diff --git a/drivers/irqchip/irq-gic-v2m.c b/drivers/irqchip/irq-gic-v2m.c
index 3c77ab676e54a4..5356739d4799d4 100644
--- a/drivers/irqchip/irq-gic-v2m.c
+++ b/drivers/irqchip/irq-gic-v2m.c
@@ -56,6 +56,7 @@
 
 /* List of flags for specific v2m implementation */
 #define GICV2M_NEEDS_SPI_OFFSET		0x00000001
+#define GICV2M_GRAVITON_ADDRESS_ONLY	0x00000002
 
 static LIST_HEAD(v2m_nodes);
 static DEFINE_SPINLOCK(v2m_lock);
@@ -98,15 +99,26 @@ static struct msi_domain_info gicv2m_msi_domain_info = {
 	.chip	= &gicv2m_msi_irq_chip,
 };
 
+static phys_addr_t gicv2m_get_msi_addr(struct v2m_data *v2m, int hwirq)
+{
+	if (v2m->flags & GICV2M_GRAVITON_ADDRESS_ONLY)
+		return v2m->res.start | ((hwirq - 32) << 3);
+	else
+		return v2m->res.start + V2M_MSI_SETSPI_NS;
+}
+
 static void gicv2m_compose_msi_msg(struct irq_data *data, struct msi_msg *msg)
 {
 	struct v2m_data *v2m = irq_data_get_irq_chip_data(data);
-	phys_addr_t addr = v2m->res.start + V2M_MSI_SETSPI_NS;
+	phys_addr_t addr = gicv2m_get_msi_addr(v2m, data->hwirq);
 
 	msg->address_hi = upper_32_bits(addr);
 	msg->address_lo = lower_32_bits(addr);
-	msg->data = data->hwirq;
 
+	if (v2m->flags & GICV2M_GRAVITON_ADDRESS_ONLY)
+		msg->data = 0;
+	else
+		msg->data = data->hwirq;
 	if (v2m->flags & GICV2M_NEEDS_SPI_OFFSET)
 		msg->data -= v2m->spi_offset;
 
@@ -188,7 +200,7 @@ static int gicv2m_irq_domain_alloc(struct irq_domain *domain, unsigned int virq,
 	hwirq = v2m->spi_start + offset;
 
 	err = iommu_dma_prepare_msi(info->desc,
-				    v2m->res.start + V2M_MSI_SETSPI_NS);
+				    gicv2m_get_msi_addr(v2m, hwirq));
 	if (err)
 		return err;
 
@@ -307,7 +319,7 @@ static int gicv2m_allocate_domains(struct irq_domain *parent)
 
 static int __init gicv2m_init_one(struct fwnode_handle *fwnode,
 				  u32 spi_start, u32 nr_spis,
-				  struct resource *res)
+				  struct resource *res, u32 flags)
 {
 	int ret;
 	struct v2m_data *v2m;
@@ -320,6 +332,7 @@ static int __init gicv2m_init_one(struct fwnode_handle *fwnode,
 
 	INIT_LIST_HEAD(&v2m->entry);
 	v2m->fwnode = fwnode;
+	v2m->flags = flags;
 
 	memcpy(&v2m->res, res, sizeof(struct resource));
 
@@ -334,7 +347,14 @@ static int __init gicv2m_init_one(struct fwnode_handle *fwnode,
 		v2m->spi_start = spi_start;
 		v2m->nr_spis = nr_spis;
 	} else {
-		u32 typer = readl_relaxed(v2m->base + V2M_MSI_TYPER);
+		u32 typer;
+
+		/* Graviton should always have explicit spi_start/nr_spis */
+		if (v2m->flags & GICV2M_GRAVITON_ADDRESS_ONLY) {
+			ret = -EINVAL;
+			goto err_iounmap;
+		}
+		typer = readl_relaxed(v2m->base + V2M_MSI_TYPER);
 
 		v2m->spi_start = V2M_MSI_TYPER_BASE_SPI(typer);
 		v2m->nr_spis = V2M_MSI_TYPER_NUM_SPI(typer);
@@ -355,18 +375,21 @@ static int __init gicv2m_init_one(struct fwnode_handle *fwnode,
 	 *
 	 * Broadom NS2 GICv2m implementation has an erratum where the MSI data
 	 * is 'spi_number - 32'
+	 *
+	 * Reading that register fails on the Graviton implementation
 	 */
-	switch (readl_relaxed(v2m->base + V2M_MSI_IIDR)) {
-	case XGENE_GICV2M_MSI_IIDR:
-		v2m->flags |= GICV2M_NEEDS_SPI_OFFSET;
-		v2m->spi_offset = v2m->spi_start;
-		break;
-	case BCM_NS2_GICV2M_MSI_IIDR:
-		v2m->flags |= GICV2M_NEEDS_SPI_OFFSET;
-		v2m->spi_offset = 32;
-		break;
+	if (!(v2m->flags & GICV2M_GRAVITON_ADDRESS_ONLY)) {
+		switch (readl_relaxed(v2m->base + V2M_MSI_IIDR)) {
+		case XGENE_GICV2M_MSI_IIDR:
+			v2m->flags |= GICV2M_NEEDS_SPI_OFFSET;
+			v2m->spi_offset = v2m->spi_start;
+			break;
+		case BCM_NS2_GICV2M_MSI_IIDR:
+			v2m->flags |= GICV2M_NEEDS_SPI_OFFSET;
+			v2m->spi_offset = 32;
+			break;
+		}
 	}
-
 	v2m->bm = kcalloc(BITS_TO_LONGS(v2m->nr_spis), sizeof(long),
 			  GFP_KERNEL);
 	if (!v2m->bm) {
@@ -419,7 +442,8 @@ static int __init gicv2m_of_init(struct fwnode_handle *parent_handle,
 			pr_info("DT overriding V2M MSI_TYPER (base:%u, num:%u)\n",
 				spi_start, nr_spis);
 
