user/sven/linux.git/include/linux/irq.h, branch v5.6.2

x86/apic/msi: Plug non-maskable MSI affinity race

2020-02-01T08:31:47Z

Evan tracked down a subtle race between the update of the MSI message and the device raising an interrupt internally on PCI devices which do not support MSI masking. The update of the MSI message is non-atomic and consists of either 2 or 3 sequential 32bit wide writes to the PCI config space. - Write address low 32bits - Write address high 32bits (If supported by device) - Write data When an interrupt is migrated then both address and data might change, so the kernel attempts to mask the MSI interrupt first. But for MSI masking is optional, so there exist devices which do not provide it. That means that if the device raises an interrupt internally between the writes then a MSI message is sent built from half updated state. On x86 this can lead to spurious interrupts on the wrong interrupt vector when the affinity setting changes both address and data. As a consequence the device interrupt can be lost causing the device to become stuck or malfunctioning. Evan tried to handle that by disabling MSI accross an MSI message update. That's not feasible because disabling MSI has issues on its own: If MSI is disabled the PCI device is routing an interrupt to the legacy INTx mechanism. The INTx delivery can be disabled, but the disablement is not working on all devices. Some devices lose interrupts when both MSI and INTx delivery are disabled. Another way to solve this would be to enforce the allocation of the same vector on all CPUs in the system for this kind of screwed devices. That could be done, but it would bring back the vector space exhaustion problems which got solved a few years ago. Fortunately the high address (if supported by the device) is only relevant when X2APIC is enabled which implies interrupt remapping. In the interrupt remapping case the affinity setting is happening at the interrupt remapping unit and the PCI MSI message is programmed only once when the PCI device is initialized. That makes it possible to solve it with a two step update: 1) Target the MSI msg to the new vector on the current target CPU 2) Target the MSI msg to the new vector on the new target CPU In both cases writing the MSI message is only changing a single 32bit word which prevents the issue of inconsistency. After writing the final destination it is necessary to check whether the device issued an interrupt while the intermediate state #1 (new vector, current CPU) was in effect. This is possible because the affinity change is always happening on the current target CPU. The code runs with interrupts disabled, so the interrupt can be detected by checking the IRR of the local APIC. If the vector is pending in the IRR then the interrupt is retriggered on the new target CPU by sending an IPI for the associated vector on the target CPU. This can cause spurious interrupts on both the local and the new target CPU. 1) If the new vector is not in use on the local CPU and the device affected by the affinity change raised an interrupt during the transitional state (step #1 above) then interrupt entry code will ignore that spurious interrupt. The vector is marked so that the 'No irq handler for vector' warning is supressed once. 2) If the new vector is in use already on the local CPU then the IRR check might see an pending interrupt from the device which is using this vector. The IPI to the new target CPU will then invoke the handler of the device, which got the affinity change, even if that device did not issue an interrupt 3) If the new vector is in use already on the local CPU and the device affected by the affinity change raised an interrupt during the transitional state (step #1 above) then the handler of the device which uses that vector on the local CPU will be invoked. expose issues in device driver interrupt handlers which are not prepared to handle a spurious interrupt correctly. This not a regression, it's just exposing something which was already broken as spurious interrupts can happen for a lot of reasons and all driver handlers need to be able to deal with them. Reported-by: Evan Green Debugged-by: Evan Green Signed-off-by: Thomas Gleixner Tested-by: Evan Green Cc: stable@vger.kernel.org Link: https://lore.kernel.org/r/87imkr4s7n.fsf@nanos.tec.linutronix.de

genirq: Introduce irq_chip_get/set_parent_state calls

2019-11-16T10:20:02Z

On certain QTI chipsets some GPIOs are direct-connect interrupts to the GIC to be used as regular interrupt lines. When the GPIOs are not used for interrupt generation the interrupt line is disabled. But disabling the interrupt at GIC does not prevent the interrupt to be reported as pending at GIC_ISPEND. Later, when drivers call enable_irq() on the interrupt, an unwanted interrupt occurs. Introduce get and set methods for irqchip's parent to clear it's pending irq state. This then can be invoked by the GPIO interrupt controller on the parents in it hierarchy to clear the interrupt before enabling the interrupt. Signed-off-by: Maulik Shah Signed-off-by: Lina Iyer Signed-off-by: Marc Zyngier Reviewed-by: Stephen Boyd Link: https://lore.kernel.org/r/1573855915-9841-7-git-send-email-ilina@codeaurora.org [updated commit text and minor code fixes]

genirq: Introduce irq_chip_{request,release}_resource_parent() apis

2019-05-01T09:41:38Z

Introduce irq_chip_{request,release}_resource_parent() apis so that these can be used in hierarchical irqchips. Signed-off-by: Lokesh Vutla Signed-off-by: Marc Zyngier

genirq: Fix typo in comment of IRQD_MOVE_PCNTXT

2019-03-21T10:52:37Z

Signed-off-by: Peter Xu Signed-off-by: Thomas Gleixner Cc: Marc Zyngier Cc: Dou Liyang Cc: Julien Thierry Link: https://lkml.kernel.org/r/20190318065123.11862-1-peterx@redhat.com

