From 5fcb42fbd224e1103bacbae4785745842cfd6304 Mon Sep 17 00:00:00 2001
From: Yangbo Lu <yangbo.lu@nxp.com>
Date: Wed, 17 Jan 2018 15:00:43 +0800
Subject: [PATCH 08/30] pci: support layerscape

This is an integrated patch for layerscape pcie support.

Signed-off-by: Po Liu <po.liu@nxp.com>
Signed-off-by: Liu Gang <Gang.Liu@nxp.com>
Signed-off-by: Minghuan Lian <Minghuan.Lian@freescale.com>
Signed-off-by: hongbo.wang <hongbo.wang@nxp.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@nxp.com>
Signed-off-by: Mingkai Hu <mingkai.hu@nxp.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com>
---
 drivers/irqchip/irq-ls-scfg-msi.c            | 257 +++++++--
 drivers/pci/host/Makefile                    |   2 +-
 drivers/pci/host/pci-layerscape-ep-debugfs.c | 758 +++++++++++++++++++++++++++
 drivers/pci/host/pci-layerscape-ep.c         | 309 +++++++++++
 drivers/pci/host/pci-layerscape-ep.h         | 115 ++++
 drivers/pci/host/pci-layerscape.c            |  48 +-
 drivers/pci/host/pcie-designware.c           |   6 +
 drivers/pci/host/pcie-designware.h           |   1 +
 drivers/pci/pci.c                            |   2 +-
 drivers/pci/pcie/portdrv_core.c              | 181 +++----
 drivers/pci/quirks.c                         |   8 +
 include/linux/pci.h                          |   1 +
 12 files changed, 1539 insertions(+), 149 deletions(-)
 create mode 100644 drivers/pci/host/pci-layerscape-ep-debugfs.c
 create mode 100644 drivers/pci/host/pci-layerscape-ep.c
 create mode 100644 drivers/pci/host/pci-layerscape-ep.h

--- a/drivers/irqchip/irq-ls-scfg-msi.c
+++ b/drivers/irqchip/irq-ls-scfg-msi.c
@@ -17,13 +17,32 @@
 #include <linux/irq.h>
 #include <linux/irqchip/chained_irq.h>
 #include <linux/irqdomain.h>
+#include <linux/of_irq.h>
 #include <linux/of_pci.h>
 #include <linux/of_platform.h>
 #include <linux/spinlock.h>
 
-#define MSI_MAX_IRQS	32
-#define MSI_IBS_SHIFT	3
-#define MSIR		4
+#define MSI_IRQS_PER_MSIR	32
+#define MSI_MSIR_OFFSET		4
+
+#define MSI_LS1043V1_1_IRQS_PER_MSIR	8
+#define MSI_LS1043V1_1_MSIR_OFFSET	0x10
+
+struct ls_scfg_msi_cfg {
+	u32 ibs_shift; /* Shift of interrupt bit select */
+	u32 msir_irqs; /* The irq number per MSIR */
+	u32 msir_base; /* The base address of MSIR */
+};
+
+struct ls_scfg_msir {
+	struct ls_scfg_msi *msi_data;
+	unsigned int index;
+	unsigned int gic_irq;
+	unsigned int bit_start;
+	unsigned int bit_end;
+	unsigned int srs; /* Shared interrupt register select */
+	void __iomem *reg;
+};
 
 struct ls_scfg_msi {
 	spinlock_t		lock;
@@ -32,8 +51,11 @@ struct ls_scfg_msi {
 	struct irq_domain	*msi_domain;
 	void __iomem		*regs;
 	phys_addr_t		msiir_addr;
-	int			irq;
-	DECLARE_BITMAP(used, MSI_MAX_IRQS);
+	struct ls_scfg_msi_cfg	*cfg;
+	u32			msir_num;
+	struct ls_scfg_msir	*msir;
+	u32			irqs_num;
+	unsigned long		*used;
 };
 
 static struct irq_chip ls_scfg_msi_irq_chip = {
@@ -49,19 +71,56 @@ static struct msi_domain_info ls_scfg_ms
 	.chip	= &ls_scfg_msi_irq_chip,
 };
 
+static int msi_affinity_flag = 1;
+
+static int __init early_parse_ls_scfg_msi(char *p)
+{
+	if (p && strncmp(p, "no-affinity", 11) == 0)
+		msi_affinity_flag = 0;
+	else
+		msi_affinity_flag = 1;
+
+	return 0;
+}
+early_param("lsmsi", early_parse_ls_scfg_msi);
+
 static void ls_scfg_msi_compose_msg(struct irq_data *data, struct msi_msg *msg)
 {
 	struct ls_scfg_msi *msi_data = irq_data_get_irq_chip_data(data);
 
 	msg->address_hi = upper_32_bits(msi_data->msiir_addr);
 	msg->address_lo = lower_32_bits(msi_data->msiir_addr);
-	msg->data = data->hwirq << MSI_IBS_SHIFT;
+	msg->data = data->hwirq;
+
+	if (msi_affinity_flag)
+		msg->data |= cpumask_first(data->common->affinity);
 }
 
 static int ls_scfg_msi_set_affinity(struct irq_data *irq_data,
 				    const struct cpumask *mask, bool force)
 {
-	return -EINVAL;
+	struct ls_scfg_msi *msi_data = irq_data_get_irq_chip_data(irq_data);
+	u32 cpu;
+
+	if (!msi_affinity_flag)
+		return -EINVAL;
+
+	if (!force)
+		cpu = cpumask_any_and(mask, cpu_online_mask);
+	else
+		cpu = cpumask_first(mask);
+
+	if (cpu >= msi_data->msir_num)
+		return -EINVAL;
+
+	if (msi_data->msir[cpu].gic_irq <= 0) {
+		pr_warn("cannot bind the irq to cpu%d\n", cpu);
+		return -EINVAL;
+	}
+
+	cpumask_copy(irq_data->common->affinity, mask);
+
+	return IRQ_SET_MASK_OK;
 }
 
 static struct irq_chip ls_scfg_msi_parent_chip = {
@@ -81,8 +140,8 @@ static int ls_scfg_msi_domain_irq_alloc(
 	WARN_ON(nr_irqs != 1);
 
 	spin_lock(&msi_data->lock);
-	pos = find_first_zero_bit(msi_data->used, MSI_MAX_IRQS);
-	if (pos < MSI_MAX_IRQS)
+	pos = find_first_zero_bit(msi_data->used, msi_data->irqs_num);
+	if (pos < msi_data->irqs_num)
 		__set_bit(pos, msi_data->used);
 	else
 		err = -ENOSPC;
@@ -106,7 +165,7 @@ static void ls_scfg_msi_domain_irq_free(
 	int pos;
 
 	pos = d->hwirq;
-	if (pos < 0 || pos >= MSI_MAX_IRQS) {
+	if (pos < 0 || pos >= msi_data->irqs_num) {
 		pr_err("failed to teardown msi. Invalid hwirq %d\n", pos);
 		return;
 	}
@@ -123,15 +182,22 @@ static const struct irq_domain_ops ls_sc
 
 static void ls_scfg_msi_irq_handler(struct irq_desc *desc)
 {
-	struct ls_scfg_msi *msi_data = irq_desc_get_handler_data(desc);
+	struct ls_scfg_msir *msir = irq_desc_get_handler_data(desc);
+	struct ls_scfg_msi *msi_data = msir->msi_data;
 	unsigned long val;
-	int pos, virq;
+	int pos, size, virq, hwirq;
 
 	chained_irq_enter(irq_desc_get_chip(desc), desc);
 
-	val = ioread32be(msi_data->regs + MSIR);
-	for_each_set_bit(pos, &val, MSI_MAX_IRQS) {
-		virq = irq_find_mapping(msi_data->parent, (31 - pos));
+	val = ioread32be(msir->reg);
+
+	pos = msir->bit_start;
+	size = msir->bit_end + 1;
+
+	for_each_set_bit_from(pos, &val, size) {
+		hwirq = ((msir->bit_end - pos) << msi_data->cfg->ibs_shift) |
+			msir->srs;
+		virq = irq_find_mapping(msi_data->parent, hwirq);
 		if (virq)
 			generic_handle_irq(virq);
 	}
@@ -143,7 +209,7 @@ static int ls_scfg_msi_domains_init(stru
 {
 	/* Initialize MSI domain parent */
 	msi_data->parent = irq_domain_add_linear(NULL,
-						 MSI_MAX_IRQS,
+						 msi_data->irqs_num,
 						 &ls_scfg_msi_domain_ops,
 						 msi_data);
 	if (!msi_data->parent) {
@@ -164,16 +230,118 @@ static int ls_scfg_msi_domains_init(stru
 	return 0;
 }
 