-		ret = gicv2m_init_one(&child->fwnode, spi_start, nr_spis, &res);
+		ret = gicv2m_init_one(&child->fwnode, spi_start, nr_spis,
+				      &res, 0);
 		if (ret) {
 			of_node_put(child);
 			break;
@@ -451,6 +475,25 @@ static struct fwnode_handle *gicv2m_get_fwnode(struct device *dev)
 	return data->fwnode;
 }
 
+static bool acpi_check_amazon_graviton_quirks(void)
+{
+	static struct acpi_table_madt *madt;
+	acpi_status status;
+	bool rc = false;
+
+#define ACPI_AMZN_OEM_ID		"AMAZON"
+
+	status = acpi_get_table(ACPI_SIG_MADT, 0,
+				(struct acpi_table_header **)&madt);
+
+	if (ACPI_FAILURE(status) || !madt)
+		return rc;
+	rc = !memcmp(madt->header.oem_id, ACPI_AMZN_OEM_ID, ACPI_OEM_ID_SIZE);
+	acpi_put_table((struct acpi_table_header *)madt);
+
+	return rc;
+}
+
 static int __init
 acpi_parse_madt_msi(union acpi_subtable_headers *header,
 		    const unsigned long end)
@@ -460,6 +503,7 @@ acpi_parse_madt_msi(union acpi_subtable_headers *header,
 	u32 spi_start = 0, nr_spis = 0;
 	struct acpi_madt_generic_msi_frame *m;
 	struct fwnode_handle *fwnode;
+	u32 flags = 0;
 
 	m = (struct acpi_madt_generic_msi_frame *)header;
 	if (BAD_MADT_ENTRY(m, end))
@@ -469,6 +513,13 @@ acpi_parse_madt_msi(union acpi_subtable_headers *header,
 	res.end = m->base_address + SZ_4K - 1;
 	res.flags = IORESOURCE_MEM;
 
+	if (acpi_check_amazon_graviton_quirks()) {
+		pr_info("applying Amazon Graviton quirk\n");
+		res.end = res.start + SZ_8K - 1;
+		flags |= GICV2M_GRAVITON_ADDRESS_ONLY;
+		gicv2m_msi_domain_info.flags &= ~MSI_FLAG_MULTI_PCI_MSI;
+	}
+
 	if (m->flags & ACPI_MADT_OVERRIDE_SPI_VALUES) {
 		spi_start = m->spi_base;
 		nr_spis = m->spi_count;
@@ -483,7 +534,7 @@ acpi_parse_madt_msi(union acpi_subtable_headers *header,
 		return -EINVAL;
 	}
 
-	ret = gicv2m_init_one(fwnode, spi_start, nr_spis, &res);
+	ret = gicv2m_init_one(fwnode, spi_start, nr_spis, &res, flags);
 	if (ret)
 		irq_domain_free_fwnode(fwnode);
 
diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
index f44cd89cfc40e0..1282f81696b240 100644
--- a/drivers/irqchip/irq-gic-v3.c
+++ b/drivers/irqchip/irq-gic-v3.c
@@ -1343,6 +1343,9 @@ static int __init gic_init_bases(void __iomem *dist_base,
 	if (gic_dist_supports_lpis()) {
 		its_init(handle, &gic_data.rdists, gic_data.domain);
 		its_cpu_init();
+	} else {
+		if (IS_ENABLED(CONFIG_ARM_GIC_V2M))
+			gicv2m_init(handle, gic_data.domain);
 	}
 
 	if (gic_prio_masking_enabled()) {
diff --git a/include/linux/irqchip/arm-gic-common.h b/include/linux/irqchip/arm-gic-common.h
index 9a1a479a2bf454..62a8821047909e 100644
--- a/include/linux/irqchip/arm-gic-common.h
+++ b/include/linux/irqchip/arm-gic-common.h
@@ -39,4 +39,9 @@ struct gic_kvm_info {
 
 const struct gic_kvm_info *gic_get_kvm_info(void);
 
+struct irq_domain;
+struct fwnode_handle;
+int gicv2m_init(struct fwnode_handle *parent_handle,
+		struct irq_domain *parent);
+
 #endif /* __LINUX_IRQCHIP_ARM_GIC_COMMON_H */
diff --git a/include/linux/irqchip/arm-gic.h b/include/linux/irqchip/arm-gic.h
index 0f049b384ccddf..7bd3bc6baa4059 100644
--- a/include/linux/irqchip/arm-gic.h
+++ b/include/linux/irqchip/arm-gic.h
@@ -160,9 +160,6 @@ int gic_of_init_child(struct device *dev, struct gic_chip_data **gic, int irq);
  */
 void gic_init(void __iomem *dist , void __iomem *cpu);
 
-int gicv2m_init(struct fwnode_handle *parent_handle,
-		struct irq_domain *parent);
-
 void gic_send_sgi(unsigned int cpu_id, unsigned int irq);
 int gic_get_cpu_id(unsigned int cpu);
 void gic_migrate_target(unsigned int new_cpu_id);