Received: by 2002:a25:ab43:0:0:0:0:0 with SMTP id u61csp7853330ybi; Thu, 6 Jun 2019 02:35:00 -0700 (PDT) X-Google-Smtp-Source: APXvYqwCO4Zvush75bEhhe+/RcsWwRCednCcs3/f19xnnWXyykK37ztpo+OdHFdYFm1CfKqqCSDK X-Received: by 2002:a17:90a:bb94:: with SMTP id v20mr23613255pjr.88.1559813700068; Thu, 06 Jun 2019 02:35:00 -0700 (PDT) ARC-Seal: i=1; a=rsa-sha256; t=1559813700; cv=none; d=google.com; s=arc-20160816; b=MC1ZRHeJcmHKuweiPUElHbbxmjKTteX4oFjmzlfh0QjE3idodm7NCLicRnj/zpuBX4 ym7kvtPC6rddNOwzL8UgrOQhJz5kmZUhnSOFrlpJAfNwcBB1+SWk7WXsg++cWjqecioP tTn8dvxdTvxp8/xaV+AkEVOp1acE5sJQV8UzEm2tDDXTIsH62VDZgJdlZ8qetH1+jLyL f6tWu3Oyq5lX9bbzYs4+36kkoonkPXrEnbfrKzsF9B8pg3wtNytaLfo+yHubnW8WLEW+ pZUXgnaiN3dciW9PYyrb5TVey88CwudB8dOtydZHSTYHepjQuP/dOWMBBDIZBSyh1pRt nEoQ== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20160816; h=list-id:precedence:sender:references:in-reply-to:message-id:date :subject:cc:to:from; bh=IftE7SGycU0i3WYkXpVw075Tro4JDxipUBWS/Dd0VtM=; b=VEeGJrh/jJd2zzzW1sqJnWqJzcLhb/t76kKFw/p1PnpQ08aPO6QwTpYO8aRxPXVQVH xAft/CeFo52wVvVkQVLn/86rstzSGapDLQmfHdRS++ybnIbODT0ips8fmnpMdiosZnOE kEpKmLCR5Pkq10w90SjjHnYJNtkQY+0Dzyn5sd7FhgXKWeLGyt8UCoaMz8ZWGe9k1esB bwOCdFuswr9h41cC5bq7JiXoEr3WSln9vqsJX4o0IJmdXxAzrZHNq4i9VK0/BrLxEKrP qvgRqDMdsABl5gB5n1iRkj8k08Ep1uBQDc7fc/ArVMSWq06Cyq4Ny8N3wXKVJDNV2xKl kziw== ARC-Authentication-Results: i=1; mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org Return-Path: Received: from vger.kernel.org (vger.kernel.org. [209.132.180.67]) by mx.google.com with ESMTP id o17si1575628pge.519.2019.06.06.02.34.42; Thu, 06 Jun 2019 02:35:00 -0700 (PDT) Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67; Authentication-Results: mx.google.com; spf=pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728002AbfFFJcl (ORCPT + 99 others); Thu, 6 Jun 2019 05:32:41 -0400 Received: from usa-sjc-mx-foss1.foss.arm.com ([217.140.101.70]:43630 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727893AbfFFJcO (ORCPT ); Thu, 6 Jun 2019 05:32:14 -0400 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.72.51.249]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id 2B862165C; Thu, 6 Jun 2019 02:32:14 -0700 (PDT) Received: from e112298-lin.cambridge.arm.com (usa-sjc-imap-foss1.foss.arm.com [10.72.51.249]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPA id E80243F71D; Thu, 6 Jun 2019 02:32:11 -0700 (PDT) From: Julien Thierry To: linux-arm-kernel@lists.infradead.org Cc: linux-kernel@vger.kernel.org, rostedt@goodmis.org, marc.zyngier@arm.com, yuzenghui@huawei.com, wanghaibin.wang@huawei.com, james.morse@arm.com, will.deacon@arm.com, catalin.marinas@arm.com, mark.rutland@arm.com, liwei391@huawei.com, Julien Thierry , Thomas Gleixner , Jason Cooper Subject: [PATCH v3 4/8] arm64: Fix interrupt tracing in the presence of NMIs Date: Thu, 6 Jun 2019 10:31:53 +0100 Message-Id: <1559813517-41540-5-git-send-email-julien.thierry@arm.com> X-Mailer: git-send-email 1.9.1 In-Reply-To: <1559813517-41540-1-git-send-email-julien.thierry@arm.com> References: <1559813517-41540-1-git-send-email-julien.thierry@arm.com> Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org In the presence of any form of instrumentation, nmi_enter() should be done before calling any traceable code and any instrumentation code. Currently, nmi_enter() is done in handle_domain_nmi(), which is much too late as instrumentation code might get called before. Move the nmi_enter/exit() calls to the arch IRQ vector handler. On arm64, it is not possible to know if the IRQ vector handler was called because of an NMI before acknowledging the interrupt. However, It is possible to know whether normal interrupts could be taken in the interrupted context (i.e. if taking an NMI in that context could introduce a potential race condition). When interrupting a context with IRQs disabled, call nmi_enter() as soon as possible. In contexts with IRQs enabled, defer this to the interrupt controller, which is in a better position to know if an interrupt taken is an NMI. Fixes: bc3c03ccb ("arm64: Enable the support of pseudo-NMIs") Signed-off-by: Julien Thierry Cc: Catalin Marinas Cc: Will Deacon Cc: Thomas