+static int ls_scfg_msi_setup_hwirq(struct ls_scfg_msi *msi_data, int index)
+{
+	struct ls_scfg_msir *msir;
+	int virq, i, hwirq;
+
+	virq = platform_get_irq(msi_data->pdev, index);
+	if (virq <= 0)
+		return -ENODEV;
+
+	msir = &msi_data->msir[index];
+	msir->index = index;
+	msir->msi_data = msi_data;
+	msir->gic_irq = virq;
+	msir->reg = msi_data->regs + msi_data->cfg->msir_base + 4 * index;
+
+	if (msi_data->cfg->msir_irqs == MSI_LS1043V1_1_IRQS_PER_MSIR) {
+		msir->bit_start = 32 - ((msir->index + 1) *
+				  MSI_LS1043V1_1_IRQS_PER_MSIR);
+		msir->bit_end = msir->bit_start +
+				MSI_LS1043V1_1_IRQS_PER_MSIR - 1;
+	} else {
+		msir->bit_start = 0;
+		msir->bit_end = msi_data->cfg->msir_irqs - 1;
+	}
+
+	irq_set_chained_handler_and_data(msir->gic_irq,
+					 ls_scfg_msi_irq_handler,
+					 msir);
+
+	if (msi_affinity_flag) {
+		/* Associate MSIR interrupt to the cpu */
+		irq_set_affinity(msir->gic_irq, get_cpu_mask(index));
+		msir->srs = 0; /* This value is determined by the CPU */
+	} else
+		msir->srs = index;
+
+	/* Release the hwirqs corresponding to this MSIR */
+	if (!msi_affinity_flag || msir->index == 0) {
+		for (i = 0; i < msi_data->cfg->msir_irqs; i++) {
+			hwirq = i << msi_data->cfg->ibs_shift | msir->index;
+			bitmap_clear(msi_data->used, hwirq, 1);
+		}
+	}
+
+	return 0;
+}
+
+static int ls_scfg_msi_teardown_hwirq(struct ls_scfg_msir *msir)
+{
+	struct ls_scfg_msi *msi_data = msir->msi_data;
+	int i, hwirq;
+
+	if (msir->gic_irq > 0)
+		irq_set_chained_handler_and_data(msir->gic_irq, NULL, NULL);
+
+	for (i = 0; i < msi_data->cfg->msir_irqs; i++) {
+		hwirq = i << msi_data->cfg->ibs_shift | msir->index;
+		bitmap_set(msi_data->used, hwirq, 1);
+	}
+
+	return 0;
+}
+
+static struct ls_scfg_msi_cfg ls1021_msi_cfg = {
+	.ibs_shift = 3,
+	.msir_irqs = MSI_IRQS_PER_MSIR,
+	.msir_base = MSI_MSIR_OFFSET,
+};
+
+static struct ls_scfg_msi_cfg ls1046_msi_cfg = {
+	.ibs_shift = 2,
+	.msir_irqs = MSI_IRQS_PER_MSIR,
+	.msir_base = MSI_MSIR_OFFSET,
+};
+
+static struct ls_scfg_msi_cfg ls1043_v1_1_msi_cfg = {
+	.ibs_shift = 2,
+	.msir_irqs = MSI_LS1043V1_1_IRQS_PER_MSIR,
+	.msir_base = MSI_LS1043V1_1_MSIR_OFFSET,
+};
+
+static const struct of_device_id ls_scfg_msi_id[] = {
+	/* The following two misspelled compatibles are obsolete */
+	{ .compatible = "fsl,1s1021a-msi", .data = &ls1021_msi_cfg},
+	{ .compatible = "fsl,1s1043a-msi", .data = &ls1021_msi_cfg},
+
+	{ .compatible = "fsl,ls1012a-msi", .data = &ls1021_msi_cfg },
+	{ .compatible = "fsl,ls1021a-msi", .data = &ls1021_msi_cfg },
+	{ .compatible = "fsl,ls1043a-msi", .data = &ls1021_msi_cfg },
+	{ .compatible = "fsl,ls1043a-v1.1-msi", .data = &ls1043_v1_1_msi_cfg },
+	{ .compatible = "fsl,ls1046a-msi", .data = &ls1046_msi_cfg },
+	{},
+};
+MODULE_DEVICE_TABLE(of, ls_scfg_msi_id);
+
 static int ls_scfg_msi_probe(struct platform_device *pdev)
 {
+	const struct of_device_id *match;
 	struct ls_scfg_msi *msi_data;
 	struct resource *res;
-	int ret;
+	int i, ret;
+
+	match = of_match_device(ls_scfg_msi_id, &pdev->dev);
+	if (!match)
+		return -ENODEV;
 
 	msi_data = devm_kzalloc(&pdev->dev, sizeof(*msi_data), GFP_KERNEL);
 	if (!msi_data)
 		return -ENOMEM;
 
+	msi_data->cfg = (struct ls_scfg_msi_cfg *) match->data;
+
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	msi_data->regs = devm_ioremap_resource(&pdev->dev, res);
 	if (IS_ERR(msi_data->regs)) {
@@ -182,23 +350,48 @@ static int ls_scfg_msi_probe(struct plat
 	}
 	msi_data->msiir_addr = res->start;
 
-	msi_data->irq = platform_get_irq(pdev, 0);
-	if (msi_data->irq <= 0) {
-		dev_err(&pdev->dev, "failed to get MSI irq\n");
-		return -ENODEV;
-	}
-
 	msi_data->pdev = pdev;
 	spin_lock_init(&msi_data->lock);
 
+	msi_data->irqs_num = MSI_IRQS_PER_MSIR *
+			     (1 << msi_data->cfg->ibs_shift);
+	msi_data->used = devm_kcalloc(&pdev->dev,
+				    BITS_TO_LONGS(msi_data->irqs_num),
+				    sizeof(*msi_data->used),
+				    GFP_KERNEL);
+	if (!msi_data->used)
+		return -ENOMEM;
+	/*
+	 * Reserve all the hwirqs
+	 * The available hwirqs will be released in ls1_msi_setup_hwirq()
+	 */
+	bitmap_set(msi_data->used, 0, msi_data->irqs_num);
+
+	msi_data->msir_num = of_irq_count(pdev->dev.of_node);
+
+	if (msi_affinity_flag) {
+		u32 cpu_num;
+
+		cpu_num = num_possible_cpus();
+		if (msi_data->msir_num >= cpu_num)
+			msi_data->msir_num = cpu_num;
+		else
+			msi_affinity_flag = 0;
+	}
+
+	msi_data->msir = devm_kcalloc(&pdev->dev, msi_data->msir_num,
+				      sizeof(*msi_data->msir),
+				      GFP_KERNEL);
+	if (!msi_data->msir)
+		return -ENOMEM;
+
+	for (i = 0; i < msi_data->msir_num; i++)
+		ls_scfg_msi_setup_hwirq(msi_data, i);
+
 	ret = ls_scfg_msi_domains_init(msi_data);
 	if (ret)
 		return ret;
 
-	irq_set_chained_handler_and_data(msi_data->irq,
-					 ls_scfg_msi_irq_handler,
-					 msi_data);
-
 	platform_set_drvdata(pdev, msi_data);
 
 	return 0;
@@ -207,8 +400,10 @@ static int ls_scfg_msi_probe(struct plat
 static int ls_scfg_msi_remove(struct platform_device *pdev)
 {
 	struct ls_scfg_msi *msi_data = platform_get_drvdata(pdev);
+	int i;
 
-	irq_set_chained_handler_and_data(msi_data->irq, NULL, NULL);
+	for (i = 0; i < msi_data->msir_num; i++)
+		ls_scfg_msi_teardown_hwirq(&msi_data->msir[i]);
 
 	irq_domain_remove(msi_data->msi_domain);
 	irq_domain_remove(msi_data->parent);
@@ -218,12 +413,6 @@ static int ls_scfg_msi_remove(struct pla
 	return 0;
 }
 
