngier --- include/linux/irqdomain.h | 3 ++ kernel/irq/irqdomain.c | 99 +++++++++++++++++++++++++++++++++++++-- 2 files changed, 98 insertions(+), 4 deletions(-) diff --git a/include/linux/irqdomain.h b/include/linux/irqdomain.h index b37350c4fe37..a52b095bd404 100644 --- a/include/linux/irqdomain.h +++ b/include/linux/irqdomain.h @@ -509,6 +509,9 @@ extern void irq_domain_free_irqs_parent(struct irq_domain *domain, unsigned int irq_base, unsigned int nr_irqs); +extern int irq_domain_disconnect_hierarchy(struct irq_domain *domain, + unsigned int virq); + static inline bool irq_domain_is_hierarchy(struct irq_domain *domain) { return domain->flags & IRQ_DOMAIN_FLAG_HIERARCHY; diff --git a/kernel/irq/irqdomain.c b/kernel/irq/irqdomain.c index 76cd7ebd1178..cf8b374b892d 100644 --- a/kernel/irq/irqdomain.c +++ b/kernel/irq/irqdomain.c @@ -1136,6 +1136,17 @@ static struct irq_data *irq_domain_insert_irq_data(struct irq_domain *domain, return irq_data; } +static void __irq_domain_free_hierarchy(struct irq_data *irq_data) +{ + struct irq_data *tmp; + + while (irq_data) { + tmp = irq_data; + irq_data = irq_data->parent_data; + kfree(tmp); + } +} + static void irq_domain_free_irq_data(unsigned int virq, unsigned int nr_irqs) { struct irq_data *irq_data, *tmp; @@ -1147,12 +1158,83 @@ static void irq_domain_free_irq_data(unsigned int virq, unsigned int nr_irqs) irq_data->parent_data = NULL; irq_data->domain = NULL; - while (tmp) { - irq_data = tmp; - tmp = tmp->parent_data; - kfree(irq_data); + __irq_domain_free_hierarchy(tmp); + } +} + +/** + * irq_domain_disconnect_hierarchy - Mark the first unused level of a hierarchy + * @domain: IRQ domain from which the hierarchy is to be disconnected + * @virq: IRQ number where the hierarchy is to be trimmed + * + * Marks the @virq level belonging to @domain as disconnected. + * Returns -EINVAL if @virq doesn't have a valid irq_data pointing + * to @domain. + * + * Its only use is to be able to trim levels of hierarchy that do not + * have any real meaning for this interrupt, and that the driver marks + * as such from its .alloc() callback. + */ +int irq_domain_disconnect_hierarchy(struct irq_domain *domain, + unsigned int virq) +{ + struct irq_data *irqd; + + irqd = irq_domain_get_irq_data(domain, virq); + if (!irqd) + return -EINVAL; + + irqd->chip = ERR_PTR(-ENOTCONN); + return 0; +} + +static int irq_domain_trim_hierarchy(unsigned int virq) +{ + struct irq_data *tail, *irqd, *irq_data; + + irq_data = irq_get_irq_data(virq); + tail = NULL; + + /* The first entry must have a valid irqchip */ + if (!irq_data->chip || IS_ERR(irq_data->chip)) + return -EINVAL; + + /* + * Validate that the irq_data chain is sane in the presence of + * a hierarchy trimming marker. + */ + for (irqd = irq_data->parent_data; irqd; irq_data = irqd, irqd = irqd->parent_data) { + /* Can't have a valid irqchip after a trim marker */ + if (irqd->chip && tail) + return -EINVAL; + + /* Can't have an empty irqchip before a trim marker */ + if (!irqd->chip && !tail) + return -EINVAL; + + if (IS_ERR(irqd->chip)) { + /* Only -ENOTCONN is a valid trim marker */ + if (PTR_ERR(irqd->chip) != -ENOTCONN) + return -EINVAL; + + tail = irq_data; } } + + /* No trim marker, nothing to do */ + if (!tail) + return 0; + + pr_info("IRQ%d: trimming hierarchy from %s\n", + virq, tail->parent_data->domain->name); + + /* Sever the inner part of the hierarchy... */ + irqd = tail; + tail = tail->parent_data; + irqd->parent_data = NULL; + __irq_domain_free_hierarchy(tail); + + return 0; } static int irq_domain_alloc_irq_data(struct irq_domain *domain, @@ -1362,6 +1444,15 @@ int __irq_domain_alloc_irqs(struct irq_domain *domain, int irq_base, mutex_unlock(&irq_domain_mutex); goto out_free_irq_data; } + + for (i = 0; i < nr_irqs; i++) { + ret = irq_domain_trim_hierarchy(virq + i); + if (ret) { + mutex_unlock(&irq_domain_mutex); + goto out_free_irq_data; + } + } + for (i = 0; i < nr_irqs; i++) irq_domain_insert_irq(virq + i); mutex_unlock(&irq_domain_mutex); From 986ec63d4482292570b579ac98b151acf8bdd1de Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Mon, 5 Oct 2020 10:27:27 +0100 Subject: [PATCH 2/4] gpio: tegra186: Allow optional irq parent callbacks Make the tegra186 GPIO driver resistent to variable depth interrupt hierarchy, which we are about to introduce. No functionnal change yet. Signed-off-by: Marc Zyngier --- drivers/gpio/gpio-tegra186.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/drivers/gpio/gpio-tegra186.c b/drivers/gpio/gpio-tegra186.c index 178e9128ded0..9500074b1f1b 100644 --- a/drivers/gpio/gpio-tegra186.c +++ b/drivers/gpio/gpio-tegra186.c @@ -430,7 +430,18 @@ static int tegra186_irq_set_type(struct irq_data *data, unsigned int type) else irq_set_handler_locked(data, handle_edge_irq); - return irq_chip_set_type_parent(data, type); + if (data->parent_data) + return irq_chip_set_type_parent(data, type); + + return 0; +} + +static int tegra186_irq_set_wake(struct irq_data *data, unsigned int on) +{ + if (data->parent_data) + return irq_chip_set_wake_parent(data, on); + + return 0; } static void tegra186_gpio_irq(struct irq_desc *desc) @@ -678,7 +689,7 @@ static int tegra186_gpio_probe(struct platform_device *pdev) gpio->intc.irq_mask = tegra186_irq_mask; gpio->intc.irq_unmask = tegra186_irq_unmask; gpio->intc.irq_set_type = tegra186_irq_set_type; - gpio->intc.irq_set_wake = irq_chip_set_wake_parent; + gpio->intc.irq_set_wake = tegra186_irq_set_wake; irq = &gpio->gpio.irq; irq->chip = &gpio->intc; From 8681cc33f817842df7ebe3c36558d97f5497a177 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Sun, 4 Oct 2020 21:16:24 +0100 Subject: [PATCH 3/4] soc/tegra: pmc: Allow optional irq parent callbacks Make the PMC driver resistent to variable depth interrupt hierarchy, which we are about to introduce. Signed-off-by: Marc Zyngier --- drivers/soc/tegra/pmc.c | 36 ++++++++++++++++++++++++++++++++---- 1 file changed, 32 insertions(+), 4 deletions(-) diff --git a/drivers/soc/tegra/pmc.c b/drivers/soc/tegra/pmc.c index d332e5d9abac..b39536c68f45 100644 --- a/drivers/soc/tegra/pmc.c +++ b/drivers/soc/tegra/pmc.c @@ -2184,6 +2184,34 @@ static int tegra186_pmc_irq_set_type(struct irq_data *data, unsigned int type) return 0; } +static void tegra_irq_mask_parent(struct irq_data *data) +{ + if (data->parent_data) + irq_chip_mask_parent(data); +} + +static void tegra_irq_unmask_parent(struct irq_data *data) +{ + if (data->parent_data) + irq_chip_unmask_parent(data); +} + +static void tegra_irq_eoi_parent(struct irq_data *data) +{ + if (data->parent_data) + irq_chip_eoi_parent(data); +} + +static int tegra_irq_set_affinity_parent(struct irq_data *data, + const struct cpumask *dest, + bool force) +{ + if (data->parent_data) + return irq_chip_set_affinity_parent(data, dest, force); + + return -EINVAL; +} + static int tegra_pmc_irq_init(struct tegra_pmc *pmc) { struct irq_domain *parent = NULL; @@ -2199,10 +2227,10 @@ static int tegra_pmc_irq_init(struct tegra_pmc *pmc) return 0; pmc->irq.name = dev_name(pmc->dev); - pmc->irq.irq_mask = irq_chip_mask_parent; - pmc->irq.irq_unmask = irq_chip_unmask_parent; - pmc->irq.irq_eoi = irq_chip_eoi_parent; - pmc->irq.irq_set_affinity = irq_chip_set_affinity_parent; + pmc->irq.irq_mask = tegra_irq_mask_parent; + pmc->irq.irq_unmask = tegra_irq_unmask_parent; + pmc->irq.irq_eoi = tegra_irq_eoi_parent; + pmc->irq.irq_set_affinity = tegra_irq_set_affinity_parent; pmc->irq.irq_set_type = pmc->soc->irq_set_type; pmc->irq.irq_set_wake = pmc->soc->irq_set_wake; From c351ab7bf2a565951172cadbdebe686137c3fd43 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Sun, 