[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Antoine Tenart]

Hi all,

This series introduce the Alpine MSIX driver, and enables it in both
the Alpine v1 and Alpine v2 device trees.

This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
https://lkml.org/lkml/2016/2/10/83

You can find the series at:
https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix

Antoine

Since v1:
  - Added an interrupt-parent property in the documentation example.
  - Updated to use bitmap_*() instead of *_bit().
  - Removed the static irq_set_affinity to use irq_chip_set_affinity_parent().
  - Updated the address field to use phys_addr_t.
  - Added a comment on why we're setting bit 16 in the address.
  - Patched the gic_set_affinity() function in irqchip/gic-v3.

Antoine Tenart (7):
  irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity
  irqchip: add the Alpine MSIX interrupt controller
  Documentation: bindings: document the Alpine MSIX driver
  arm64: dts: alpine: add the MSIX node in the Alpine v2 dtsi
  ARM: dts: alpine: add the MSIX node
  arm64: alpine: select the Alpine MSI controller driver
  arm: alpine: select the Alpine MSI controller driver

 .../interrupt-controller/al,alpine-msix.txt        |  24 ++
 arch/arm/boot/dts/alpine.dtsi                      |  10 +
 arch/arm/mach-alpine/Kconfig                       |   1 +
 arch/arm64/Kconfig.platforms                       |   1 +
 arch/arm64/boot/dts/al/alpine-v2.dtsi              |  10 +
 drivers/irqchip/Kconfig                            |   6 +
 drivers/irqchip/Makefile                           |   1 +
 drivers/irqchip/irq-alpine-msi.c                   | 293 +++++++++++++++++++++
 drivers/irqchip/irq-gic-v3.c                       |   2 +-
 9 files changed, 347 insertions(+), 1 deletion(-)
 create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
 create mode 100644 drivers/irqchip/irq-alpine-msi.c

-- 
2.7.0

[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Antoine Tenart]

Hi all,

This series introduce the Alpine MSIX driver, and enables it in both
the Alpine v1 and Alpine v2 device trees.

This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
https://lkml.org/lkml/2016/2/10/83

You can find the series at:
https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix

Antoine

Since v1:
  - Added an interrupt-parent property in the documentation example.
  - Updated to use bitmap_*() instead of *_bit().
  - Removed the static irq_set_affinity to use irq_chip_set_affinity_parent().
  - Updated the address field to use phys_addr_t.
  - Added a comment on why we're setting bit 16 in the address.
  - Patched the gic_set_affinity() function in irqchip/gic-v3.

Antoine Tenart (7):
  irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity
  irqchip: add the Alpine MSIX interrupt controller
  Documentation: bindings: document the Alpine MSIX driver
  arm64: dts: alpine: add the MSIX node in the Alpine v2 dtsi
  ARM: dts: alpine: add the MSIX node
  arm64: alpine: select the Alpine MSI controller driver
  arm: alpine: select the Alpine MSI controller driver

 .../interrupt-controller/al,alpine-msix.txt        |  24 ++
 arch/arm/boot/dts/alpine.dtsi                      |  10 +
 arch/arm/mach-alpine/Kconfig                       |   1 +
 arch/arm64/Kconfig.platforms                       |   1 +
 arch/arm64/boot/dts/al/alpine-v2.dtsi              |  10 +
 drivers/irqchip/Kconfig                            |   6 +
 drivers/irqchip/Makefile                           |   1 +
 drivers/irqchip/irq-alpine-msi.c                   | 293 +++++++++++++++++++++
 drivers/irqchip/irq-gic-v3.c                       |   2 +-
 9 files changed, 347 insertions(+), 1 deletion(-)
 create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
 create mode 100644 drivers/irqchip/irq-alpine-msi.c

-- 
2.7.0

[PATCH v2 1/7] irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity

[Antoine Tenart]

Always return IRQ_SET_MASK_OK_DONE instead of IRQ_SET_MASK_OK when the
affinity has been updated. When using stacked irqchips, returning
IRQ_SET_MASK_OK_DONE means skipping all descendant irqchips.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
---
 drivers/irqchip/irq-gic-v3.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index d7be6ddc34f6..ca120a99bd5a 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -634,7 +634,7 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
 	else
 		gic_dist_wait_for_rwp();
 
-	return IRQ_SET_MASK_OK;
+	return IRQ_SET_MASK_OK_DONE;
 }
 #else
 #define gic_set_affinity	NULL
-- 
2.7.0

[PATCH v2 1/7] irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity

[Antoine Tenart]

Always return IRQ_SET_MASK_OK_DONE instead of IRQ_SET_MASK_OK when the
affinity has been updated. When using stacked irqchips, returning
IRQ_SET_MASK_OK_DONE means skipping all descendant irqchips.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
---
 drivers/irqchip/irq-gic-v3.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index d7be6ddc34f6..ca120a99bd5a 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -634,7 +634,7 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *mask_val,
 	else
 		gic_dist_wait_for_rwp();
 
-	return IRQ_SET_MASK_OK;
+	return IRQ_SET_MASK_OK_DONE;
 }
 #else
 #define gic_set_affinity	NULL
-- 
2.7.0

[PATCH v2 2/7] irqchip: add the Alpine MSIX interrupt controller

[Antoine Tenart]

This patch adds the Alpine MSIX interrupt controller driver.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 drivers/irqchip/Kconfig          |   6 +
 drivers/irqchip/Makefile         |   1 +
 drivers/irqchip/irq-alpine-msi.c | 293 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 300 insertions(+)
 create mode 100644 drivers/irqchip/irq-alpine-msi.c

diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig
index 715923d5236c..f20e5b28eb5f 100644
--- a/drivers/irqchip/Kconfig
+++ b/drivers/irqchip/Kconfig
@@ -60,6 +60,12 @@ config ARM_VIC_NR
 	  The maximum number of VICs available in the system, for
 	  power management.
 