-static const struct of_device_id ls_scfg_msi_id[] = {
-	{ .compatible = "fsl,1s1021a-msi", },
-	{ .compatible = "fsl,1s1043a-msi", },
-	{},
-};
-
 static struct platform_driver ls_scfg_msi_driver = {
 	.driver = {
 		.name = "ls-scfg-msi",
--- a/drivers/pci/host/Makefile
+++ b/drivers/pci/host/Makefile
@@ -17,7 +17,7 @@ obj-$(CONFIG_PCIE_XILINX) += pcie-xilinx
 obj-$(CONFIG_PCIE_XILINX_NWL) += pcie-xilinx-nwl.o
 obj-$(CONFIG_PCI_XGENE) += pci-xgene.o
 obj-$(CONFIG_PCI_XGENE_MSI) += pci-xgene-msi.o
-obj-$(CONFIG_PCI_LAYERSCAPE) += pci-layerscape.o
+obj-$(CONFIG_PCI_LAYERSCAPE) += pci-layerscape.o pci-layerscape-ep.o pci-layerscape-ep-debugfs.o
 obj-$(CONFIG_PCI_VERSATILE) += pci-versatile.o
 obj-$(CONFIG_PCIE_IPROC) += pcie-iproc.o
 obj-$(CONFIG_PCIE_IPROC_MSI) += pcie-iproc-msi.o
--- /dev/null
+++ b/drivers/pci/host/pci-layerscape-ep-debugfs.c
@@ -0,0 +1,758 @@
+/*
+ * PCIe Endpoint driver for Freescale Layerscape SoCs
+ *
+ * Copyright (C) 2015 Freescale Semiconductor.
+ *
+  * Author: Minghuan Lian <Minghuan.Lian@freescale.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/debugfs.h>
+#include <linux/time.h>
+#include <linux/uaccess.h>
+#include <linux/kthread.h>
+#include <linux/slab.h>
+#include <linux/dmaengine.h>
+#include <linux/dma-mapping.h>
+#include <linux/freezer.h>
+
+#include <linux/completion.h>
+
+#include "pci-layerscape-ep.h"
+
+#define PCIE_ATU_INDEX3		(0x3 << 0)
+#define PCIE_ATU_INDEX2		(0x2 << 0)
+#define PCIE_ATU_INDEX1		(0x1 << 0)
+#define PCIE_ATU_INDEX0		(0x0 << 0)
+
+#define PCIE_BAR0_SIZE		(4 * 1024) /* 4K */
+#define PCIE_BAR1_SIZE		(8 * 1024) /* 8K for MSIX */
+#define PCIE_BAR2_SIZE		(4 * 1024) /* 4K */
+#define PCIE_BAR4_SIZE		(1 * 1024 * 1024) /* 1M */
+#define PCIE_MSI_OB_SIZE	(4 * 1024) /* 4K */
+
+#define PCIE_MSI_MSG_ADDR_OFF	0x54
+#define PCIE_MSI_MSG_DATA_OFF	0x5c
+
+enum test_type {
+	TEST_TYPE_DMA,
+	TEST_TYPE_MEMCPY
+};
+
+enum test_dirt {
+	TEST_DIRT_READ,
+	TEST_DIRT_WRITE
+};
+
+enum test_status {
+	TEST_IDLE,
+	TEST_BUSY
+};
+
+struct ls_ep_test {
+	struct ls_ep_dev	*ep;
+	void __iomem		*cfg;
+	void __iomem		*buf;
+	void __iomem		*out;
+	void __iomem		*msi;
+	dma_addr_t		cfg_addr;
+	dma_addr_t		buf_addr;
+	dma_addr_t		out_addr;
+	dma_addr_t		bus_addr;
+	dma_addr_t		msi_addr;
+	u64			msi_msg_addr;
+	u16			msi_msg_data;
+	struct task_struct	*thread;
+	spinlock_t		lock;
+	struct completion	done;
+	u32			len;
+	int			loop;
+	char			data;
+	enum test_dirt		dirt;
+	enum test_type		type;
+	enum test_status	status;
+	u64			result; /* Mbps */
+	char			cmd[256];
+};
+
+static int ls_pcie_ep_trigger_msi(struct ls_ep_test *test)
+{
+	if (!test->msi)
+		return -EINVAL;
+
+	iowrite32(test->msi_msg_data, test->msi);
+
+	return 0;
+}
+
+static int ls_pcie_ep_test_try_run(struct ls_ep_test *test)
+{
+	int ret;
+
+	spin_lock(&test->lock);
+	if (test->status == TEST_IDLE) {
+		test->status = TEST_BUSY;
+		ret = 0;
+	} else
+		ret = -EBUSY;
+	spin_unlock(&test->lock);
+
+	return ret;
+}
+
+static void ls_pcie_ep_test_done(struct ls_ep_test *test)
+{
+	spin_lock(&test->lock);
+	test->status = TEST_IDLE;
+	spin_unlock(&test->lock);
+}
+
+static void ls_pcie_ep_test_dma_cb(void *arg)
+{
+	struct ls_ep_test *test = arg;
+
+	complete(&test->done);
+}
+
+static int ls_pcie_ep_test_dma(struct ls_ep_test *test)
+{
+	dma_cap_mask_t mask;
+	struct dma_chan *chan;
+	struct dma_device *dma_dev;
+	dma_addr_t src, dst;
+	enum dma_data_direction direction;
+	enum dma_ctrl_flags dma_flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+	struct timespec start, end, period;
+	int i = 0;
+
+	dma_cap_zero(mask);
+	dma_cap_set(DMA_MEMCPY, mask);
+
+	chan = dma_request_channel(mask, NULL, test);
+	if (!chan) {
+		pr_err("failed to request dma channel\n");
+		return -EINVAL;
+	}
+
+	memset(test->buf, test->data, test->len);
+
+	if (test->dirt == TEST_DIRT_WRITE) {
+		src = test->buf_addr;
+		dst = test->out_addr;
+		direction = DMA_TO_DEVICE;
+	} else {
+		src = test->out_addr;
+		dst = test->buf_addr;
+		direction = DMA_FROM_DEVICE;
+	}
+
+	dma_dev = chan->device;
+	dma_flags = DMA_CTRL_ACK | DMA_PREP_INTERRUPT;
+
+	dma_sync_single_for_device(&test->ep->dev, test->buf_addr,
+				   test->len, direction);
+
+	set_freezable();
+
+	getrawmonotonic(&start);
+	while (!kthread_should_stop() && (i < test->loop)) {
+		struct dma_async_tx_descriptor *dma_desc;
+		dma_cookie_t	dma_cookie = {0};
+		unsigned long tmo;
+		int status;
+
+		init_completion(&test->done);
+
+		dma_desc = dma_dev->device_prep_dma_memcpy(chan,
+							   dst, src,
+							   test->len,
+							   dma_flags);
+		if (!dma_desc) {
+			pr_err("DMA desc constr failed...\n");
+			goto _err;
+		}
+
+		dma_desc->callback = ls_pcie_ep_test_dma_cb;
+		dma_desc->callback_param = test;
+		dma_cookie = dmaengine_submit(dma_desc);
+
+		if (dma_submit_error(dma_cookie)) {
+			pr_err("DMA submit error....\n");
+			goto _err;
+		}
+
+		/* Trigger the transaction */
+		dma_async_issue_pending(chan);
+
+		tmo = wait_for_completion_timeout(&test->done,
+					  msecs_to_jiffies(5 * test->len));
+		if (tmo == 0) {
+			pr_err("Self-test copy timed out, disabling\n");
+			goto _err;
+		}
+
+		status = dma_async_is_tx_complete(chan, dma_cookie,
+						  NULL, NULL);
+		if (status != DMA_COMPLETE) {
+			pr_err("got completion callback, but status is %s\n",
+			       status == DMA_ERROR ? "error" : "in progress");
+			goto _err;
+		}
+
+		i++;
+	}
+
+	getrawmonotonic(&end);
+	period = timespec_sub(end, start);
+	test->result = test->len * 8ULL * i * 1000;
+	do_div(test->result, period.tv_sec * 1000 * 1000 * 1000 + period.tv_nsec);
+	dma_release_channel(chan);
+
+	return 0;
+
+_err:
+	dma_release_channel(chan);
+	test->result = 0;
+	return -EINVAL;
+}
+
+static int ls_pcie_ep_test_cpy(struct ls_ep_test *test)
+{
+	void *dst, *src;
+	struct timespec start, end, period;
+	int i = 0;
+
+	memset(test->buf, test->data, test->len);
+
+	if (test->dirt == TEST_DIRT_WRITE) {
+		dst = test->out;
+		src = test->buf;
+	} else {
+		dst = test->buf;
+		src = test->out;
+	}
+
+	getrawmonotonic(&start);
+	while (!kthread_should_stop() && i < test->loop) {
+		memcpy(dst, src, test->len);
+		i++;
+	}
+	getrawmonotonic(&end);
+
+	period = timespec_sub(end, start);
+	test->result = test->len * 8ULL * i * 1000;
+	do_div(test->result, period.tv_sec * 1000 * 1000 * 1000 + period.tv_nsec);
+
+	return 0;
+}
+
+int ls_pcie_ep_test_thread(void *arg)
+{
+	int ret;
+
+	struct ls_ep_test *test = arg;
+
+	if (test->type == TEST_TYPE_DMA)
+		ret = ls_pcie_ep_test_dma(test);
+	else
+		ret = ls_pcie_ep_test_cpy(test);
+
+	if (ret) {
+		pr_err("\n%s \ttest failed\n",
+		       test->cmd);
+		test->result = 0;
+	} else
+		pr_err("\n%s \tthroughput:%lluMbps\n",
+		       test->cmd, test->result);
+
+	ls_pcie_ep_test_done(test);
+
+	ls_pcie_ep_trigger_msi(test);
+
+	do_exit(0);
+}
+
+static int ls_pcie_ep_free_test(struct ls_ep_dev *ep)
+{
+	struct ls_ep_test *test = ep->driver_data;
+
+	if (!test)
+		return 0;
+
+	if (test->status == TEST_BUSY) {
+		kthread_stop(test->thread);
+		dev_info(&ep->dev,
+			 "test is running please wait and run again\n");
+		return -EBUSY;
+	}
+
+	if (test->buf)
+		free_pages((unsigned long)test->buf,
+			   get_order(PCIE_BAR4_SIZE));
+
+	if (test->cfg)
+		free_pages((unsigned long)test->cfg,
+			   get_order(PCIE_BAR2_SIZE));
+
+	if (test->out)
+		iounmap(test->out);
+
+	kfree(test);
+	ep->driver_data = NULL;
+
+	return 0;
+}
+
+static int ls_pcie_ep_init_test(struct ls_ep_dev *ep, u64 bus_addr)
+{
+	struct ls_pcie *pcie = ep->pcie;
+	struct ls_ep_test *test = ep->driver_data;
+	int err;
+
+	if (test) {
+		dev_info(&ep->dev,
