Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753065AbaLEIsh (ORCPT ); Fri, 5 Dec 2014 03:48:37 -0500 Received: from www.linutronix.de ([62.245.132.108]:34019 "EHLO Galois.linutronix.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752549AbaLEIsf (ORCPT ); Fri, 5 Dec 2014 03:48:35 -0500 Message-Id: <20141205084147.313026156@linutronix.de> User-Agent: quilt/0.63-1 Date: Fri, 05 Dec 2014 08:48:32 -0000 From: Thomas Gleixner To: LKML Cc: x86@kernel.org, Jiang Liu , Joerg Roedel , Borislav Petkov Subject: [patch 3/5] iommu/vt-d: Move iommu preparatory allocations to irq_remap_ops.prepare References: <20141204234737.728961990@linutronix.de> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-15 Content-Disposition: inline; filename=iommu-intel-remap-move-domain-setup.patch X-Linutronix-Spam-Score: -1.0 X-Linutronix-Spam-Level: - X-Linutronix-Spam-Status: No , -1.0 points, 5.0 required, ALL_TRUSTED=-1,SHORTCIRCUIT=-0.0001,URIBL_BLOCKED=0.001 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org The whole iommu setup for irq remapping is a convoluted mess. The iommu detect function gets called from mem_init() and the prepare callback gets called from enable_IR_x2apic() for unknown reasons. Of course AMD and Intel setup differs in nonsensical ways. Intels prepare callback is explicit while AMDs prepare callback is implicit in setup_irq_remapping_ops() just to be called in the prepare call again. Because all of this gets called from enable_IR_x2apic() and the dmar prepare function merily parses the ACPI tables, but does not allocate memory we end up with memory allocation from irq disabled context later on. AMDs iommu code at least allocates the required memory from the prepare function. That has issues as well, but thats not scope of this patch. The goal of this change is to distangle the allocation from the actual enablement. There is no point to allocate memory from irq disabled regions with GFP_ATOMIC just because it does not matter at that point in the boot stage. It matters with physical hotplug later on. There is another issue with the current setup. Due to the conversion to stacked irqdomains we end up with a call into the irqdomain allocation code from irq disabled context, but that code does GFP_KERNEL allocations rightfully as there is no reason to do preperatory allocations with GFP_ATOMIC. That change caused the allocator code to complain about GFP_KERNEL allocations invoked in atomic context. Boris provided a temporary hackaround which changed the GFP flags if irq_domain_add() got called from atomic context. Not pretty and we really dont want to get this into a mainline release for obvious reasons. Move the ACPI table parsing and the resulting memory allocations from the enable to the prepare function. That allows to get rid of the horrible hackaround in irq_domain_add() later. Signed-off-by: Thomas Gleixner --- drivers/iommu/intel_irq_remapping.c | 64 ++++++++++++++++++++++++------------ 1 file changed, 44 insertions(+), 20 deletions(-) Index: tip/drivers/iommu/intel_irq_remapping.c =================================================================== --- tip.orig/drivers/iommu/intel_irq_remapping.c +++ tip/drivers/iommu/intel_irq_remapping.c @@ -539,22 +539,57 @@ static int __init intel_irq_remapping_su return 1; } -static int __init intel_enable_irq_remapping(void) +static void __init intel_cleanup_irq_remapping(void) +{ + struct dmar_drhd_unit *drhd; + struct intel_iommu *iommu; + + for_each_iommu(iommu, drhd) { + if (ecap_ir_support(iommu->ecap)) { + iommu_disable_irq_remapping(iommu); + intel_teardown_irq_remapping(iommu); + } + } + + if (x2apic_supported()) + pr_warn("Failed to enable irq remapping. You are vulnerable to irq-injection attacks.\n"); +} + +static int __init intel_prepare_irq_remapping(void) { struct dmar_drhd_unit *drhd; struct intel_iommu *iommu; - bool x2apic_present; - int setup = 0; - int eim = 0; - x2apic_present = x2apic_supported(); + if (dmar_table_init() < 0) + return -1; if (parse_ioapics_under_ir() != 1) { - printk(KERN_INFO "Not enable interrupt remapping\n"); + printk(KERN_INFO "Not enabling interrupt remapping\n"); goto error; } - if (x2apic_present) { + for_each_iommu(iommu, drhd) { + if (!ecap_ir_support(iommu->ecap)) + continue; + + /* Do the allocations early */ + if (intel_setup_irq_remapping(iommu)) + goto error; + } + return 0; +error: + intel_cleanup_irq_remapping(); + return -1; +} + +static int __init intel_enable_irq_remapping(void) +{ + struct dmar_drhd_unit *drhd; + struct intel_iommu *iommu; + int setup = 0; + int eim = 0; + + if (x2apic_supported()) { pr_info("Queued invalidation will be enabled to support x2apic and Intr-remapping.\n"); eim = !dmar_x2apic_optout(); @@ -622,9 +657,6 @@ static int __init intel_enable_irq_remap if (!ecap_ir_support(iommu->ecap)) continue; - if (intel_setup_irq_remapping(iommu)) - goto error; - iommu_set_irq_remapping(iommu, eim); setup = 1; } @@ -639,15 +671,7 @@ static int __init intel_enable_irq_remap return eim ? IRQ_REMAP_X2APIC_MODE : IRQ_REMAP_XAPIC_MODE; error: - for_each_iommu(iommu, drhd) - if (ecap_ir_support(iommu->ecap)) { - iommu_disable_irq_remapping(iommu); - intel_teardown_irq_remapping(iommu); - } - - if (x2apic_present) - pr_warn("Failed to enable irq remapping. You are vulnerable to irq-injection attacks.\n"); - + intel_cleanup_irq_remapping(); return -1; } @@ -947,7 +971,7 @@ static struct irq_domain *intel_get_irq_ struct irq_remap_ops intel_irq_remap_ops = { .supported = intel_irq_remapping_supported, - .prepare = dmar_table_init, + .prepare = intel_prepare_irq_remapping, .enable = intel_enable_irq_remapping, .disable = disable_irq_remapping, .reenable = reenable_irq_remapping, -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/