+config ALPINE_MSI
+	bool
+	depends on PCI && PCI_MSI
+	select GENERIC_IRQ_CHIP
+	select PCI_MSI_IRQ_DOMAIN
+
 config ATMEL_AIC_IRQ
 	bool
 	select GENERIC_IRQ_CHIP
diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
index 18caacb60d58..57f68e0eea74 100644
--- a/drivers/irqchip/Makefile
+++ b/drivers/irqchip/Makefile
@@ -1,5 +1,6 @@
 obj-$(CONFIG_IRQCHIP)			+= irqchip.o
 
+obj-$(CONFIG_ALPINE_MSI)		+= irq-alpine-msi.o
 obj-$(CONFIG_ARCH_BCM2835)		+= irq-bcm2835.o
 obj-$(CONFIG_ARCH_BCM2835)		+= irq-bcm2836.o
 obj-$(CONFIG_ARCH_EXYNOS)		+= exynos-combiner.o
diff --git a/drivers/irqchip/irq-alpine-msi.c b/drivers/irqchip/irq-alpine-msi.c
new file mode 100644
index 000000000000..f8712722a606
--- /dev/null
+++ b/drivers/irqchip/irq-alpine-msi.c
@@ -0,0 +1,293 @@
+/*
+ * Annapurna Labs MSIX support services
+ *
+ * Copyright (C) 2016, Amazon.com, Inc. or its affiliates. All Rights Reserved.
+ *
+ * Antoine Tenart <antoine.tenart@free-electrons.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/irqchip.h>
+#include <linux/irqchip/arm-gic.h>
+#include <linux/msi.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_pci.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+
+#include <asm/irq.h>
+#include <asm-generic/msi.h>
+
+/* MSIX message address format: local GIC target */
+#define ALPINE_MSIX_SPI_TARGET_CLUSTER0		BIT(16)
+
+struct alpine_msix_data {
+	spinlock_t msi_map_lock;
+	phys_addr_t addr;
+	u32 spi_first;		/* The SGI number that MSIs start */
+	u32 num_spis;		/* The number of SGIs for MSIs */
+	unsigned long *msi_map;
+};
+
+static void alpine_msix_mask_msi_irq(struct irq_data *d)
+{
+	pci_msi_mask_irq(d);
+	irq_chip_mask_parent(d);
+}
+
+static void alpine_msix_unmask_msi_irq(struct irq_data *d)
+{
+	pci_msi_unmask_irq(d);
+	irq_chip_unmask_parent(d);
+}
+
+static struct irq_chip alpine_msix_irq_chip = {
+	.name			= "MSIx",
+	.irq_mask		= alpine_msix_mask_msi_irq,
+	.irq_unmask		= alpine_msix_unmask_msi_irq,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+};
+
+static int alpine_msix_allocate_sgi(struct alpine_msix_data *priv, int num_req)
+{
+	int first;
+
+	spin_lock(&priv->msi_map_lock);
+
+	first = bitmap_find_next_zero_area(priv->msi_map, priv->num_spis, 0,
+					   num_req, 0);
+	if (first >= priv->num_spis) {
+		spin_unlock(&priv->msi_map_lock);
+		return -ENOSPC;
+	}
+
+	bitmap_set(priv->msi_map, first, num_req);
+
+	spin_unlock(&priv->msi_map_lock);
+
+	return priv->spi_first + first;
+}
+
+static void alpine_msix_free_sgi(struct alpine_msix_data *priv, unsigned sgi,
+				 int num_req)
+{
+	int first = sgi - priv->spi_first;
+
+	spin_lock(&priv->msi_map_lock);
+
+	bitmap_clear(priv->msi_map, first, num_req);
+
+	spin_unlock(&priv->msi_map_lock);
+}
+
+static void alpine_msix_compose_msi_msg(struct irq_data *data,
+					struct msi_msg *msg)
+{
+	struct alpine_msix_data *priv = irq_data_get_irq_chip_data(data);
+	phys_addr_t msg_addr = priv->addr;
+
+	msg_addr |= (data->hwirq << 3);
+
+	msg->address_hi = upper_32_bits(msg_addr);
+	msg->address_lo = lower_32_bits(msg_addr);
+	msg->data = 0;
+}
+
+static struct msi_domain_info alpine_msix_domain_info = {
+	.flags	= MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
+		  MSI_FLAG_PCI_MSIX,
+	.chip	= &alpine_msix_irq_chip,
+};
+
+static struct irq_chip middle_irq_chip = {
+	.name			= "alpine_msix_middle",
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+	.irq_compose_msi_msg	= alpine_msix_compose_msi_msg,
+};
+
+static int alpine_msix_gic_domain_alloc(struct irq_domain *domain,
+					unsigned int virq, int sgi)
+{
+	struct irq_fwspec fwspec;
+	struct irq_data *d;
+	int ret;
+
+	if (!is_of_node(domain->parent->fwnode))
+		return -EINVAL;
+
+	fwspec.fwnode = domain->parent->fwnode;
+	fwspec.param_count = 3;
+	fwspec.param[0] = 0;
+	fwspec.param[1] = sgi;
+	fwspec.param[2] = IRQ_TYPE_EDGE_RISING;
+
+	ret = irq_domain_alloc_irqs_parent(domain, virq, 1, &fwspec);
+	if (ret)
+		return ret;
+
+	d = irq_domain_get_irq_data(domain->parent, virq);
+	d->chip->irq_set_type(d, IRQ_TYPE_EDGE_RISING);
+
+	return 0;
+}
+
+static int alpine_msix_middle_domain_alloc(struct irq_domain *domain,
+					   unsigned int virq,
+					   unsigned int nr_irqs, void *args)
+{
+	struct alpine_msix_data *priv = domain->host_data;
+	int sgi, err, i;
+
+	sgi = alpine_msix_allocate_sgi(priv, nr_irqs);
+	if (sgi < 0)
+		return sgi;
+
+	for (i = 0; i < nr_irqs; i++) {
+		err = alpine_msix_gic_domain_alloc(domain, virq + i, sgi + i);
+		if (err)
+			goto err_sgi;
+
+		irq_domain_set_hwirq_and_chip(domain, virq + i, sgi + i,
+					      &middle_irq_chip, priv);
+	}
+
+	return 0;
+
+err_sgi:
+	while (--i >= 0)
+		irq_domain_free_irqs_parent(domain, virq, i);
+	alpine_msix_free_sgi(priv, sgi, nr_irqs);
+	return err;
+}
+
+static void alpine_msix_middle_domain_free(struct irq_domain *domain,
+					   unsigned int virq,
+					   unsigned int nr_irqs)
+{
+	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
+	struct alpine_msix_data *priv = irq_data_get_irq_chip_data(d);
+
+	irq_domain_free_irqs_parent(domain, virq, nr_irqs);
+	alpine_msix_free_sgi(priv, d->hwirq, nr_irqs);
+}
+
+static const struct irq_domain_ops alpine_msix_middle_domain_ops = {
+	.alloc	= alpine_msix_middle_domain_alloc,
+	.free	= alpine_msix_middle_domain_free,
+};
+
+static int alpine_msix_init_domains(struct alpine_msix_data *priv,
+				    struct device_node *node)
+{
+	struct irq_domain *middle_domain, *msi_domain, *gic_domain;
+	struct device_node *gic_node;
+
+	gic_node = of_irq_find_parent(node);
+	if (!gic_node) {
+		pr_err("Failed to find the GIC node\n");
+		return -ENODEV;
+	}
+
+	gic_domain = irq_find_host(gic_node);
+	if (!gic_domain) {
+		pr_err("Failed to find the GIC domain\n");
+		return -ENXIO;
+	}
+
+	middle_domain = irq_domain_add_tree(NULL,
+					    &alpine_msix_middle_domain_ops,
+					    priv);
+	if (!middle_domain) {
+		pr_err("Failed to create the MSIX middle domain\n");
+		return -ENOMEM;
+	}
+
+	middle_domain->parent = gic_domain;
+
+	msi_domain = pci_msi_create_irq_domain(of_node_to_fwnode(node),
+					       &alpine_msix_domain_info,
+					       middle_domain);
+	if (!msi_domain) {
+		pr_err("Failed to create MSI domain\n");
+		irq_domain_remove(msi_domain);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int alpine_msix_init(struct device_node *node,
+			    struct device_node *parent)
+{
+	struct alpine_msix_data *priv;
+	struct resource res;
+	int ret;
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	spin_lock_init(&priv->msi_map_lock);
+
+	ret = of_address_to_resource(node, 0, &res);
+	if (ret) {
+		pr_err("Failed to allocate resource\n");
+		goto err_priv;
+	}
+
+	/*
+	 * The 20 least significant bits of addr provide direct information
+	 * regarding the interrupt destination.
+	 *
+	 * To select the primary GIC as the target GIC, bits [18:17] must be set
+	 * to 0x0. In this case, bit 16 (SPI_TARGET_CLUSTER0) must be set.
+	 */
+	priv->addr = res.start & GENMASK_ULL(63,20);
+	priv->addr |= ALPINE_MSIX_SPI_TARGET_CLUSTER0;
+
+	if (of_property_read_u32(node, "al,msi-base-spi", &priv->spi_first)) {
+		pr_err("Unable to parse MSI base\n");
+		ret = -EINVAL;
+		goto err_priv;
+	}
+
+	if (of_property_read_u32(node, "al,msi-num-spis", &priv->num_spis)) {
+		pr_err("Unable to parse MSI numbers\n");
+		ret = -EINVAL;
+		goto err_priv;
+	}
+
+	priv->msi_map = kzalloc(sizeof(*priv->msi_map) * BITS_TO_LONGS(priv->num_spis),
+				GFP_KERNEL);
+	if (!priv->msi_map) {
+		ret = -ENOMEM;
+		goto err_priv;
+	}
+
+	pr_debug("Registering %d msixs, starting at %d\n",
+		 priv->num_spis, priv->spi_first);
+
+	ret = alpine_msix_init_domains(priv, node);
+	if (ret)
+		goto err_map;
+
+	return 0;
+
+err_map:
+	kfree(priv->msi_map);
+err_priv:
+	kfree(priv);
+	return ret;
+}
+IRQCHIP_DECLARE(alpine_msix, "al,alpine-msix", alpine_msix_init);
-- 
2.7.0