+			 "Please use 'free' to remove the exiting test\n");
+		return -EBUSY;
+	}
+
+	test = kzalloc(sizeof(*test), GFP_KERNEL);
+	if (!test)
+		return -ENOMEM;
+	ep->driver_data = test;
+	test->ep = ep;
+	spin_lock_init(&test->lock);
+	test->status = TEST_IDLE;
+
+	test->buf = dma_alloc_coherent(pcie->dev, get_order(PCIE_BAR4_SIZE),
+					&test->buf_addr,
+					GFP_KERNEL);
+	if (!test->buf) {
+		dev_info(&ep->dev, "failed to get mem for bar4\n");
+		err = -ENOMEM;
+		goto _err;
+	}
+
+	test->cfg = (void *)__get_free_pages(GFP_KERNEL | __GFP_ZERO,
+					     get_order(PCIE_BAR2_SIZE));
+	if (!test->cfg) {
+		dev_info(&ep->dev, "failed to get mem for bar4\n");
+		err = -ENOMEM;
+		goto _err;
+	}
+	test->cfg_addr = virt_to_phys(test->cfg);
+
+	test->out_addr = pcie->out_base;
+	test->out = ioremap(test->out_addr, PCIE_BAR4_SIZE);
+	if (!test->out) {
+		dev_info(&ep->dev, "failed to map out\n");
+		err = -ENOMEM;
+		goto _err;
+	}
+
+	test->bus_addr = bus_addr;
+
+	test->msi_addr = test->out_addr + PCIE_BAR4_SIZE;
+	test->msi = ioremap(test->msi_addr, PCIE_MSI_OB_SIZE);
+	if (!test->msi)
+		dev_info(&ep->dev, "failed to map MSI outbound region\n");
+
+	test->msi_msg_addr = ioread32(pcie->dbi + PCIE_MSI_MSG_ADDR_OFF) |
+		(((u64)ioread32(pcie->dbi + PCIE_MSI_MSG_ADDR_OFF + 4)) << 32);
+	test->msi_msg_data = ioread16(pcie->dbi + PCIE_MSI_MSG_DATA_OFF);
+
+	ls_pcie_ep_dev_cfg_enable(ep);
+
+	/* outbound iATU for memory */
+	ls_pcie_iatu_outbound_set(pcie, 0, PCIE_ATU_TYPE_MEM,
+				  test->out_addr, bus_addr, PCIE_BAR4_SIZE);
+	/* outbound iATU for MSI */
+	ls_pcie_iatu_outbound_set(pcie, 1, PCIE_ATU_TYPE_MEM,
+				  test->msi_addr, test->msi_msg_addr,
+				  PCIE_MSI_OB_SIZE);
+
+	/* ATU 0 : INBOUND : map BAR0 */
+	ls_pcie_iatu_inbound_set(pcie, 0, 0, test->cfg_addr);
+	/* ATU 2 : INBOUND : map BAR2 */
+	ls_pcie_iatu_inbound_set(pcie, 2, 2, test->cfg_addr);
+	/* ATU 3 : INBOUND : map BAR4 */
+	ls_pcie_iatu_inbound_set(pcie, 3, 4, test->buf_addr);
+
+	return 0;
+
+_err:
+	ls_pcie_ep_free_test(ep);
+	return err;
+}
+
+static int ls_pcie_ep_start_test(struct ls_ep_dev *ep, char *cmd)
+{
+	struct ls_ep_test *test = ep->driver_data;
+	enum test_type type;
+	enum test_dirt dirt;
+	u32 cnt, len, loop;
+	unsigned int data;
+	char dirt_str[2];
+	int ret;
+
+	if (strncmp(cmd, "dma", 3) == 0)
+		type = TEST_TYPE_DMA;
+	else
+		type = TEST_TYPE_MEMCPY;
+
+	cnt = sscanf(&cmd[4], "%1s %u %u %x", dirt_str, &len, &loop, &data);
+	if (cnt != 4) {
+		dev_info(&ep->dev, "format error %s", cmd);
+		dev_info(&ep->dev, "dma/cpy <r/w> <packet_size> <loop> <data>\n");
+		return -EINVAL;
+	}
+
+	if (strncmp(dirt_str, "r", 1) == 0)
+		dirt = TEST_DIRT_READ;
+	else
+		dirt = TEST_DIRT_WRITE;
+
+	if (len > PCIE_BAR4_SIZE) {
+		dev_err(&ep->dev, "max len is %d", PCIE_BAR4_SIZE);
+		return -EINVAL;
+	}
+
+	if (!test) {
+		dev_err(&ep->dev, "Please first run init command\n");
+		return -EINVAL;
+	}
+
+	if (ls_pcie_ep_test_try_run(test)) {
+		dev_err(&ep->dev, "There is already a test running\n");
+		return -EINVAL;
+	}
+
+	test->len = len;
+	test->loop = loop;
+	test->type = type;
+	test->data = (char)data;
+	test->dirt = dirt;
+	strcpy(test->cmd, cmd);
+	test->thread = kthread_run(ls_pcie_ep_test_thread, test,
+				   "pcie ep test");
+	if (IS_ERR(test->thread)) {
+		dev_err(&ep->dev, "fork failed for pcie ep test\n");
+		ls_pcie_ep_test_done(test);
+		ret = PTR_ERR(test->thread);
+	}
+
+	return ret;
+}
+
+
+/**
+ * ls_pcie_reg_ops_read - read for regs data
+ * @filp: the opened file
+ * @buffer: where to write the data for the user to read
+ * @count: the size of the user's buffer
+ * @ppos: file position offset
+ **/
+static ssize_t ls_pcie_ep_dbg_regs_read(struct file *filp, char __user *buffer,
+				    size_t count, loff_t *ppos)
+{
+	struct ls_ep_dev *ep = filp->private_data;
+	struct ls_pcie *pcie = ep->pcie;
+	char *buf;
+	int desc = 0, i, len;
+
+	buf = kmalloc(4 * 1024, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	ls_pcie_ep_dev_cfg_enable(ep);
+
+	desc += sprintf(buf + desc, "%s", "reg info:");
+	for (i = 0; i < 0x200; i += 4) {
+		if (i % 16 == 0)
+			desc += sprintf(buf + desc, "\n%08x:", i);
+		desc += sprintf(buf + desc, " %08x", readl(pcie->dbi + i));
+	}
+
+	desc += sprintf(buf + desc, "\n%s", "outbound iATU info:\n");
+	for (i = 0; i < 6; i++) {
+		writel(PCIE_ATU_REGION_OUTBOUND | i,
+		       pcie->dbi + PCIE_ATU_VIEWPORT);
+		desc += sprintf(buf + desc, "iATU%d", i);
+		desc += sprintf(buf + desc, "\tLOWER PHYS 0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LOWER_BASE));
+		desc += sprintf(buf + desc, "\tUPPER PHYS 0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_UPPER_BASE));
+		desc += sprintf(buf + desc, "\tLOWER BUS  0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LOWER_TARGET));
+		desc += sprintf(buf + desc, "\tUPPER BUS  0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_UPPER_TARGET));
+		desc += sprintf(buf + desc, "\tLIMIT      0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LIMIT));
+		desc += sprintf(buf + desc, "\tCR1        0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_CR1));
+		desc += sprintf(buf + desc, "\tCR2        0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_CR2));
+	}
+
+	desc += sprintf(buf + desc, "\n%s", "inbound iATU info:\n");
+	for (i = 0; i < 6; i++) {
+		writel(PCIE_ATU_REGION_INBOUND | i,
+		       pcie->dbi + PCIE_ATU_VIEWPORT);
+		desc += sprintf(buf + desc, "iATU%d", i);
+		desc += sprintf(buf + desc, "\tLOWER BUS  0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LOWER_BASE));
+		desc += sprintf(buf + desc, "\tUPPER BUSs 0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_UPPER_BASE));
+		desc += sprintf(buf + desc, "\tLOWER PHYS 0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LOWER_TARGET));
+		desc += sprintf(buf + desc, "\tUPPER PHYS 0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_UPPER_TARGET));
+		desc += sprintf(buf + desc, "\tLIMIT      0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_LIMIT));
+		desc += sprintf(buf + desc, "\tCR1        0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_CR1));
+		desc += sprintf(buf + desc, "\tCR2        0x%08x\n",
+		      readl(pcie->dbi + PCIE_ATU_CR2));
+	}
+
+	len = simple_read_from_buffer(buffer, count, ppos, buf, desc);
+	kfree(buf);
+
+	return len;
+}
+
+/**
+ * ls_pcie_ep_dbg_regs_write - write into regs datum
+ * @filp: the opened file
+ * @buffer: where to find the user's data
+ * @count: the length of the user's data
+ * @ppos: file position offset
+ **/
+static ssize_t ls_pcie_ep_dbg_regs_write(struct file *filp,
+					 const char __user *buffer,
+					 size_t count, loff_t *ppos)
+{
+	struct ls_ep_dev *ep = filp->private_data;
+	struct ls_pcie *pcie = ep->pcie;
+	char buf[256];
+
+	if (count >= sizeof(buf))
+		return -ENOSPC;
+
+	memset(buf, 0, sizeof(buf));
+
+	if (copy_from_user(buf, buffer, count))
+		return -EFAULT;
+
+	ls_pcie_ep_dev_cfg_enable(ep);
+
+	if (strncmp(buf, "reg", 3) == 0) {
+		u32 reg, value;
+		int cnt;
+
+		cnt = sscanf(&buf[3], "%x %x", &reg, &value);
+		if (cnt == 2) {
+			writel(value, pcie->dbi + reg);
+			value = readl(pcie->dbi + reg);
+			dev_info(&ep->dev, "reg 0x%08x: 0x%08x\n",
+				 reg, value);
+		} else {
+			dev_info(&ep->dev, "reg <reg> <value>\n");
+		}
+	} else if (strncmp(buf, "atu", 3) == 0) {
+		/* to do */
+		dev_info(&ep->dev, " Not support atu command\n");
+	} else {
+		dev_info(&ep->dev, "Unknown command %s\n", buf);
+		dev_info(&ep->dev, "Available commands:\n");
+		dev_info(&ep->dev, "   reg <reg> <value>\n");
+	}
+
+	return count;
+}
+
+static const struct file_operations ls_pcie_ep_dbg_regs_fops = {
+	.owner = THIS_MODULE,