Merge tag 'gpio-v5.1-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio

2019-03-08T18:09:53Z

Pull GPIO updates from Linus Walleij: "This is the bulk of GPIO changes for the v5.1 cycle: Core changes: - The big change this time around is the irqchip handling in the qualcomm pin controllers, closely coupled with the gpiochip. This rework, in a classic fall-between-the-chairs fashion has been sidestepped for too long. The Qualcomm IRQchips using the SPMI and SSBI transport mechanisms have been rewritten to use hierarchical irqchip. This creates the base from which I intend to gradually pull support for hierarchical irqchips into the gpiolib irqchip helpers to cut down on duplicate code. We have too many hacks in the kernel because people have been working around the missing hierarchical irqchip for years, and once it was there, noone understood it for a while. We are now slowly adapting to using it. This is why this pull requests include changes to MFD, SPMI, IRQchip core and some ARM Device Trees pertaining to the Qualcomm chip family. Since Qualcomm have so many chips and such large deployments it is paramount that this platform gets this right, and now it (hopefully) does. - Core support for pull-up and pull-down configuration, also from the device tree. When a simple GPIO chip supports an "off or on" pull-up or pull-down resistor, we provide a way to set this up using machine descriptors or device tree. If more elaborate control of pull up/down (such as resistance shunt setting) is required, drivers should be phased over to use pin control. We do not yet provide a userspace ABI for this pull up-down setting but I suspect the makers are going to ask for it soon enough. PCA953x is the first user of this new API. - The GPIO mockup driver has been revamped after some discussion improving the IRQ simulator in the process. The idea is to make it possible to use the mockup for both testing and virtual prototyping, e.g. when you do not yet have a GPIO expander to play with but really want to get something to develop code around before hardware is available. It's neat. The blackbox testing usecase is currently making its way into kernelci. - ACPI GPIO core preserves non direction flags when updating flags. - A new device core helper for devm_platform_ioremap_resource() is funneled through the GPIO tree with Greg's ACK. New drivers: - TQ-Systems QTMX86 GPIO controllers (using port-mapped I/O) - Gateworks PLD GPIO driver (vaccumed up from OpenWrt) - AMD G-Series PCH (Platform Controller Hub) GPIO driver. - Fintek F81804 & F81966 subvariants. - PCA953x now supports NXP PCAL6416. Driver improvements: - IRQ support on the Nintendo Wii (Hollywood) GPIO. - get_direction() support for the MVEBU driver. - Set the right output level on SAMA5D2. - Drop the unused irq trigger setting on the Spreadtrum driver. - Wakeup support for PCA953x. - A slew of cleanups in the various Intel drivers" * tag 'gpio-v5.1-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio: (110 commits) gpio: gpio-omap: fix level interrupt idling gpio: amd-fch: Set proper output level for direction_output x86: apuv2: remove unused variable gpio: pca953x: Use PCA_LATCH_INT platform/x86: fix PCENGINES_APU2 Kconfig warning gpio: pca953x: Fix dereference of irq data in shutdown gpio: amd-fch: Fix type error found by sparse gpio: amd-fch: Drop const from resource gpio: mxc: add check to return defer probe if clock tree NOT ready gpio: ftgpio: Register per-instance irqchip gpio: ixp4xx: Add DT bindings x86: pcengines apuv2 gpio/leds/keys platform driver gpio: AMD G-Series PCH gpio driver drivers: depend on HAS_IOMEM for devm_platform_ioremap_resource() gpio: tqmx86: Set proper output level for direction_output gpio: sprd: Change to use SoC compatible string gpio: sprd: Use SoC compatible string instead of wildcard string gpio: of: Handle both enable-gpio{,s} gpio: of: Restrict enable-gpio quirk to regulator-gpio gpio: davinci: use devm_platform_ioremap_resource() ...

genirq: introduce irq_chip_mask_ack_parent()