[PATCH v2 2/7] irqchip: add the Alpine MSIX interrupt controller

[Antoine Tenart]

This patch adds the Alpine MSIX interrupt controller driver.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 drivers/irqchip/Kconfig          |   6 +
 drivers/irqchip/Makefile         |   1 +
 drivers/irqchip/irq-alpine-msi.c | 293 +++++++++++++++++++++++++++++++++++++++
 3 files changed, 300 insertions(+)
 create mode 100644 drivers/irqchip/irq-alpine-msi.c

diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig
index 715923d5236c..f20e5b28eb5f 100644
--- a/drivers/irqchip/Kconfig
+++ b/drivers/irqchip/Kconfig
@@ -60,6 +60,12 @@ config ARM_VIC_NR
 	  The maximum number of VICs available in the system, for
 	  power management.
 
+config ALPINE_MSI
+	bool
+	depends on PCI && PCI_MSI
+	select GENERIC_IRQ_CHIP
+	select PCI_MSI_IRQ_DOMAIN
+
 config ATMEL_AIC_IRQ
 	bool
 	select GENERIC_IRQ_CHIP
diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile
index 18caacb60d58..57f68e0eea74 100644
--- a/drivers/irqchip/Makefile
+++ b/drivers/irqchip/Makefile
@@ -1,5 +1,6 @@
 obj-$(CONFIG_IRQCHIP)			+= irqchip.o
 
