From patchwork Thu Jun 19 10:38:13 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Daniel Thompson X-Patchwork-Id: 32199 Return-Path: X-Original-To: linaro@patches.linaro.org Delivered-To: linaro@patches.linaro.org Received: from mail-pb0-f69.google.com (mail-pb0-f69.google.com [209.85.160.69]) by ip-10-151-82-157.ec2.internal (Postfix) with ESMTPS id 67341203F4 for ; Thu, 19 Jun 2014 10:38:54 +0000 (UTC) Received: by mail-pb0-f69.google.com with SMTP id uo5sf7274960pbc.4 for ; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20130820; h=x-gm-message-state:mime-version:delivered-to:from:to:cc:subject :date:message-id:in-reply-to:references:x-original-sender :x-original-authentication-results:precedence:mailing-list:list-id :list-post:list-help:list-archive:list-unsubscribe; bh=f6ZmXbiTBpF7Tfqo+4aElcq1besViB6KT5q97C9XAdQ=; b=fDJep+XW9RkMcwk9ye11JJUeVaoH67W2SZQXKMQNz99lqRFoXpkong3QLa/ewtln2k JH9XGsws4GoAaOsaF8e5OVrQtNNt81hd3VGv8X5zQddtqTJHbKILrDUnaHXrBfgOtlbj frWPiGDYRbpCHeXkKKazUWvsMNKriIPwmN3wxalVB7O7lRlVRWIzIyeJfKFXme+fvjOe Xh4zWqhOKmrqFy7/tEGNdMOzoV7tl2qfXCNMLDtppeAhR+fUZqaJr5eI5p9GjYStAzxV XuUNCkWzi6A/nYO9kI20G1f1DR2Ry8ObpQBk9drldjlMCiFb7tiuzCbhhyUbG6H2aqnq 0igg== X-Gm-Message-State: ALoCoQm1jLduyb+zOiptH5XT63jJYCs/HbF4ZR19u+gzPYvdJC4rTF7vVhfZL8R9lA/9+aH6nRhn X-Received: by 10.66.189.163 with SMTP id gj3mr2148107pac.32.1403174333518; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) MIME-Version: 1.0 X-BeenThere: patchwork-forward@linaro.org Received: by 10.140.24.42 with SMTP id 39ls435584qgq.26.gmail; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) X-Received: by 10.52.185.39 with SMTP id ez7mr2795551vdc.7.1403174333420; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) Received: from mail-vc0-f173.google.com (mail-vc0-f173.google.com [209.85.220.173]) by mx.google.com with ESMTPS id yn17si2171435vdb.49.2014.06.19.03.38.53 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Thu, 19 Jun 2014 03:38:53 -0700 (PDT) Received-SPF: pass (google.com: domain of patch+caf_=patchwork-forward=linaro.org@linaro.org designates 209.85.220.173 as permitted sender) client-ip=209.85.220.173; Received: by mail-vc0-f173.google.com with SMTP id lf12so2030025vcb.18 for ; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) X-Received: by 10.52.185.72 with SMTP id fa8mr2840260vdc.12.1403174333316; Thu, 19 Jun 2014 03:38:53 -0700 (PDT) X-Forwarded-To: patchwork-forward@linaro.org X-Forwarded-For: patch@linaro.org patchwork-forward@linaro.org Delivered-To: patches@linaro.org Received: by 10.221.54.6 with SMTP id vs6csp351985vcb; Thu, 19 Jun 2014 03:38:52 -0700 (PDT) X-Received: by 10.180.88.131 with SMTP id bg3mr4582266wib.57.1403174332375; Thu, 19 Jun 2014 03:38:52 -0700 (PDT) Received: from mail-we0-f175.google.com (mail-we0-f175.google.com [74.125.82.175]) by mx.google.com with ESMTPS id lb4si6683480wjb.84.2014.06.19.03.38.52 for (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128); Thu, 19 Jun 2014 03:38:52 -0700 (PDT) Received-SPF: pass (google.com: domain of daniel.thompson@linaro.org designates 74.125.82.175 as permitted sender) client-ip=74.125.82.175; Received: by mail-we0-f175.google.com with SMTP id k48so2158481wev.34 for ; Thu, 19 Jun 2014 03:38:52 -0700 (PDT) X-Received: by 10.180.210.174 with SMTP id mv14mr5236671wic.47.1403174331925; Thu, 19 Jun 2014 03:38:51 -0700 (PDT) Received: from sundance.lan (cpc4-aztw19-0-0-cust157.18-1.cable.virginm.net. [82.33.25.158]) by mx.google.com with ESMTPSA id hi2sm7881147wjb.29.2014.06.19.03.38.49 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Thu, 19 Jun 2014 03:38:51 -0700 (PDT) From: Daniel Thompson To: Jason Wessel Cc: Daniel Thompson , kgdb-bugreport@lists.sourceforge.net, patches@linaro.org, linux-arm-kernel@lists.infradead.org, linaro-kernel@lists.linaro.org, linux-kernel@vger.kernel.org, John Stultz , Anton Vorontsov , Colin Cross , Dirk Behme , kernel-team@android.com, Russell King , Rob Herring , Pawel Moll , Mark Rutland , Ian Campbell , Kumar Gala , Thomas Gleixner , Greg Kroah-Hartman , Jiri