+	.open = simple_open,
+	.read =  ls_pcie_ep_dbg_regs_read,
+	.write = ls_pcie_ep_dbg_regs_write,
+};
+
+static ssize_t ls_pcie_ep_dbg_test_read(struct file *filp,
+				   char __user *buffer,
+				   size_t count, loff_t *ppos)
+{
+	struct ls_ep_dev *ep = filp->private_data;
+	struct ls_ep_test *test = ep->driver_data;
+	char buf[512];
+	int desc = 0, len;
+
+	if (!test) {
+		dev_info(&ep->dev, " there is NO test\n");
+		return 0;
+	}
+
+	if (test->status != TEST_IDLE) {
+		dev_info(&ep->dev, "test %s is running\n", test->cmd);
+		return 0;
+	}
+
+	desc = sprintf(buf, "MSI ADDR:0x%llx MSI DATA:0x%x\n",
+		test->msi_msg_addr, test->msi_msg_data);
+
+	desc += sprintf(buf + desc, "%s throughput:%lluMbps\n",
+			test->cmd, test->result);
+
+	len = simple_read_from_buffer(buffer, count, ppos,
+				      buf, desc);
+
+	return len;
+}
+
+static ssize_t ls_pcie_ep_dbg_test_write(struct file *filp,
+					const char __user *buffer,
+					size_t count, loff_t *ppos)
+{
+	struct ls_ep_dev *ep = filp->private_data;
+	char buf[256];
+
+	if (count >= sizeof(buf))
+		return -ENOSPC;
+
+	memset(buf, 0, sizeof(buf));
+
+	if (copy_from_user(buf, buffer, count))
+		return -EFAULT;
+
+	if (strncmp(buf, "init", 4) == 0) {
+		int i = 4;
+		u64 bus_addr;
+
+		while (buf[i] == ' ')
+			i++;
+
+		if (kstrtou64(&buf[i], 0, &bus_addr))
+			dev_info(&ep->dev, "command: init <bus_addr>\n");
+		else {
+			if (ls_pcie_ep_init_test(ep, bus_addr))
+				dev_info(&ep->dev, "failed to init test\n");
+		}
+	} else if (strncmp(buf, "free", 4) == 0)
+		ls_pcie_ep_free_test(ep);
+	else if (strncmp(buf, "dma", 3) == 0 ||
+		 strncmp(buf, "cpy", 3) == 0)
+		ls_pcie_ep_start_test(ep, buf);
+	else {
+		dev_info(&ep->dev, "Unknown command: %s\n", buf);
+		dev_info(&ep->dev, "Available commands:\n");
+		dev_info(&ep->dev, "\tinit <bus_addr>\n");
+		dev_info(&ep->dev, "\t<dma/cpy> <r/w> <packet_size> <loop>\n");
+		dev_info(&ep->dev, "\tfree\n");
+	}
+
+	return count;
+}
+
+static const struct file_operations ls_pcie_ep_dbg_test_fops = {
+	.owner = THIS_MODULE,
+	.open = simple_open,
+	.read = ls_pcie_ep_dbg_test_read,
+	.write = ls_pcie_ep_dbg_test_write,
+};
+
+static ssize_t ls_pcie_ep_dbg_dump_read(struct file *filp,
+				   char __user *buffer,
+				   size_t count, loff_t *ppos)
+{
+	struct ls_ep_dev *ep = filp->private_data;
+	struct ls_ep_test *test = ep->driver_data;
+	char *buf;
+	int desc = 0, i, len;
+
+	buf = kmalloc(4 * 1024, GFP_KERNEL);
+	if (!buf)
+		return -ENOMEM;
+
+	if (!test) {
+		dev_info(&ep->dev, " there is NO test\n");
+		kfree(buf);
+		return 0;
+	}
+
+	desc += sprintf(buf + desc, "%s", "dump info:");
+	for (i = 0; i < 256; i += 4) {
+		if (i % 16 == 0)
+			desc += sprintf(buf + desc, "\n%08x:", i);
+		desc += sprintf(buf + desc, " %08x", readl(test->buf + i));
+	}
+
+	desc += sprintf(buf + desc, "\n");
+	len = simple_read_from_buffer(buffer, count, ppos, buf, desc);
+
+	kfree(buf);
+
+	return len;
+}
+
+static const struct file_operations ls_pcie_ep_dbg_dump_fops = {
+	.owner = THIS_MODULE,
+	.open = simple_open,
+	.read = ls_pcie_ep_dbg_dump_read,
+};
+
+static int ls_pcie_ep_dev_dbgfs_init(struct ls_ep_dev *ep)
+{
+	struct ls_pcie *pcie = ep->pcie;
+	struct dentry *pfile;
+
+	ls_pcie_ep_dev_cfg_enable(ep);
+
+	ep->dir = debugfs_create_dir(dev_name(&ep->dev), pcie->dir);
+	if (!ep->dir)
+		return -ENOMEM;
+
+	pfile = debugfs_create_file("regs", 0600, ep->dir, ep,
+				    &ls_pcie_ep_dbg_regs_fops);
+	if (!pfile)
+		dev_info(&ep->dev, "debugfs regs for failed\n");
+
+	pfile = debugfs_create_file("test", 0600, ep->dir, ep,
+				    &ls_pcie_ep_dbg_test_fops);
+	if (!pfile)
+		dev_info(&ep->dev, "debugfs test for failed\n");
+
+	pfile = debugfs_create_file("dump", 0600, ep->dir, ep,
+				    &ls_pcie_ep_dbg_dump_fops);
+	if (!pfile)
+		dev_info(&ep->dev, "debugfs dump for failed\n");
+
+	return 0;
+}
+
+int ls_pcie_ep_dbgfs_init(struct ls_pcie *pcie)
+{
+	struct ls_ep_dev *ep;
+
+	pcie->dir = debugfs_create_dir(dev_name(pcie->dev), NULL);
+	if (!pcie->dir)
+		return -ENOMEM;
+
+	list_for_each_entry(ep, &pcie->ep_list, node)
+		ls_pcie_ep_dev_dbgfs_init(ep);
+
+	return 0;
+}
+
+int ls_pcie_ep_dbgfs_remove(struct ls_pcie *pcie)
+{
+	debugfs_remove_recursive(pcie->dir);
+	return 0;
+}
+
+MODULE_AUTHOR("Minghuan Lian <Minghuan.Lian@freescale.com>");
+MODULE_DESCRIPTION("Freescale Layerscape PCIe EP controller driver");
+MODULE_LICENSE("GPL v2");
--- /dev/null
+++ b/drivers/pci/host/pci-layerscape-ep.c
@@ -0,0 +1,309 @@
+/*
+ * PCIe Endpoint driver for Freescale Layerscape SoCs
+ *
+ * Copyright (C) 2015 Freescale Semiconductor.
+ *
+  * Author: Minghuan Lian <Minghuan.Lian@freescale.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+#include <linux/kernel.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/module.h>
+#include <linux/of_pci.h>
+#include <linux/of_platform.h>
+#include <linux/of_irq.h>
+#include <linux/of_address.h>
+#include <linux/pci.h>
+#include <linux/platform_device.h>
+#include <linux/resource.h>
+#include <linux/debugfs.h>
+#include <linux/time.h>
+#include <linux/uaccess.h>
+
+#include "pci-layerscape-ep.h"
+
+struct ls_ep_dev *
+ls_pci_ep_find(struct ls_pcie *pcie, int dev_id)
+{
+	struct ls_ep_dev *ep;
+
+	list_for_each_entry(ep, &pcie->ep_list, node) {
+		if (ep->dev_id == dev_id)
+			return ep;
+	}
+
+	return NULL;
+}
+
+static void ls_pcie_try_cfg2(struct ls_pcie *pcie, int pf, int vf)
+{
+	if (pcie->sriov)
+		writel(PCIE_LCTRL0_VAL(pf, vf),
+		       pcie->dbi + PCIE_LUT_BASE + PCIE_LUT_LCTRL0);
+}
+
+static bool ls_pcie_is_bridge(struct ls_pcie *pcie)
+{
+	u32 header_type = 0;
+
+	header_type = readl(pcie->dbi + (PCI_HEADER_TYPE & ~0x3));
+	header_type = (header_type >> 16) & 0x7f;
+
+	return header_type == PCI_HEADER_TYPE_BRIDGE;
+}
+
+void ls_pcie_iatu_outbound_set(struct ls_pcie *pcie, int idx, int type,
+			       u64 cpu_addr, u64 pci_addr, u32 size)
+{
+	writel(PCIE_ATU_REGION_OUTBOUND | idx,
+	       pcie->dbi + PCIE_ATU_VIEWPORT);
+	writel(lower_32_bits(cpu_addr),
+	       pcie->dbi +  PCIE_ATU_LOWER_BASE);
+	writel(upper_32_bits(cpu_addr),
+	       pcie->dbi + PCIE_ATU_UPPER_BASE);
+	writel(lower_32_bits(cpu_addr + size - 1),
+	       pcie->dbi + PCIE_ATU_LIMIT);
+	writel(lower_32_bits(pci_addr),
+	       pcie->dbi + PCIE_ATU_LOWER_TARGET);
+	writel(upper_32_bits(pci_addr),
+	       pcie->dbi + PCIE_ATU_UPPER_TARGET);
+	writel(type, pcie->dbi + PCIE_ATU_CR1);
+	writel(PCIE_ATU_ENABLE, pcie->dbi + PCIE_ATU_CR2);
+}
+
+/* Use bar match mode and MEM type as default */
+void ls_pcie_iatu_inbound_set(struct ls_pcie *pcie, int idx,
+				     int bar, u64 phys)
+{
+	writel(PCIE_ATU_REGION_INBOUND | idx, pcie->dbi + PCIE_ATU_VIEWPORT);
+	writel((u32)phys, pcie->dbi + PCIE_ATU_LOWER_TARGET);
+	writel(phys >> 32, pcie->dbi + PCIE_ATU_UPPER_TARGET);
+	writel(PCIE_ATU_TYPE_MEM, pcie->dbi + PCIE_ATU_CR1);
+	writel(PCIE_ATU_ENABLE | PCIE_ATU_BAR_MODE_ENABLE |
+	       PCIE_ATU_BAR_NUM(bar), pcie->dbi + PCIE_ATU_CR2);
+}
+
+void ls_pcie_ep_dev_cfg_enable(struct ls_ep_dev *ep)
+{
+	ls_pcie_try_cfg2(ep->pcie, ep->pf_idx, ep->vf_idx);
+}
+
+void ls_pcie_ep_setup_bar(void *bar_base, int bar, u32 size)
+{
+	if (size < 4 * 1024)
+		return;
+
+	switch (bar) {
+	case 0:
+		writel(size - 1, bar_base + PCI_BASE_ADDRESS_0);
+		break;
+	case 1:
+		writel(size - 1, bar_base + PCI_BASE_ADDRESS_1);
+		break;
+	case 2:
+		writel(size - 1, bar_base + PCI_BASE_ADDRESS_2);
+		writel(0, bar_base + PCI_BASE_ADDRESS_3);
+		break;
+	case 4:
+		writel(size - 1, bar_base + PCI_BASE_ADDRESS_4);
+		writel(0, bar_base + PCI_BASE_ADDRESS_5);
+		break;
+	default:
+		break;
+	}
+}
+