4 Oct 2020 18:27:04 +0100 Subject: [PATCH 4/4] soc/tegra: pmc: Don't create fake interrupt hierarchy levels The Tegra PMC driver does ungodly things with the interrupt hierarchy, repeatedly corrupting it by pulling hwirq numbers out of thin air, overriding existing IRQ mappings and changing the handling flow of unsuspecting users. All of this is done in the name of preserving the interrupt hierarchy even when these levels do not exist in the HW. Together with the use of proper IRQs for IPIs, this leads to an unbootable system as the rescheduling IPI gets repeatedly repurposed for random drivers... Instead, let's simply mark the level from which the hierarchy does not make sense for the HW, and let the core code trim the usused levels from the hierarchy. Signed-off-by: Marc Zyngier --- drivers/soc/tegra/pmc.c | 55 ++++++----------------------------------- 1 file changed, 7 insertions(+), 48 deletions(-) diff --git a/drivers/soc/tegra/pmc.c b/drivers/soc/tegra/pmc.c index b39536c68f45..b0bba8ab75bb 100644 --- a/drivers/soc/tegra/pmc.c +++ b/drivers/soc/tegra/pmc.c @@ -1990,44 +1990,17 @@ static int tegra_pmc_irq_alloc(struct irq_domain *domain, unsigned int virq, event->id, &pmc->irq, pmc); - /* - * GPIOs don't have an equivalent interrupt in the - * parent controller (GIC). However some code, such - * as the one in irq_get_irqchip_state(), require a - * valid IRQ chip to be set. Make sure that's the - * case by passing NULL here, which will install a - * dummy IRQ chip for the interrupt in the parent - * domain. - */ - if (domain->parent) - irq_domain_set_hwirq_and_chip(domain->parent, - virq, 0, NULL, - NULL); - + /* GPIO hierarchies stop at the PMC level */ + if (!err && domain->parent) + err = irq_domain_disconnect_hierarchy(domain->parent, + virq); break; } } - /* - * For interrupts that don't have associated wake events, assign a - * dummy hardware IRQ number. This is used in the ->irq_set_type() - * and ->irq_set_wake() callbacks to return early for these IRQs. - */ - if (i == soc->num_wake_events) { - err = irq_domain_set_hwirq_and_chip(domain, virq, ULONG_MAX, - &pmc->irq, pmc); - - /* - * Interrupts without a wake event don't have a corresponding - * interrupt in the parent controller (GIC). Pass NULL for the - * chip here, which causes a dummy IRQ chip to be installed - * for the interrupt in the parent domain, to make this - * explicit. - */ - if (domain->parent) - irq_domain_set_hwirq_and_chip(domain->parent, virq, 0, - NULL, NULL); - } + /* If there is no wake-up event, there is no PMC mapping */ + if (i == soc->num_wake_events) + err = irq_domain_disconnect_hierarchy(domain, virq); return err; } @@ -2043,9 +2016,6 @@ static int tegra210_pmc_irq_set_wake(struct irq_data *data, unsigned int on) unsigned int offset, bit; u32 value; - if (data->hwirq == ULONG_MAX) - return 0; - offset = data->hwirq / 32; bit = data->hwirq % 32; @@ -2080,9 +2050,6 @@ static int tegra210_pmc_irq_set_type(struct irq_data *data, unsigned int type) unsigned int offset, bit; u32 value; - if (data->hwirq == ULONG_MAX) - return 0; - offset = data->hwirq / 32; bit = data->hwirq % 32; @@ -2123,10 +2090,6 @@ static int tegra186_pmc_irq_set_wake(struct irq_data *data, unsigned int on) unsigned int offset, bit; u32 value; - /* nothing to do if there's no associated wake event */ - if (WARN_ON(data->hwirq == ULONG_MAX)) - return 0; - offset = data->hwirq / 32; bit = data->hwirq % 32; @@ -2154,10 +2117,6 @@ static int tegra186_pmc_irq_set_type(struct irq_data *data, unsigned int type) struct tegra_pmc *pmc = irq_data_get_irq_chip_data(data); u32 value; - /* nothing to do if there's no associated wake event */ - if (data->hwirq == ULONG_MAX) - return 0; - value = readl(pmc->wake + WAKE_AOWAKE_CNTRL(data->hwirq)); switch (type) {