Slaby , "David A. Long" , Nicolas Pitre , Catalin Marinas , Frederic Weisbecker , Linus Walleij , Christoffer Dall , kernel@stlinux.com, devicetree@vger.kernel.org, linux-serial@vger.kernel.org, Jason Cooper , Nicolas Pitre , Sricharan R Subject: [PATCH v4 03/13] irqchip: gic: Provide support for interrupt grouping Date: Thu, 19 Jun 2014 11:38:13 +0100 Message-Id: <1403174303-25456-4-git-send-email-daniel.thompson@linaro.org> X-Mailer: git-send-email 1.9.3 In-Reply-To: <1403174303-25456-1-git-send-email-daniel.thompson@linaro.org> References: <1401961994-18033-1-git-send-email-daniel.thompson@linaro.org> <1403174303-25456-1-git-send-email-daniel.thompson@linaro.org> X-Removed-Original-Auth: Dkim didn't pass. X-Original-Sender: daniel.thompson@linaro.org X-Original-Authentication-Results: mx.google.com; spf=pass (google.com: domain of patch+caf_=patchwork-forward=linaro.org@linaro.org designates 209.85.220.173 as permitted sender) smtp.mail=patch+caf_=patchwork-forward=linaro.org@linaro.org Precedence: list Mailing-list: list patchwork-forward@linaro.org; contact patchwork-forward+owners@linaro.org List-ID: X-Google-Group-Id: 836684582541 List-Post: , List-Help: , List-Archive: List-Unsubscribe: , All GIC hardware except GICv1-without-TrustZone support provides a means to group exceptions into group 0 (which can optionally be signally using use FIQ) and group 1. The kernel currently provides no means to exploit this. This patch alters the initialization of the GIC to place all interrupts into group 1 which is the foundational requirement to meaningfully use FIQ. Note that the hardware functionality is unavailable to the kernel when a secure monitor is present because access to the grouping registers are prohibited outside "secure world" (this feature allows grouping to be used to allow hardware peripherals to send interrupts into the secure world). The GIC driver will automatically detect this and disable its attempts to group interrupts. On systems without TrustZone support the kernel has the power to route interrupt sources to FIQ, potentially allowing a driver to exploit the NMI-like properties of FIQ. Tested on Freescale i.MX6 (quad A9), STiH416 (dual A9) and a self-written qemu GICv2 model. Signed-off-by: Daniel Thompson Cc: Thomas Gleixner Cc: Jason Cooper Cc: Nicolas Pitre Cc: Christoffer Dall Cc: Sricharan R Acked-by: Dirk Behme --- drivers/irqchip/irq-gic.c | 99 ++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 94 insertions(+), 5 deletions(-) diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c index 7e11c9d..bbffca3 100644 --- a/drivers/irqchip/irq-gic.c +++ b/drivers/irqchip/irq-gic.c @@ -42,6 +42,9 @@ #include #include +#ifdef CONFIG_FIQ +#include +#endif #include #include #include @@ -68,6 +71,9 @@ struct gic_chip_data { #ifdef CONFIG_GIC_NON_BANKED void __iomem *(*get_base)(union gic_base *); #endif +#ifdef CONFIG_FIQ + bool fiq_enable; +#endif }; static DEFINE_RAW_SPINLOCK(irq_controller_lock); @@ -131,6 +137,16 @@ static inline void gic_set_base_accessor(struct gic_chip_data *data, #define gic_set_base_accessor(d, f) #endif +#ifdef CONFIG_FIQ +static inline bool gic_data_fiq_enable(struct gic_chip_data *data) +{ + return data->fiq_enable; +} +#else +static inline bool gic_data_fiq_enable( + struct gic_chip_data *data) { return false; } +#endif + static inline void __iomem *gic_dist_base(struct irq_data *d) { struct gic_chip_data *gic_data = irq_data_get_irq_chip_data(d); @@ -349,6 +365,42 @@ static struct irq_chip gic_chip = { .irq_set_wake = gic_set_wake, }; +#ifdef CONFIG_FIQ +static void __init gic_init_fiq(struct gic_chip_data *gic, + irq_hw_number_t first_irq, + unsigned int num_irqs) +{ + void __iomem *dist_base = gic_data_dist_base(gic_data); + unsigned int i; + + /* + * FIQ can only be supported on platforms without an extended irq_eoi + * method (otherwise we take a lock during eoi handling). + */ + if (gic_arch_extn.irq_eoi) + return; + + /* + * If grouping is not available (not implemented or prohibited by + * security mode) these registers