+void ls_pcie_ep_dev_setup_bar(struct ls_ep_dev *ep, int bar, u32 size)
+{
+	struct ls_pcie *pcie = ep->pcie;
+	void *bar_base;
+
+	if (size < 4 * 1024)
+		return;
+
+	if (pcie->sriov)
+		bar_base = pcie->dbi;
+	else
+		bar_base = pcie->dbi + PCIE_NO_SRIOV_BAR_BASE;
+
+	ls_pcie_ep_dev_cfg_enable(ep);
+	ls_pcie_ep_setup_bar(bar_base, bar, size);
+}
+
+static int ls_pcie_ep_dev_init(struct ls_pcie *pcie, int pf_idx, int vf_idx)
+{
+	struct ls_ep_dev *ep;
+
+	ep = devm_kzalloc(pcie->dev, sizeof(*ep), GFP_KERNEL);
+	if (!ep)
+		return -ENOMEM;
+
+	ep->pcie = pcie;
+	ep->pf_idx = pf_idx;
+	ep->vf_idx = vf_idx;
+	if (vf_idx)
+		ep->dev_id = pf_idx + 4 + 4 * (vf_idx - 1);
+	else
+		ep->dev_id = pf_idx;
+
+	if (ep->vf_idx)
+		dev_set_name(&ep->dev, "pf%d-vf%d",
+			     ep->pf_idx,
+			     ep->vf_idx);
+	else
+		dev_set_name(&ep->dev, "pf%d",
+			     ep->pf_idx);
+
+	list_add_tail(&ep->node, &pcie->ep_list);
+
+	return 0;
+}
+
+static int ls_pcie_ep_init(struct ls_pcie *pcie)
+{
+	u32 sriov_header;
+	int pf, vf, i, j;
+
+	sriov_header = readl(pcie->dbi + PCIE_SRIOV_POS);
+
+	if (PCI_EXT_CAP_ID(sriov_header) == PCI_EXT_CAP_ID_SRIOV) {
+		pcie->sriov = PCIE_SRIOV_POS;
+		pf = PCIE_PF_NUM;
+		vf = PCIE_VF_NUM;
+	} else {
+		pcie->sriov = 0;
+		pf = 1;
+		vf = 0;
+	}
+
+	for (i = 0; i < pf; i++) {
+		for (j = 0; j <= vf; j++)
+			ls_pcie_ep_dev_init(pcie, i, j);
+	}
+
+	return 0;
+}
+
+static struct ls_pcie_ep_drvdata ls1043_drvdata = {
+	.lut_offset = 0x10000,
+	.ltssm_shift = 24,
+	.lut_dbg = 0x7fc,
+};
+
+static struct ls_pcie_ep_drvdata ls1046_drvdata = {
+	.lut_offset = 0x80000,
+	.ltssm_shift = 24,
+	.lut_dbg = 0x407fc,
+};
+
+static struct ls_pcie_ep_drvdata ls2080_drvdata = {
+	.lut_offset = 0x80000,
+	.ltssm_shift = 0,
+	.lut_dbg = 0x7fc,
+};
+
+static const struct of_device_id ls_pcie_ep_of_match[] = {
+	{ .compatible = "fsl,ls1021a-pcie", },
+	{ .compatible = "fsl,ls1043a-pcie", .data = &ls1043_drvdata },
+	{ .compatible = "fsl,ls1046a-pcie", .data = &ls1046_drvdata },
+	{ .compatible = "fsl,ls2080a-pcie", .data = &ls2080_drvdata },
+	{ .compatible = "fsl,ls2085a-pcie", .data = &ls2080_drvdata },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, ls_pcie_ep_of_match);
+
+static int ls_pcie_ep_probe(struct platform_device *pdev)
+{
+	struct ls_pcie *pcie;
+	struct resource *dbi_base, *cfg_res;
+	const struct of_device_id *match;
+	int ret;
+
+	match = of_match_device(ls_pcie_ep_of_match, &pdev->dev);
+	if (!match)
+		return -ENODEV;
+
+	pcie = devm_kzalloc(&pdev->dev, sizeof(*pcie), GFP_KERNEL);
+	if (!pcie)
+		return -ENOMEM;
+
+	pcie->dev = &pdev->dev;
+	INIT_LIST_HEAD(&pcie->ep_list);
+
+	dbi_base = platform_get_resource_byname(pdev, IORESOURCE_MEM, "regs");
+	pcie->dbi = devm_ioremap_resource(&pdev->dev, dbi_base);
+	if (IS_ERR(pcie->dbi)) {
+		dev_err(&pdev->dev, "missing *regs* space\n");
+		return PTR_ERR(pcie->dbi);
+	}
+
+	pcie->drvdata = match->data;
+	pcie->lut = pcie->dbi + pcie->drvdata->lut_offset;
+
+	if (ls_pcie_is_bridge(pcie))
+		return -ENODEV;
+
+	dev_info(pcie->dev, "in EP mode\n");
+
+	cfg_res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "config");
+	if (cfg_res)
+		pcie->out_base = cfg_res->start;
+	else {
+		dev_err(&pdev->dev, "missing *config* space\n");
+		return -ENODEV;
+	}
+
+	ret = ls_pcie_ep_init(pcie);
+	if (ret)
+		return ret;
+
+	ls_pcie_ep_dbgfs_init(pcie);
+
+	platform_set_drvdata(pdev, pcie);
+
+	return 0;
+}
+
+static int ls_pcie_ep_dev_remove(struct ls_ep_dev *ep)
+{
+	list_del(&ep->node);
+
+	return 0;
+}
+
+static int ls_pcie_ep_remove(struct platform_device *pdev)
+{
+	struct ls_pcie *pcie = platform_get_drvdata(pdev);
+	struct ls_ep_dev *ep, *tmp;
+
+	if (!pcie)
+		return 0;
+
+	ls_pcie_ep_dbgfs_remove(pcie);
+
+	list_for_each_entry_safe(ep, tmp, &pcie->ep_list, node)
+		ls_pcie_ep_dev_remove(ep);
+
+	return 0;
+}
+
+static struct platform_driver ls_pcie_ep_driver = {
+	.driver = {
+		.name = "ls-pcie-ep",
+		.owner = THIS_MODULE,
+		.of_match_table = ls_pcie_ep_of_match,
+	},
+	.probe = ls_pcie_ep_probe,
+	.remove = ls_pcie_ep_remove,
+};
+
+module_platform_driver(ls_pcie_ep_driver);
+
+MODULE_AUTHOR("Minghuan Lian <Minghuan.Lian@freescale.com>");
+MODULE_DESCRIPTION("Freescale Layerscape PCIe EP driver");
+MODULE_LICENSE("GPL v2");
--- /dev/null
+++ b/drivers/pci/host/pci-layerscape-ep.h
@@ -0,0 +1,115 @@
+/*
+ * PCIe Endpoint driver for Freescale Layerscape SoCs
+ *
+ * Copyright (C) 2015 Freescale Semiconductor.
+ *
+  * Author: Minghuan Lian <Minghuan.Lian@freescale.com>
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */
+
+
+#ifndef _PCIE_LAYERSCAPE_EP_H
+#define _PCIE_LAYERSCAPE_EP_H
+
+#include <linux/device.h>
+
+/* Synopsis specific PCIE configuration registers */
+#define PCIE_ATU_VIEWPORT		0x900
+#define PCIE_ATU_REGION_INBOUND		(0x1 << 31)
+#define PCIE_ATU_REGION_OUTBOUND	(0x0 << 31)
+#define PCIE_ATU_REGION_INDEX3		(0x3 << 0)
+#define PCIE_ATU_REGION_INDEX2		(0x2 << 0)
+#define PCIE_ATU_REGION_INDEX1		(0x1 << 0)
+#define PCIE_ATU_REGION_INDEX0		(0x0 << 0)
+#define PCIE_ATU_CR1			0x904
+#define PCIE_ATU_TYPE_MEM		(0x0 << 0)
+#define PCIE_ATU_TYPE_IO		(0x2 << 0)
+#define PCIE_ATU_TYPE_CFG0		(0x4 << 0)
+#define PCIE_ATU_TYPE_CFG1		(0x5 << 0)
+#define PCIE_ATU_CR2			0x908
+#define PCIE_ATU_ENABLE			(0x1 << 31)
+#define PCIE_ATU_BAR_MODE_ENABLE	(0x1 << 30)
+#define PCIE_ATU_LOWER_BASE		0x90C
+#define PCIE_ATU_UPPER_BASE		0x910
+#define PCIE_ATU_LIMIT			0x914
+#define PCIE_ATU_LOWER_TARGET		0x918
+#define PCIE_ATU_BUS(x)			(((x) & 0xff) << 24)
+#define PCIE_ATU_DEV(x)			(((x) & 0x1f) << 19)
+#define PCIE_ATU_FUNC(x)		(((x) & 0x7) << 16)
+#define PCIE_ATU_UPPER_TARGET		0x91C
+
+/* PEX internal configuration registers */
+#define PCIE_DBI_RO_WR_EN	0x8bc /* DBI Read-Only Write Enable Register */
+
+/* PEX LUT registers */
+#define PCIE_LUT_BASE		0x80000
+#define PCIE_LUT_DBG		0x7FC /* PEX LUT Debug register */
+
+#define PCIE_LUT_LCTRL0		0x7F8
+
+#define PCIE_ATU_BAR_NUM(bar)	((bar) << 8)
+#define PCIE_LCTRL0_CFG2_ENABLE	(1 << 31)
+#define PCIE_LCTRL0_VF(vf)	((vf) << 22)
+#define PCIE_LCTRL0_PF(pf)	((pf) << 16)
+#define PCIE_LCTRL0_VF_ACTIVE	(1 << 21)
+#define PCIE_LCTRL0_VAL(pf, vf)	(PCIE_LCTRL0_PF(pf) |			   \
+				 PCIE_LCTRL0_VF(vf) |			   \
+				 ((vf) == 0 ? 0 : PCIE_LCTRL0_VF_ACTIVE) | \
+				 PCIE_LCTRL0_CFG2_ENABLE)
+
+#define PCIE_NO_SRIOV_BAR_BASE	0x1000
+
+#define PCIE_SRIOV_POS		0x178
+#define PCIE_PF_NUM		2
+#define PCIE_VF_NUM		64
+
+struct ls_pcie_ep_drvdata {
+	u32 lut_offset;
+	u32 ltssm_shift;
+	u32 lut_dbg;
+};
+
+struct ls_pcie {
+	struct list_head	ep_list;
+	struct device		*dev;
+	struct dentry		*dir;
+	const struct ls_pcie_ep_drvdata *drvdata;
+	void __iomem		*dbi;
+	void __iomem		*lut;
+	phys_addr_t		out_base;
+	int			sriov;
+	int			index;
+};
+
+struct ls_ep_dev {
+	struct list_head	node;
+	struct ls_pcie		*pcie;
+	struct device		dev;
+	struct dentry		*dir;
+	int			pf_idx;
+	int			vf_idx;
+	int			dev_id;
+	void			*driver_data;
+};
+
+struct ls_ep_dev *ls_pci_ep_find(struct ls_pcie *pcie, int dev_id);
+
+void ls_pcie_iatu_outbound_set(struct ls_pcie *pcie, int idx, int type,
+			      u64 cpu_addr, u64 pci_addr, u32 size);
+
+/* Use bar match mode and MEM type as default */
+void ls_pcie_iatu_inbound_set(struct ls_pcie *pcie, int idx,
+				     int bar, u64 phys);
+
+void ls_pcie_ep_dev_setup_bar(struct ls_ep_dev *ep, int bar, u32 size);
+
+
+void ls_pcie_ep_dev_cfg_enable(struct ls_ep_dev *ep);
+
+int ls_pcie_ep_dbgfs_init(struct ls_pcie *pcie);
+int ls_pcie_ep_dbgfs_remove(struct ls_pcie *pcie);
+
+#endif /* _PCIE_LAYERSCAPE_EP_H */
--- a/drivers/pci/host/pci-layerscape.c
+++ b/drivers/pci/host/pci-layerscape.c
@@ -33,14 +33,18 @@
 