Gleixner Cc: Jason Cooper Cc: Marc Zyngier Cc: Mark Rutland --- arch/arm64/kernel/entry.S | 44 +++++++++++++++++++++++++++++++++----------- arch/arm64/kernel/irq.c | 17 +++++++++++++++++ drivers/irqchip/irq-gic-v3.c | 6 ++++++ kernel/irq/irqdesc.c | 8 ++++++-- 4 files changed, 62 insertions(+), 13 deletions(-) diff --git a/arch/arm64/kernel/entry.S b/arch/arm64/kernel/entry.S index 89ab6bd..46358a3 100644 --- a/arch/arm64/kernel/entry.S +++ b/arch/arm64/kernel/entry.S @@ -435,6 +435,20 @@ tsk .req x28 // current thread_info irq_stack_exit .endm +#ifdef CONFIG_ARM64_PSEUDO_NMI + /* + * Set res to 0 if irqs were masked in interrupted context. + * Otherwise set res to non-0 value. + */ + .macro test_irqs_unmasked res:req, pmr:req +alternative_if ARM64_HAS_IRQ_PRIO_MASKING + sub \res, \pmr, #GIC_PRIO_IRQON +alternative_else + mov \res, xzr +alternative_endif + .endm +#endif + .text /* @@ -631,19 +645,19 @@ ENDPROC(el1_sync) el1_irq: kernel_entry 1 enable_da_f -#ifdef CONFIG_TRACE_IRQFLAGS + #ifdef CONFIG_ARM64_PSEUDO_NMI alternative_if ARM64_HAS_IRQ_PRIO_MASKING ldr x20, [sp, #S_PMR_SAVE] -alternative_else - mov x20, #GIC_PRIO_IRQON -alternative_endif - cmp x20, #GIC_PRIO_IRQOFF - /* Irqs were disabled, don't trace */ - b.ls 1f +alternative_else_nop_endif + test_irqs_unmasked res=x0, pmr=x20 + cbz x0, 1f + bl asm_nmi_enter +1: #endif + +#ifdef CONFIG_TRACE_IRQFLAGS bl trace_hardirqs_off -1: #endif irq_handler @@ -662,14 +676,22 @@ alternative_else_nop_endif bl preempt_schedule_irq // irq en/disable is done inside 1: #endif -#ifdef CONFIG_TRACE_IRQFLAGS + #ifdef CONFIG_ARM64_PSEUDO_NMI /* * if IRQs were disabled when we received the interrupt, we have an NMI * and we are not re-enabling interrupt upon eret. Skip tracing. */ - cmp x20, #GIC_PRIO_IRQOFF - b.ls 1f + test_irqs_unmasked res=x0, pmr=x20 + cbz x0, 1f + bl asm_nmi_exit +1: +#endif + +#ifdef CONFIG_TRACE_IRQFLAGS +#ifdef CONFIG_ARM64_PSEUDO_NMI + test_irqs_unmasked res=x0, pmr=x20 + cbnz x0, 1f #endif bl trace_hardirqs_on 1: diff --git a/arch/arm64/kernel/irq.c b/arch/arm64/kernel/irq.c index 92fa817..fdd9cb2 100644 --- a/arch/arm64/kernel/irq.c +++ b/arch/arm64/kernel/irq.c @@ -27,8 +27,10 @@ #include #include #include +#include #include #include +#include #include unsigned long irq_err_count; @@ -76,3 +78,18 @@ void __init init_IRQ(void) if (!handle_arch_irq) panic("No interrupt controller found."); } + +/* + * Stubs to make nmi_enter/exit() code callable from ASM + */ +asmlinkage void notrace asm_nmi_enter(void) +{ + nmi_enter(); +} +NOKPROBE_SYMBOL(asm_nmi_enter); + +asmlinkage void notrace asm_nmi_exit(void) +{ + nmi_exit(); +} +NOKPROBE_SYMBOL(asm_nmi_exit); diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c index f44cd89..0bf0fc4 100644 --- a/drivers/irqchip/irq-gic-v3.c +++ b/drivers/irqchip/irq-gic-v3.c @@ -495,7 +495,13 @@ static asmlinkage void __exception_irq_entry gic_handle_irq(struct pt_regs *regs if (gic_supports_nmi() && unlikely(gic_read_rpr() == GICD_INT_NMI_PRI)) { + if (interrupts_enabled(regs)) + nmi_enter(); + gic_handle_nmi(irqnr, regs); + + if (interrupts_enabled(regs)) + nmi_exit(); return; } diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c index c52b737..a92b335 100644 --- a/kernel/irq/irqdesc.c +++ b/kernel/irq/irqdesc.c @@ -680,6 +680,8 @@ int __handle_domain_irq(struct irq_domain *domain, unsigned int hwirq, * @hwirq: The HW irq number to convert to a logical one * @regs: Register file coming from the low-level handling code * + * This function must be called from an NMI context. + * * Returns: 0 on success, or -EINVAL if conversion has failed */ int handle_domain_nmi(struct irq_domain *domain, unsigned int hwirq, @@ -689,7 +691,10 @@ int handle_domain_nmi(struct irq_domain *domain, unsigned int hwirq, unsigned int irq; int ret = 0; - nmi_enter(); + /* + * NMI context needs to be setup earlier in order to deal with tracing. + */ + WARN_ON(!in_nmi()); irq = irq_find_mapping(domain, hwirq); @@ -702,7 +707,6 @@ int handle_domain_nmi(struct irq_domain *domain, unsigned int hwirq, else ret = -EINVAL; - nmi_exit(); set_irq_regs(old_regs); return ret; } -- 1.9.1