a read-as-zero/write-ignored. + * However as a precaution we restore the reset default regardless of + * the result of the test. + */ + writel_relaxed(1, dist_base + GIC_DIST_IGROUP + 0); + gic->fiq_enable = readl_relaxed(dist_base + GIC_DIST_IGROUP + 0); + writel_relaxed(0, dist_base + GIC_DIST_IGROUP + 0); + pr_debug("gic: FIQ support %s\n", + gic->fiq_enable ? "enabled" : "disabled"); + + if (!gic->fiq_enable) + return; +} +#else /* CONFIG_FIQ */ +static inline void gic_init_fiq(struct gic_chip_data *gic, + irq_hw_number_t first_irq, + unsigned int num_irqs) {} +#endif /* CONFIG_FIQ */ + void __init gic_cascade_irq(unsigned int gic_nr, unsigned int irq) { if (gic_nr >= MAX_GIC_NR) @@ -408,13 +460,28 @@ static void __init gic_dist_init(struct gic_chip_data *gic) writel_relaxed(0xa0a0a0a0, base + GIC_DIST_PRI + i * 4 / 4); /* + * Optionally set all global interrupts to be group 1. + */ + if (gic_data_fiq_enable(gic)) + for (i = 32; i < gic_irqs; i += 32) + writel_relaxed(0xffffffff, + base + GIC_DIST_IGROUP + i * 4 / 32); + + /* * Disable all interrupts. Leave the PPI and SGIs alone * as these enables are banked registers. */ for (i = 32; i < gic_irqs; i += 32) writel_relaxed(0xffffffff, base + GIC_DIST_ENABLE_CLEAR + i * 4 / 32); - writel_relaxed(1, base + GIC_DIST_CTRL); + /* + * Set EnableGrp1/EnableGrp0 (bit 1 and 0) or EnableGrp (bit 0 only, + * bit 1 ignored) + */ + if (gic_data_fiq_enable(gic)) + writel_relaxed(3, base + GIC_DIST_CTRL); + else + writel_relaxed(1, base + GIC_DIST_CTRL); } static void gic_cpu_init(struct gic_chip_data *gic) @@ -452,8 +519,20 @@ static void gic_cpu_init(struct gic_chip_data *gic) for (i = 0; i < 32; i += 4) writel_relaxed(0xa0a0a0a0, dist_base + GIC_DIST_PRI + i * 4 / 4); + /* + * Set all PPI and SGI interrupts to be group 1. + * + * If grouping is not available (not implemented or prohibited by + * security mode) these registers are read-as-zero/write-ignored. + */ + if (gic_data_fiq_enable(gic)) + writel_relaxed(0xffffffff, dist_base + GIC_DIST_IGROUP + 0); + writel_relaxed(0xf0, base + GIC_CPU_PRIMASK); - writel_relaxed(1, base + GIC_CPU_CTRL); + if (gic_data_fiq_enable(gic)) + writel_relaxed(0x1f, base + GIC_CPU_CTRL); + else + writel_relaxed(1, base + GIC_CPU_CTRL); } void gic_cpu_if_down(void) @@ -537,7 +616,10 @@ static void gic_dist_restore(unsigned int gic_nr) writel_relaxed(gic_data[gic_nr].saved_spi_enable[i], dist_base + GIC_DIST_ENABLE_SET + i * 4); - writel_relaxed(1, dist_base + GIC_DIST_CTRL); + if (gic_data_fiq_enable(&gic_data[gic_nr])) + writel_relaxed(3, dist_base + GIC_DIST_CTRL); + else + writel_relaxed(1, dist_base + GIC_DIST_CTRL); } static void gic_cpu_save(unsigned int gic_nr) @@ -594,7 +676,7 @@ static void gic_cpu_restore(unsigned int gic_nr) writel_relaxed(0xa0a0a0a0, dist_base + GIC_DIST_PRI + i * 4); writel_relaxed(0xf0, cpu_base + GIC_CPU_PRIMASK); - writel_relaxed(1, cpu_base + GIC_CPU_CTRL); + writel_relaxed(0x1f, cpu_base + GIC_CPU_CTRL); } static int gic_notifier(struct notifier_block *self, unsigned long cmd, void *v) @@ -656,6 +738,7 @@ static void gic_raise_softirq(const struct cpumask *mask, unsigned int irq) { int cpu; unsigned long flags, map = 0; + unsigned long softint; raw_spin_lock_irqsave(&irq_controller_lock, flags); @@ -670,7 +753,11 @@ static void gic_raise_softirq(const struct cpumask *mask, unsigned int irq) dmb(ishst); /* this always happens on GIC0 */ - writel_relaxed(map << 16 | irq, gic_data_dist_base(&gic_data[0]) + GIC_DIST_SOFTINT); + softint = map << 16 | irq; + if (gic_data_fiq_enable(&gic_data[0])) + softint |= 0x8000; + writel_relaxed(softint, + gic_data_dist_base(&gic_data[0]) + GIC_DIST_SOFTINT); raw_spin_unlock_irqrestore(&irq_controller_lock, flags); } @@ -1014,6 +1101,8 @@ void __init gic_init_bases(unsigned int gic_nr, int irq_start, gic->domain = irq_domain_add_legacy(node, gic_irqs, irq_base, hwirq_base, &gic_irq_domain_ops, gic); + + gic_init_fiq(gic, irq_base, gic_irqs); } else { gic->domain = irq_domain_add_linear(node, nr_routable_irqs, &gic_irq_domain_ops,