 /* PEX Internal Configuration Registers */
 #define PCIE_STRFMR1		0x71c /* Symbol Timer & Filter Mask Register1 */
+#define PCIE_ABSERR		0x8d0 /* Bridge Slave Error Response Register */
+#define PCIE_ABSERR_SETTING	0x9401 /* Forward error of non-posted request */
 #define PCIE_DBI_RO_WR_EN	0x8bc /* DBI Read-Only Write Enable Register */
 
-/* PEX LUT registers */
-#define PCIE_LUT_DBG		0x7FC /* PEX LUT Debug Register */
+#define PCIE_IATU_NUM		6
+
+static void ls_pcie_host_init(struct pcie_port *pp);
 
 struct ls_pcie_drvdata {
 	u32 lut_offset;
 	u32 ltssm_shift;
+	u32 lut_dbg;
 	struct pcie_host_ops *ops;
 };
 
@@ -86,6 +90,14 @@ static void ls_pcie_drop_msg_tlp(struct
 	iowrite32(val, pcie->pp.dbi_base + PCIE_STRFMR1);
 }
 
+static void ls_pcie_disable_outbound_atus(struct ls_pcie *pcie)
+{
+	int i;
+
+	for (i = 0; i < PCIE_IATU_NUM; i++)
+		dw_pcie_disable_outbound_atu(&pcie->pp, i);
+}
+
 static int ls1021_pcie_link_up(struct pcie_port *pp)
 {
 	u32 state;
@@ -134,7 +146,7 @@ static int ls_pcie_link_up(struct pcie_p
 	struct ls_pcie *pcie = to_ls_pcie(pp);
 	u32 state;
 
-	state = (ioread32(pcie->lut + PCIE_LUT_DBG) >>
+	state = (ioread32(pcie->lut + pcie->drvdata->lut_dbg) >>
 		 pcie->drvdata->ltssm_shift) &
 		 LTSSM_STATE_MASK;
 