+obj-$(CONFIG_ALPINE_MSI)		+= irq-alpine-msi.o
 obj-$(CONFIG_ARCH_BCM2835)		+= irq-bcm2835.o
 obj-$(CONFIG_ARCH_BCM2835)		+= irq-bcm2836.o
 obj-$(CONFIG_ARCH_EXYNOS)		+= exynos-combiner.o
diff --git a/drivers/irqchip/irq-alpine-msi.c b/drivers/irqchip/irq-alpine-msi.c
new file mode 100644
index 000000000000..f8712722a606
--- /dev/null
+++ b/drivers/irqchip/irq-alpine-msi.c
@@ -0,0 +1,293 @@
+/*
+ * Annapurna Labs MSIX support services
+ *
+ * Copyright (C) 2016, Amazon.com, Inc. or its affiliates. All Rights Reserved.
+ *
+ * Antoine Tenart <antoine.tenart@free-electrons.com>
+ *
+ * This file is licensed under the terms of the GNU General Public
+ * License version 2. This program is licensed "as is" without any
+ * warranty of any kind, whether express or implied.
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/irqchip.h>
+#include <linux/irqchip/arm-gic.h>
+#include <linux/msi.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+#include <linux/of_pci.h>
+#include <linux/pci.h>
+#include <linux/slab.h>
+
+#include <asm/irq.h>
+#include <asm-generic/msi.h>
+
+/* MSIX message address format: local GIC target */
+#define ALPINE_MSIX_SPI_TARGET_CLUSTER0		BIT(16)
+
+struct alpine_msix_data {
+	spinlock_t msi_map_lock;
+	phys_addr_t addr;
+	u32 spi_first;		/* The SGI number that MSIs start */
+	u32 num_spis;		/* The number of SGIs for MSIs */
+	unsigned long *msi_map;
+};
+
+static void alpine_msix_mask_msi_irq(struct irq_data *d)
+{
+	pci_msi_mask_irq(d);
+	irq_chip_mask_parent(d);
+}
+
+static void alpine_msix_unmask_msi_irq(struct irq_data *d)
+{
+	pci_msi_unmask_irq(d);
+	irq_chip_unmask_parent(d);
+}
+
+static struct irq_chip alpine_msix_irq_chip = {
+	.name			= "MSIx",
+	.irq_mask		= alpine_msix_mask_msi_irq,
+	.irq_unmask		= alpine_msix_unmask_msi_irq,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+};
+
+static int alpine_msix_allocate_sgi(struct alpine_msix_data *priv, int num_req)
+{
+	int first;
+
+	spin_lock(&priv->msi_map_lock);
+
+	first = bitmap_find_next_zero_area(priv->msi_map, priv->num_spis, 0,
+					   num_req, 0);
+	if (first >= priv->num_spis) {
+		spin_unlock(&priv->msi_map_lock);
+		return -ENOSPC;
+	}
+
+	bitmap_set(priv->msi_map, first, num_req);
+
+	spin_unlock(&priv->msi_map_lock);
+
+	return priv->spi_first + first;
+}
+
+static void alpine_msix_free_sgi(struct alpine_msix_data *priv, unsigned sgi,
+				 int num_req)
+{
+	int first = sgi - priv->spi_first;
+
+	spin_lock(&priv->msi_map_lock);
+
+	bitmap_clear(priv->msi_map, first, num_req);
+
+	spin_unlock(&priv->msi_map_lock);
+}
+
+static void alpine_msix_compose_msi_msg(struct irq_data *data,
+					struct msi_msg *msg)
+{
+	struct alpine_msix_data *priv = irq_data_get_irq_chip_data(data);
+	phys_addr_t msg_addr = priv->addr;
+
+	msg_addr |= (data->hwirq << 3);
+
+	msg->address_hi = upper_32_bits(msg_addr);
+	msg->address_lo = lower_32_bits(msg_addr);
+	msg->data = 0;
+}
+
+static struct msi_domain_info alpine_msix_domain_info = {
+	.flags	= MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
+		  MSI_FLAG_PCI_MSIX,
+	.chip	= &alpine_msix_irq_chip,
+};
+
+static struct irq_chip middle_irq_chip = {
+	.name			= "alpine_msix_middle",
+	.irq_mask		= irq_chip_mask_parent,
+	.irq_unmask		= irq_chip_unmask_parent,
+	.irq_eoi		= irq_chip_eoi_parent,
+	.irq_set_affinity	= irq_chip_set_affinity_parent,
+	.irq_compose_msi_msg	= alpine_msix_compose_msi_msg,
+};
+
+static int alpine_msix_gic_domain_alloc(struct irq_domain *domain,
+					unsigned int virq, int sgi)
+{
+	struct irq_fwspec fwspec;
+	struct irq_data *d;
+	int ret;
+
+	if (!is_of_node(domain->parent->fwnode))
+		return -EINVAL;
+
+	fwspec.fwnode = domain->parent->fwnode;
+	fwspec.param_count = 3;
+	fwspec.param[0] = 0;
+	fwspec.param[1] = sgi;
+	fwspec.param[2] = IRQ_TYPE_EDGE_RISING;
+
+	ret = irq_domain_alloc_irqs_parent(domain, virq, 1, &fwspec);
+	if (ret)
+		return ret;
+
+	d = irq_domain_get_irq_data(domain->parent, virq);
+	d->chip->irq_set_type(d, IRQ_TYPE_EDGE_RISING);
+
+	return 0;
+}
+
+static int alpine_msix_middle_domain_alloc(struct irq_domain *domain,
+					   unsigned int virq,
+					   unsigned int nr_irqs, void *args)
+{
+	struct alpine_msix_data *priv = domain->host_data;
+	int sgi, err, i;
+
+	sgi = alpine_msix_allocate_sgi(priv, nr_irqs);
+	if (sgi < 0)
+		return sgi;
+
+	for (i = 0; i < nr_irqs; i++) {
+		err = alpine_msix_gic_domain_alloc(domain, virq + i, sgi + i);
+		if (err)
+			goto err_sgi;
+
+		irq_domain_set_hwirq_and_chip(domain, virq + i, sgi + i,
+					      &middle_irq_chip, priv);
+	}
+
+	return 0;
+
+err_sgi:
+	while (--i >= 0)
+		irq_domain_free_irqs_parent(domain, virq, i);
+	alpine_msix_free_sgi(priv, sgi, nr_irqs);
+	return err;
+}
+
+static void alpine_msix_middle_domain_free(struct irq_domain *domain,
+					   unsigned int virq,
+					   unsigned int nr_irqs)
+{
+	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
+	struct alpine_msix_data *priv = irq_data_get_irq_chip_data(d);
+
+	irq_domain_free_irqs_parent(domain, virq, nr_irqs);
+	alpine_msix_free_sgi(priv, d->hwirq, nr_irqs);
+}
+
+static const struct irq_domain_ops alpine_msix_middle_domain_ops = {
+	.alloc	= alpine_msix_middle_domain_alloc,
+	.free	= alpine_msix_middle_domain_free,
+};
+
+static int alpine_msix_init_domains(struct alpine_msix_data *priv,
+				    struct device_node *node)
+{
+	struct irq_domain *middle_domain, *msi_domain, *gic_domain;
+	struct device_node *gic_node;
+
+	gic_node = of_irq_find_parent(node);
+	if (!gic_node) {
+		pr_err("Failed to find the GIC node\n");
+		return -ENODEV;
+	}
+
+	gic_domain = irq_find_host(gic_node);
+	if (!gic_domain) {
+		pr_err("Failed to find the GIC domain\n");
+		return -ENXIO;
+	}
+
+	middle_domain = irq_domain_add_tree(NULL,
+					    &alpine_msix_middle_domain_ops,
+					    priv);
+	if (!middle_domain) {
+		pr_err("Failed to create the MSIX middle domain\n");
+		return -ENOMEM;
+	}
+
+	middle_domain->parent = gic_domain;
+
+	msi_domain = pci_msi_create_irq_domain(of_node_to_fwnode(node),
+					       &alpine_msix_domain_info,
+					       middle_domain);
+	if (!msi_domain) {
+		pr_err("Failed to create MSI domain\n");
+		irq_domain_remove(msi_domain);
+		return -ENOMEM;
+	}
+
+	return 0;
+}
+
+static int alpine_msix_init(struct device_node *node,
+			    struct device_node *parent)
+{
+	struct alpine_msix_data *priv;
+	struct resource res;
+	int ret;
+
+	priv = kzalloc(sizeof(*priv), GFP_KERNEL);
+	if (!priv)
+		return -ENOMEM;
+
+	spin_lock_init(&priv->msi_map_lock);
+
+	ret = of_address_to_resource(node, 0, &res);
+	if (ret) {
+		pr_err("Failed to allocate resource\n");
+		goto err_priv;
+	}
+
+	/*
+	 * The 20 least significant bits of addr provide direct information
+	 * regarding the interrupt destination.
+	 *
+	 * To select the primary GIC as the target GIC, bits [18:17] must be set
+	 * to 0x0. In this case, bit 16 (SPI_TARGET_CLUSTER0) must be set.
+	 */
+	priv->addr = res.start & GENMASK_ULL(63,20);
+	priv->addr |= ALPINE_MSIX_SPI_TARGET_CLUSTER0;
+
+	if (of_property_read_u32(node, "al,msi-base-spi", &priv->spi_first)) {
+		pr_err("Unable to parse MSI base\n");
+		ret = -EINVAL;
+		goto err_priv;
+	}
+
+	if (of_property_read_u32(node, "al,msi-num-spis", &priv->num_spis)) {
+		pr_err("Unable to parse MSI numbers\n");
+		ret = -EINVAL;
+		goto err_priv;
+	}
+
+	priv->msi_map = kzalloc(sizeof(*priv->msi_map) * BITS_TO_LONGS(priv->num_spis),
+				GFP_KERNEL);
+	if (!priv->msi_map) {
+		ret = -ENOMEM;
+		goto err_priv;
+	}
+
+	pr_debug("Registering %d msixs, starting at %d\n",
+		 priv->num_spis, priv->spi_first);
+
+	ret = alpine_msix_init_domains(priv, node);
+	if (ret)
+		goto err_map;
+
+	return 0;
+
+err_map:
+	kfree(priv->msi_map);
+err_priv:
+	kfree(priv);
+	return ret;
+}
+IRQCHIP_DECLARE(alpine_msix, "al,alpine-msix", alpine_msix_init);
-- 
2.7.0