2019-02-13T08:23:05Z

The hierarchical irqchip never before ran into a situation where the parent is not "simple", i.e. does not implement .irq_ack() and .irq_mask() like most, but the qcom-pm8xxx.c happens to implement only .irq_mask_ack(). Since we want to make ssbi-gpio a hierarchical child of this irqchip, it must *also* only implement .irq_mask_ack() and call down to the parent, and for this we of course need irq_chip_mask_ack_parent(). Cc: Marc Zyngier Cc: Thomas Gleixner Acked-by: Marc Zyngier Signed-off-by: Brian Masney Signed-off-by: Linus Walleij

genirq: Provide NMI handlers

2019-02-05T14:37:01Z

Provide flow handlers that are NMI safe for interrupts and percpu_devid interrupts. Signed-off-by: Julien Thierry Acked-by: Marc Zyngier Cc: Thomas Gleixner Cc: Marc Zyngier Cc: Peter Zijlstra Signed-off-by: Marc Zyngier

genirq: Provide basic NMI management for interrupt lines

2019-02-05T14:36:57Z

Add functionality to allocate interrupt lines that will deliver IRQs as Non-Maskable Interrupts. These allocations are only successful if the irqchip provides the necessary support and allows NMI delivery for the interrupt line. Interrupt lines allocated for NMI delivery must be enabled/disabled through enable_nmi/disable_nmi_nosync to keep their state consistent. To treat a PERCPU IRQ as NMI, the interrupt must not be shared nor threaded, the irqchip directly managing the IRQ must be the root irqchip and the irqchip cannot be behind a slow bus. Signed-off-by: Julien Thierry Reviewed-by: Marc Zyngier Cc: Thomas Gleixner Cc: Peter Zijlstra Cc: Ingo Molnar Cc: Marc Zyngier Signed-off-by: Marc Zyngier

genirq/core: Introduce struct irq_affinity_desc

2018-12-19T10:32:08Z

The interrupt affinity management uses straight cpumask pointers to convey the automatically assigned affinity masks for managed interrupts. The core interrupt descriptor allocation also decides based on the pointer being non NULL whether an interrupt is managed or not. Devices which use managed interrupts usually have two classes of interrupts: - Interrupts for multiple device queues - Interrupts for general device management Currently both classes are treated the same way, i.e. as managed interrupts. The general interrupts get the default affinity mask assigned while the device queue interrupts are spread out over the possible CPUs. Treating the general interrupts as managed is both a limitation and under certain circumstances a bug. Assume the following situation: default_irq_affinity = 4..7 So if CPUs 4-7 are offlined, then the core code will shut down the device management interrupts because the last CPU in their affinity mask went offline. It's also a limitation because it's desired to allow manual placement of the general device interrupts for various reasons. If they are marked managed then the interrupt affinity setting from both user and kernel space is disabled. To remedy that situation it's required to convey more information than the cpumasks through various interfaces related to interrupt descriptor allocation. Instead of adding yet another argument, create a new data structure 'irq_affinity_desc' which for now just contains the cpumask. This struct can be expanded to convey auxilliary information in the next step. No functional change, just preparatory work. [ tglx: Simplified logic and clarified changelog ] Suggested-by: Thomas Gleixner Suggested-by: Bjorn Helgaas Signed-off-by: Dou Liyang Signed-off-by: Thomas Gleixner Cc: linux-pci@vger.kernel.org Cc: kashyap.desai@broadcom.com Cc: shivasharan.srikanteshwara@broadcom.com Cc: sumit.saxena@broadcom.com Cc: ming.lei@redhat.com Cc: hch@lst.de Cc: douliyang1@huawei.com Link: https://lkml.kernel.org/r/20181204155122.6327-2-douliyangs@gmail.com

irq/matrix: Spread managed interrupts on allocation

2018-09-18T16:27:24Z

Linux spreads out the non managed interrupt across the possible target CPUs to avoid vector space exhaustion. Managed interrupts are treated differently, as for them the vectors are reserved (with guarantee) when the interrupt descriptors are initialized. When the interrupt is requested a real vector is assigned. The assignment logic uses the first CPU in the affinity mask for assignment. If the interrupt has more than one CPU in the affinity mask, which happens when a multi queue device has less queues than CPUs, then doing the same search as for non managed interrupts makes sense as it puts the interrupt on the least interrupt plagued CPU. For single CPU affine vectors that's obviously a NOOP. Restructre the matrix allocation code so it does the 'best CPU' search, add the sanity check for an empty affinity mask and adapt the call site in the x86 vector management code. [ tglx: Added the empty mask check to the core and improved change log ] Signed-off-by: Dou Liyang Signed-off-by: Thomas Gleixner Cc: hpa@zytor.com Link: https://lkml.kernel.org/r/20180908175838.14450-2-dou_liyang@163.com