From: Geert Uytterhoeven Date: Wed, 9 Sep 2015 11:42:54 +0000 (+0200) Subject: irqchip/renesas-intc-irqpin: Use a separate lockdep class X-Git-Tag: Ubuntu-5.2.0-15.16~10613^2~27 X-Git-Url: https://git.proxmox.com/?a=commitdiff_plain;h=769b5cf78e6c653c2f513649ee6c4e7a06723872;p=mirror_ubuntu-eoan-kernel.git irqchip/renesas-intc-irqpin: Use a separate lockdep class The renesas-intc-irqpin interrupt controller is cascaded to the GIC. Hence when propagating wake-up settings to its parent interrupt controller, the following lockdep warning is printed: ============================================= [ INFO: possible recursive locking detected ] 4.2.0-armadillo-10725-g50fcd7643c034198 #781 Not tainted --------------------------------------------- s2ram/1179 is trying to acquire lock: (&irq_desc_lock_class){-.-...}, at: [] __irq_get_desc_lock+0x78/0x94 but task is already holding lock: (&irq_desc_lock_class){-.-...}, at: [] __irq_get_desc_lock+0x78/0x94 other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&irq_desc_lock_class); lock(&irq_desc_lock_class); *** DEADLOCK *** May be due to missing lock nesting notation 7 locks held by s2ram/1179: #0: (sb_writers#7){.+.+.+}, at: [] __sb_start_write+0x64/0xb8 #1: (&of->mutex){+.+.+.}, at: [] kernfs_fop_write+0x78/0x1a0 #2: (s_active#23){.+.+.+}, at: [] kernfs_fop_write+0x80/0x1a0 #3: (autosleep_lock){+.+.+.}, at: [] pm_autosleep_lock+0x18/0x20 #4: (pm_mutex){+.+.+.}, at: [] pm_suspend+0x54/0x248 #5: (&dev->mutex){......}, at: [] __device_suspend+0xdc/0x240 #6: (&irq_desc_lock_class){-.-...}, at: [] __irq_get_desc_lock+0x78/0x94 stack backtrace: CPU: 0 PID: 1179 Comm: s2ram Not tainted 4.2.0-armadillo-10725-g50fcd7643c034198 Hardware name: Generic R8A7740 (Flattened Device Tree) [] (dump_backtrace) from [] (show_stack+0x18/0x1c) [] (show_stack) from [] (dump_stack+0x20/0x28) [] (dump_stack) from [] (__lock_acquire+0x67c/0x1b88) [] (__lock_acquire) from [] (lock_acquire+0x9c/0xbc) [] (lock_acquire) from [] (_raw_spin_lock_irqsave+0x44/0x58) [] (_raw_spin_lock_irqsave) from [] (__irq_get_desc_lock+0x78/0x94 [] (__irq_get_desc_lock) from [] (irq_set_irq_wake+0x28/0x100) [] (irq_set_irq_wake) from [] (intc_irqpin_irq_set_wake+0x24/0x4c) [] (intc_irqpin_irq_set_wake) from [] (set_irq_wake_real+0x3c/0x50 [] (set_irq_wake_real) from [] (irq_set_irq_wake+0x64/0x100) [] (irq_set_irq_wake) from [] (gpio_keys_suspend+0x60/0xa0) [] (gpio_keys_suspend) from [] (platform_pm_suspend+0x3c/0x5c) Avoid this false positive by using a separate lockdep class for INTC External IRQ Pin interrupts. Signed-off-by: Geert Uytterhoeven Cc: Grygorii Strashko Cc: Magnus Damm Cc: Jason Cooper Link: http://lkml.kernel.org/r/1441798974-25716-3-git-send-email-geert%2Brenesas@glider.be Signed-off-by: Thomas Gleixner --- diff --git a/drivers/irqchip/irq-renesas-intc-irqpin.c b/drivers/irqchip/irq-renesas-intc-irqpin.c index 0670ab4e3897..67797eebbd5e 100644 --- a/drivers/irqchip/irq-renesas-intc-irqpin.c +++ b/drivers/irqchip/irq-renesas-intc-irqpin.c @@ -332,6 +332,12 @@ static irqreturn_t intc_irqpin_shared_irq_handler(int irq, void *dev_id) return status; } +/* + * This lock class tells lockdep that INTC External IRQ Pin irqs are in a + * different category than their parents, so it won't report false recursion. + */ +static struct lock_class_key intc_irqpin_irq_lock_class; + static int intc_irqpin_irq_domain_map(struct irq_domain *h, unsigned int virq, irq_hw_number_t hw) { @@ -342,6 +348,7 @@ static int intc_irqpin_irq_domain_map(struct irq_domain *h, unsigned int virq, intc_irqpin_dbg(&p->irq[hw], "map"); irq_set_chip_data(virq, h->host_data); + irq_set_lockdep_class(virq, &intc_irqpin_irq_lock_class); irq_set_chip_and_handler(virq, &p->irq_chip, handle_level_irq); set_irq_flags(virq, IRQF_VALID); /* kill me now */ return 0;