[PATCH v2 3/7] Documentation: bindings: document the Alpine MSIX driver

[Antoine Tenart]

Following the addition of the Alpine MSIX driver, this patch adds the
corresponding bindings documentation.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 .../interrupt-controller/al,alpine-msix.txt        | 24 ++++++++++++++++++++++
 1 file changed, 24 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt

diff --git a/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
new file mode 100644
index 000000000000..c7d3d0192f5d
--- /dev/null
+++ b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
@@ -0,0 +1,24 @@
+Alpine MSIX controller
+
+Required properties:
+
+- compatible: should be "al,alpine-msix"
+- reg: physical base address and size of the registers
+- interrupt-parent: specifies the parent interrupt controller.
+- interrupt-controller: identifies the node as an interrupt controller
+- msi-controller: identifies the node as an PCI Message Signaled Interrupt
+		  controller
+- al,msi-base-spi: SPI base of the MSI frame
+- al,msi-num-spis: number of SPIs assigned to the MSI frame
+
+Example:
+
+msix: msix {
+	compatible = "al,alpine-msix";
+	reg = <0x0 0xfbe00000 0x0 0x100000>;
+	interrupt-parent = <&gic>;
+	interrupt-controller;
+	msi-controller;
+	al,msi-base-spi = <160>;
+	al,msi-num-spis = <160>;
+};
-- 
2.7.0

[PATCH v2 3/7] Documentation: bindings: document the Alpine MSIX driver

[Antoine Tenart]

Following the addition of the Alpine MSIX driver, this patch adds the
corresponding bindings documentation.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 .../interrupt-controller/al,alpine-msix.txt        | 24 ++++++++++++++++++++++
 1 file changed, 24 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt

diff --git a/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
new file mode 100644
index 000000000000..c7d3d0192f5d
--- /dev/null
+++ b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
@@ -0,0 +1,24 @@
+Alpine MSIX controller
+
+Required properties:
+
+- compatible: should be "al,alpine-msix"
+- reg: physical base address and size of the registers
+- interrupt-parent: specifies the parent interrupt controller.
+- interrupt-controller: identifies the node as an interrupt controller
+- msi-controller: identifies the node as an PCI Message Signaled Interrupt
+		  controller
+- al,msi-base-spi: SPI base of the MSI frame
+- al,msi-num-spis: number of SPIs assigned to the MSI frame
+
+Example:
+
+msix: msix {
+	compatible = "al,alpine-msix";
+	reg = <0x0 0xfbe00000 0x0 0x100000>;
+	interrupt-parent = <&gic>;
+	interrupt-controller;
+	msi-controller;
+	al,msi-base-spi = <160>;
+	al,msi-num-spis = <160>;
+};
-- 
2.7.0

[PATCH v2 4/7] arm64: dts: alpine: add the MSIX node in the Alpine v2 dtsi

[Antoine Tenart]

Following the addition of the Alpine MSIX controller driver, add the
corresponding node in the Alpine v2 device tree.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm64/boot/dts/al/alpine-v2.dtsi | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/arch/arm64/boot/dts/al/alpine-v2.dtsi b/arch/arm64/boot/dts/al/alpine-v2.dtsi
index c021e077343b..5b7bef684256 100644
--- a/arch/arm64/boot/dts/al/alpine-v2.dtsi
+++ b/arch/arm64/boot/dts/al/alpine-v2.dtsi
@@ -139,6 +139,16 @@
 			/* 32 bit non prefetchable memory space */
 			ranges = <0x2000000 0x0 0xfe000000 0x0 0xfe000000 0x0 0x1000000>;
 			bus-range = <0x00 0x00>;
+			msi-parent = <&msix>;
+		};
+
+		msix: msix@fbe00000 {
+			compatible = "al,alpine-msix";
+			reg = <0x0 0xfbe00000 0x0 0x100000>;
+			interrupt-controller;
+			msi-controller;
+			al,msi-base-spi = <160>;
+			al,msi-num-spis = <160>;
 		};
 
 		io-fabric {
-- 
2.7.0

[PATCH v2 4/7] arm64: dts: alpine: add the MSIX node in the Alpine v2 dtsi

[Antoine Tenart]

Following the addition of the Alpine MSIX controller driver, add the
corresponding node in the Alpine v2 device tree.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm64/boot/dts/al/alpine-v2.dtsi | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/arch/arm64/boot/dts/al/alpine-v2.dtsi b/arch/arm64/boot/dts/al/alpine-v2.dtsi
index c021e077343b..5b7bef684256 100644
--- a/arch/arm64/boot/dts/al/alpine-v2.dtsi
+++ b/arch/arm64/boot/dts/al/alpine-v2.dtsi
@@ -139,6 +139,16 @@
 			/* 32 bit non prefetchable memory space */
 			ranges = <0x2000000 0x0 0xfe000000 0x0 0xfe000000 0x0 0x1000000>;
 			bus-range = <0x00 0x00>;
+			msi-parent = <&msix>;
+		};
+
+		msix: msix at fbe00000 {
+			compatible = "al,alpine-msix";
+			reg = <0x0 0xfbe00000 0x0 0x100000>;
+			interrupt-controller;
+			msi-controller;
+			al,msi-base-spi = <160>;
+			al,msi-num-spis = <160>;
 		};
 
 		io-fabric {
-- 
2.7.0

[PATCH v2 5/7] ARM: dts: alpine: add the MSIX node

[Antoine Tenart]

With the newly available MSIX driver for Alpine, add the corresponding
node in the Alpine device tree.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
---
 arch/arm/boot/dts/alpine.dtsi | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/arch/arm/boot/dts/alpine.dtsi b/arch/arm/boot/dts/alpine.dtsi
index 9af2d60e9a7f..db8752fc480e 100644
--- a/arch/arm/boot/dts/alpine.dtsi
+++ b/arch/arm/boot/dts/alpine.dtsi
@@ -155,6 +155,16 @@
 			ranges = <0x02000000 0x0 0xfe000000 0x0 0xfe000000 0x0 0x1000000>;
 
 			bus-range = <0x00 0x00>;
+			msi-parent = <&msix>;
+		};
+
+		msix: msix@fbe00000 {
+			compatible = "al,alpine-msix";
+			reg = <0x0 0xfbe00000 0x0 0x100000>;
+			interrupt-controller;
+			msi-controller;
+			al,msi-base-spi = <96>;
+			al,msi-num-spis = <64>;
 		};
 	};
 };
-- 
2.7.0

[PATCH v2 5/7] ARM: dts: alpine: add the MSIX node

[Antoine Tenart]

With the newly available MSIX driver for Alpine, add the corresponding
node in the Alpine device tree.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
---
 arch/arm/boot/dts/alpine.dtsi | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/arch/arm/boot/dts/alpine.dtsi b/arch/arm/boot/dts/alpine.dtsi
index 9af2d60e9a7f..db8752fc480e 100644
--- a/arch/arm/boot/dts/alpine.dtsi
+++ b/arch/arm/boot/dts/alpine.dtsi
@@ -155,6 +155,16 @@
 			ranges = <0x02000000 0x0 0xfe000000 0x0 0xfe000000 0x0 0x1000000>;
 
 			bus-range = <0x00 0x00>;
+			msi-parent = <&msix>;
+		};
+
+		msix: msix at fbe00000 {
+			compatible = "al,alpine-msix";
+			reg = <0x0 0xfbe00000 0x0 0x100000>;
+			interrupt-controller;
+			msi-controller;
+			al,msi-base-spi = <96>;
+			al,msi-num-spis = <64>;
 		};
 	};
 };
-- 
2.7.0

[PATCH v2 6/7] arm64: alpine: select the Alpine MSI controller driver

[Antoine Tenart]

Select the Alpine MSI controller driver when using an Alpine platform.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm64/Kconfig.platforms | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/Kconfig.platforms b/arch/arm64/Kconfig.platforms
index bfbefa3e0dbe..6761e12e5e13 100644
--- a/arch/arm64/Kconfig.platforms
+++ b/arch/arm64/Kconfig.platforms
@@ -2,6 +2,7 @@ menu "Platform selection"
 
 config ARCH_ALPINE
 	bool "Annapurna Labs Alpine platform"
+	select ALPINE_MSI
 	help
 	  This enables support for the Annapurna Labs Alpine
 	  Soc family.
-- 
2.7.0

[PATCH v2 6/7] arm64: alpine: select the Alpine MSI controller driver

[Antoine Tenart]

Select the Alpine MSI controller driver when using an Alpine platform.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm64/Kconfig.platforms | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/Kconfig.platforms b/arch/arm64/Kconfig.platforms
index bfbefa3e0dbe..6761e12e5e13 100644
--- a/arch/arm64/Kconfig.platforms
+++ b/arch/arm64/Kconfig.platforms
@@ -2,6 +2,7 @@ menu "Platform selection"
 
 config ARCH_ALPINE
 	bool "Annapurna Labs Alpine platform"
+	select ALPINE_MSI
 	help
 	  This enables support for the Annapurna Labs Alpine
 	  Soc family.
-- 
2.7.0

[PATCH v2 7/7] arm: alpine: select the Alpine MSI controller driver

[Antoine Tenart]

Select the Alpine MSI controller driver when using an Alpine platform.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm/mach-alpine/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/mach-alpine/Kconfig b/arch/arm/mach-alpine/Kconfig
index 5c2d54f59f53..b41838a58ae4 100644
--- a/arch/arm/mach-alpine/Kconfig
+++ b/arch/arm/mach-alpine/Kconfig
@@ -1,6 +1,7 @@
 config ARCH_ALPINE
 	bool "Annapurna Labs Alpine platform"
 	depends on ARCH_MULTI_V7
+	select ALPINE_MSI
 	select ARM_AMBA
 	select ARM_GIC
 	select GENERIC_IRQ_CHIP
-- 
2.7.0

[PATCH v2 7/7] arm: alpine: select the Alpine MSI controller driver

[Antoine Tenart]

Select the Alpine MSI controller driver when using an Alpine platform.

Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
---
 arch/arm/mach-alpine/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm/mach-alpine/Kconfig b/arch/arm/mach-alpine/Kconfig
index 5c2d54f59f53..b41838a58ae4 100644
--- a/arch/arm/mach-alpine/Kconfig
+++ b/arch/arm/mach-alpine/Kconfig
@@ -1,6 +1,7 @@
 config ARCH_ALPINE
 	bool "Annapurna Labs Alpine platform"
 	depends on ARCH_MULTI_V7
+	select ALPINE_MSI
 	select ARM_AMBA
 	select ARM_GIC
 	select GENERIC_IRQ_CHIP
-- 
2.7.0

Re: [PATCH v2 1/7] irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> Always return IRQ_SET_MASK_OK_DONE instead of IRQ_SET_MASK_OK when the
> affinity has been updated. When using stacked irqchips, returning
> IRQ_SET_MASK_OK_DONE means skipping all descendant irqchips.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>

Acked-by: Marc Zyngier <marc.zyngier@arm.com>

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v2 1/7] irqchip/gic-v3: always return IRQ_SET_MASK_OK_DONE in gic_set_affinity

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> Always return IRQ_SET_MASK_OK_DONE instead of IRQ_SET_MASK_OK when the
> affinity has been updated. When using stacked irqchips, returning
> IRQ_SET_MASK_OK_DONE means skipping all descendant irqchips.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>

Acked-by: Marc Zyngier <marc.zyngier@arm.com>

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v2 2/7] irqchip: add the Alpine MSIX interrupt controller

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> This patch adds the Alpine MSIX interrupt controller driver.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
> Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>

Acked-by: Marc Zyngier <marc.zyngier@arm.com>

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v2 2/7] irqchip: add the Alpine MSIX interrupt controller

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> This patch adds the Alpine MSIX interrupt controller driver.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
> Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>

Acked-by: Marc Zyngier <marc.zyngier@arm.com>

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v2 3/7] Documentation: bindings: document the Alpine MSIX driver

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> Following the addition of the Alpine MSIX driver, this patch adds the
> corresponding bindings documentation.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
> Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
> ---
>  .../interrupt-controller/al,alpine-msix.txt        | 24 ++++++++++++++++++++++
>  1 file changed, 24 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> 
> diff --git a/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> new file mode 100644
> index 000000000000..c7d3d0192f5d
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> @@ -0,0 +1,24 @@
> +Alpine MSIX controller
> +
> +Required properties:
> +
> +- compatible: should be "al,alpine-msix"
> +- reg: physical base address and size of the registers
> +- interrupt-parent: specifies the parent interrupt controller.
> +- interrupt-controller: identifies the node as an interrupt controller
> +- msi-controller: identifies the node as an PCI Message Signaled Interrupt
> +		  controller
> +- al,msi-base-spi: SPI base of the MSI frame
> +- al,msi-num-spis: number of SPIs assigned to the MSI frame

It would probably be good to reference the GIC bindings so that the
"SPI" acronym makes actual sense. It will also disambiguate the "base"
aspect (is this number relative to IRQ0? or SPI0? - I assume the latter,
but that's by looking at the code).

> +
> +Example:
> +
> +msix: msix {
> +	compatible = "al,alpine-msix";
> +	reg = <0x0 0xfbe00000 0x0 0x100000>;
> +	interrupt-parent = <&gic>;
> +	interrupt-controller;
> +	msi-controller;
> +	al,msi-base-spi = <160>;
> +	al,msi-num-spis = <160>;
> +};
> 

Otherwise looks good to me.

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v2 3/7] Documentation: bindings: document the Alpine MSIX driver

[Marc Zyngier]

On 10/02/16 13:22, Antoine Tenart wrote:
> Following the addition of the Alpine MSIX driver, this patch adds the
> corresponding bindings documentation.
> 
> Signed-off-by: Antoine Tenart <antoine.tenart@free-electrons.com>
> Signed-off-by: Tsahee Zidenberg <tsahee@annapurnalabs.com>
> ---
>  .../interrupt-controller/al,alpine-msix.txt        | 24 ++++++++++++++++++++++
>  1 file changed, 24 insertions(+)
>  create mode 100644 Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> 
> diff --git a/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> new file mode 100644
> index 000000000000..c7d3d0192f5d
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/interrupt-controller/al,alpine-msix.txt
> @@ -0,0 +1,24 @@
> +Alpine MSIX controller
> +
> +Required properties:
> +
> +- compatible: should be "al,alpine-msix"
> +- reg: physical base address and size of the registers
> +- interrupt-parent: specifies the parent interrupt controller.
> +- interrupt-controller: identifies the node as an interrupt controller
> +- msi-controller: identifies the node as an PCI Message Signaled Interrupt
> +		  controller
> +- al,msi-base-spi: SPI base of the MSI frame
> +- al,msi-num-spis: number of SPIs assigned to the MSI frame

It would probably be good to reference the GIC bindings so that the
"SPI" acronym makes actual sense. It will also disambiguate the "base"
aspect (is this number relative to IRQ0? or SPI0? - I assume the latter,
but that's by looking at the code).

> +
> +Example:
> +
> +msix: msix {
> +	compatible = "al,alpine-msix";
> +	reg = <0x0 0xfbe00000 0x0 0x100000>;
> +	interrupt-parent = <&gic>;
> +	interrupt-controller;
> +	msi-controller;
> +	al,msi-base-spi = <160>;
> +	al,msi-num-spis = <160>;
> +};
> 

Otherwise looks good to me.

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Marc Zyngier]

Antoine,

On 10/02/16 13:22, Antoine Tenart wrote:
> Hi all,
> 
> This series introduce the Alpine MSIX driver, and enables it in both
> the Alpine v1 and Alpine v2 device trees.
> 
> This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> https://lkml.org/lkml/2016/2/10/83
> 
> You can find the series at:
> https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix

So I'm quite happy with this as it is (modulo the documentation fix on
patch 3). I can take the first 3 patches through my GIC tree, but the
last 4 patches depend on your other series.

Would it be a problem to get these 4 patches merged after 4.6-rc1, so
that all the code can be merged before?

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Marc Zyngier]

Antoine,

On 10/02/16 13:22, Antoine Tenart wrote:
> Hi all,
> 
> This series introduce the Alpine MSIX driver, and enables it in both
> the Alpine v1 and Alpine v2 device trees.
> 
> This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> https://lkml.org/lkml/2016/2/10/83
> 
> You can find the series at:
> https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix

So I'm quite happy with this as it is (modulo the documentation fix on
patch 3). I can take the first 3 patches through my GIC tree, but the
last 4 patches depend on your other series.

Would it be a problem to get these 4 patches merged after 4.6-rc1, so
that all the code can be merged before?

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...

Re: [PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Antoine Tenart]

[-- Attachment #1: Type: text/plain, Size: 999 bytes --]

Hello Marc,

On Thu, Feb 11, 2016 at 01:41:48PM +0000, Marc Zyngier wrote:
> On 10/02/16 13:22, Antoine Tenart wrote:
> > 
> > This series introduce the Alpine MSIX driver, and enables it in both
> > the Alpine v1 and Alpine v2 device trees.
> > 
> > This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> > https://lkml.org/lkml/2016/2/10/83
> > 
> > You can find the series at:
> > https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix
> 
> So I'm quite happy with this as it is (modulo the documentation fix on
> patch 3). I can take the first 3 patches through my GIC tree, but the
> last 4 patches depend on your other series.

I'll fix the documentation.

> Would it be a problem to get these 4 patches merged after 4.6-rc1, so
> that all the code can be merged before?

I think we can live with that :)

Thanks,

Antoine

-- 
Antoine Ténart, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]

[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Antoine Tenart]

Hello Marc,

On Thu, Feb 11, 2016 at 01:41:48PM +0000, Marc Zyngier wrote:
> On 10/02/16 13:22, Antoine Tenart wrote:
> > 
> > This series introduce the Alpine MSIX driver, and enables it in both
> > the Alpine v1 and Alpine v2 device trees.
> > 
> > This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> > https://lkml.org/lkml/2016/2/10/83
> > 
> > You can find the series at:
> > https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix
> 
> So I'm quite happy with this as it is (modulo the documentation fix on
> patch 3). I can take the first 3 patches through my GIC tree, but the
> last 4 patches depend on your other series.

I'll fix the documentation.

> Would it be a problem to get these 4 patches merged after 4.6-rc1, so
> that all the code can be merged before?

I think we can live with that :)

Thanks,

Antoine

-- 
Antoine T?nart, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160211/062ef4de/attachment.sig>

Re: [PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Thomas Petazzoni]

Hello,

On Thu, 11 Feb 2016 15:45:44 +0100, Antoine Tenart wrote:

> On Thu, Feb 11, 2016 at 01:41:48PM +0000, Marc Zyngier wrote:
> > On 10/02/16 13:22, Antoine Tenart wrote:
> > > 
> > > This series introduce the Alpine MSIX driver, and enables it in both
> > > the Alpine v1 and Alpine v2 device trees.
> > > 
> > > This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> > > https://lkml.org/lkml/2016/2/10/83
> > > 
> > > You can find the series at:
> > > https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix
> > 
> > So I'm quite happy with this as it is (modulo the documentation fix on
> > patch 3). I can take the first 3 patches through my GIC tree, but the
> > last 4 patches depend on your other series.
> 
> I'll fix the documentation.
> 
> > Would it be a problem to get these 4 patches merged after 4.6-rc1, so
> > that all the code can be merged before?
> 
> I think we can live with that :)

Practically speaking, patches 4-7 have no build dependency on patches
1-3, so they can be merged independently. As long as the DT binding
introduced in patches 1-3 is validated, then patches 4-7 can be merged
in parallel.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[PATCH v2 0/7] irqchip: introduce the Alpine MSIX driver

[Thomas Petazzoni]

Hello,

On Thu, 11 Feb 2016 15:45:44 +0100, Antoine Tenart wrote:

> On Thu, Feb 11, 2016 at 01:41:48PM +0000, Marc Zyngier wrote:
> > On 10/02/16 13:22, Antoine Tenart wrote:
> > > 
> > > This series introduce the Alpine MSIX driver, and enables it in both
> > > the Alpine v1 and Alpine v2 device trees.
> > > 
> > > This series depends on "[PATCH v2 0/3] arm64: introduce the Alpine support":
> > > https://lkml.org/lkml/2016/2/10/83
> > > 
> > > You can find the series at:
> > > https://github.com/atenart/linux.git 4.5-rc1/alpinev2-msix
> > 
> > So I'm quite happy with this as it is (modulo the documentation fix on
> > patch 3). I can take the first 3 patches through my GIC tree, but the
> > last 4 patches depend on your other series.
> 
> I'll fix the documentation.
> 
> > Would it be a problem to get these 4 patches merged after 4.6-rc1, so
> > that all the code can be merged before?
> 
> I think we can live with that :)

Practically speaking, patches 4-7 have no build dependency on patches
1-3, so they can be merged independently. As long as the DT binding
introduced in patches 1-3 is validated, then patches 4-7 can be merged
in parallel.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com