@@ -144,6 +156,12 @@ static int ls_pcie_link_up(struct pcie_p
 	return 1;
 }
 
+/* Forward error response of outbound non-posted requests */
+static void ls_pcie_fix_error_response(struct ls_pcie *pcie)
+{
+	iowrite32(PCIE_ABSERR_SETTING, pcie->pp.dbi_base + PCIE_ABSERR);
+}
+
 static void ls_pcie_host_init(struct pcie_port *pp)
 {
 	struct ls_pcie *pcie = to_ls_pcie(pp);
@@ -153,6 +171,10 @@ static void ls_pcie_host_init(struct pci
 	ls_pcie_clear_multifunction(pcie);
 	ls_pcie_drop_msg_tlp(pcie);
 	iowrite32(0, pcie->pp.dbi_base + PCIE_DBI_RO_WR_EN);
+
+	ls_pcie_disable_outbound_atus(pcie);
+	ls_pcie_fix_error_response(pcie);
+	dw_pcie_setup_rc(pp);
 }
 
 static int ls_pcie_msi_host_init(struct pcie_port *pp,
@@ -196,20 +218,40 @@ static struct ls_pcie_drvdata ls1021_drv
 static struct ls_pcie_drvdata ls1043_drvdata = {
 	.lut_offset = 0x10000,
 	.ltssm_shift = 24,
+	.lut_dbg = 0x7fc,
+	.ops = &ls_pcie_host_ops,
+};
+
+static struct ls_pcie_drvdata ls1046_drvdata = {
+	.lut_offset = 0x80000,
+	.ltssm_shift = 24,
+	.lut_dbg = 0x407fc,
 	.ops = &ls_pcie_host_ops,
 };
 
 static struct ls_pcie_drvdata ls2080_drvdata = {
 	.lut_offset = 0x80000,
 	.ltssm_shift = 0,
+	.lut_dbg = 0x7fc,
+	.ops = &ls_pcie_host_ops,
+};
+
+static struct ls_pcie_drvdata ls2088_drvdata = {
+	.lut_offset = 0x80000,
+	.ltssm_shift = 0,
+	.lut_dbg = 0x407fc,
 	.ops = &ls_pcie_host_ops,
 };
 
 static const struct of_device_id ls_pcie_of_match[] = {
+	{ .compatible = "fsl,ls1012a-pcie", .data = &ls1046_drvdata },
 	{ .compatible = "fsl,ls1021a-pcie", .data = &ls1021_drvdata },
 	{ .compatible = "fsl,ls1043a-pcie", .data = &ls1043_drvdata },
+	{ .compatible = "fsl,ls1046a-pcie", .data = &ls1046_drvdata },
 	{ .compatible = "fsl,ls2080a-pcie", .data = &ls2080_drvdata },
 	{ .compatible = "fsl,ls2085a-pcie", .data = &ls2080_drvdata },
+	{ .compatible = "fsl,ls2088a-pcie", .data = &ls2088_drvdata },
+	{ .compatible = "fsl,ls1088a-pcie", .data = &ls2088_drvdata },
 	{ },
 };
 
--- a/drivers/pci/host/pcie-designware.c
+++ b/drivers/pci/host/pcie-designware.c
@@ -478,6 +478,12 @@ int dw_pcie_wait_for_link(struct pcie_po
 	return -ETIMEDOUT;
 }
 
+void dw_pcie_disable_outbound_atu(struct pcie_port *pp, int index)
+{
+	dw_pcie_writel_rc(pp, PCIE_ATU_VIEWPORT, PCIE_ATU_REGION_OUTBOUND | index);
+	dw_pcie_writel_rc(pp, PCIE_ATU_CR2, 0);
+}
+
 int dw_pcie_link_up(struct pcie_port *pp)
 {
 	u32 val;
--- a/drivers/pci/host/pcie-designware.h
+++ b/drivers/pci/host/pcie-designware.h
@@ -82,5 +82,6 @@ int dw_pcie_wait_for_link(struct pcie_po
 int dw_pcie_link_up(struct pcie_port *pp);
 void dw_pcie_setup_rc(struct pcie_port *pp);
 int dw_pcie_host_init(struct pcie_port *pp);
+void dw_pcie_disable_outbound_atu(struct pcie_port *pp, int index);
 
 #endif /* _PCIE_DESIGNWARE_H */
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -454,7 +454,7 @@ struct resource *pci_find_parent_resourc
 	pci_bus_for_each_resource(bus, r, i) {
 		if (!r)
 			continue;
-		if (res->start && resource_contains(r, res)) {
+		if (resource_contains(r, res)) {
 
 			/*
 			 * If the window is prefetchable but the BAR is
--- a/drivers/pci/pcie/portdrv_core.c
+++ b/drivers/pci/pcie/portdrv_core.c
@@ -44,52 +44,30 @@ static void release_pcie_device(struct d
 }
 
 /**
- * pcie_port_msix_add_entry - add entry to given array of MSI-X entries
- * @entries: Array of MSI-X entries
- * @new_entry: Index of the entry to add to the array
- * @nr_entries: Number of entries already in the array
+ * pcibios_check_service_irqs - check irqs in the device tree
+ * @dev: PCI Express port to handle
+ * @irqs: Array of irqs to populate
+ * @mask: Bitmask of port capabilities returned by get_port_device_capability()
+ *
+ * Return value: 0 means no service irqs in the device tree
  *
- * Return value: Position of the added entry in the array
  */
-static int pcie_port_msix_add_entry(
-	struct msix_entry *entries, int new_entry, int nr_entries)
+int __weak pcibios_check_service_irqs(struct pci_dev *dev, int *irqs, int mask)
 {
-	int j;
-
-	for (j = 0; j < nr_entries; j++)
-		if (entries[j].entry == new_entry)
-			return j;
-
-	entries[j].entry = new_entry;
-	return j;
+	return 0;
 }
 
 /**
  * pcie_port_enable_msix - try to set up MSI-X as interrupt mode for given port
  * @dev: PCI Express port to handle
- * @vectors: Array of interrupt vectors to populate
+ * @irqs: Array of interrupt vectors to populate
  * @mask: Bitmask of port capabilities returned by get_port_device_capability()
  *
  * Return value: 0 on success, error code on failure
  */
-static int pcie_port_enable_msix(struct pci_dev *dev, int *vectors, int mask)
+static int pcie_port_enable_msix(struct pci_dev *dev, int *irqs, int mask)
 {
-	struct msix_entry *msix_entries;
-	int idx[PCIE_PORT_DEVICE_MAXSERVICES];
-	int nr_entries, status, pos, i, nvec;
-	u16 reg16;
-	u32 reg32;
-
-	nr_entries = pci_msix_vec_count(dev);
-	if (nr_entries < 0)
-		return nr_entries;
-	BUG_ON(!nr_entries);
-	if (nr_entries > PCIE_PORT_MAX_MSIX_ENTRIES)
-		nr_entries = PCIE_PORT_MAX_MSIX_ENTRIES;
-
-	msix_entries = kzalloc(sizeof(*msix_entries) * nr_entries, GFP_KERNEL);
-	if (!msix_entries)
-		return -ENOMEM;
+	int nr_entries, entry, nvec = 0;
 
 	/*
 	 * Allocate as many entries as the port wants, so that we can check
@@ -97,20 +75,13 @@ static int pcie_port_enable_msix(struct
 	 * equal to the number of entries this port actually uses, we'll happily
 	 * go through without any tricks.
 	 */
-	for (i = 0; i < nr_entries; i++)
-		msix_entries[i].entry = i;
-
-	status = pci_enable_msix_exact(dev, msix_entries, nr_entries);
-	if (status)
-		goto Exit;
-
-	for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++)
-		idx[i] = -1;
-	status = -EIO;
-	nvec = 0;
+	nr_entries = pci_alloc_irq_vectors(dev, 1, PCIE_PORT_MAX_MSIX_ENTRIES,
+			PCI_IRQ_MSIX);
+	if (nr_entries < 0)
+		return nr_entries;
 
 	if (mask & (PCIE_PORT_SERVICE_PME | PCIE_PORT_SERVICE_HP)) {
-		int entry;
+		u16 reg16;
 
 		/*
 		 * The code below follows the PCI Express Base Specification 2.0
@@ -125,18 +96,16 @@ static int pcie_port_enable_msix(struct
 		pcie_capability_read_word(dev, PCI_EXP_FLAGS, &reg16);
 		entry = (reg16 & PCI_EXP_FLAGS_IRQ) >> 9;
 		if (entry >= nr_entries)
-			goto Error;
+			goto out_free_irqs;
 
-		i = pcie_port_msix_add_entry(msix_entries, entry, nvec);
-		if (i == nvec)
-			nvec++;
+		irqs[PCIE_PORT_SERVICE_PME_SHIFT] = pci_irq_vector(dev, entry);
+		irqs[PCIE_PORT_SERVICE_HP_SHIFT] = pci_irq_vector(dev, entry);
 
-		idx[PCIE_PORT_SERVICE_PME_SHIFT] = i;
-		idx[PCIE_PORT_SERVICE_HP_SHIFT] = i;
+		nvec = max(nvec, entry + 1);
 	}
 
 	if (mask & PCIE_PORT_SERVICE_AER) {
-		int entry;
+		u32 reg32, pos;
 
 		/*
 		 * The code below follows Section 7.10.10 of the PCI Express
@@ -151,13 +120,11 @@ static int pcie_port_enable_msix(struct
 		pci_read_config_dword(dev, pos + PCI_ERR_ROOT_STATUS, &reg32);
 		entry = reg32 >> 27;
 		if (entry >= nr_entries)
-			goto Error;
+			goto out_free_irqs;
 
-		i = pcie_port_msix_add_entry(msix_entries, entry, nvec);
-		if (i == nvec)
-			nvec++;
+		irqs[PCIE_PORT_SERVICE_AER_SHIFT] = pci_irq_vector(dev, entry);
 
-		idx[PCIE_PORT_SERVICE_AER_SHIFT] = i;
+		nvec = max(nvec, entry + 1);
 	}
 
 	/*
@@ -165,41 +132,54 @@ static int pcie_port_enable_msix(struct
 	 * what we have.  Otherwise, the port has some extra entries not for the
 	 * services we know and we need to work around that.
 	 */
-	if (nvec == nr_entries) {
-		status = 0;
-	} else {
+	if (nvec != nr_entries) {
 		/* Drop the temporary MSI-X setup */
-		pci_disable_msix(dev);
+		pci_free_irq_vectors(dev);
 
 		/* Now allocate the MSI-X vectors for real */
-		status = pci_enable_msix_exact(dev, msix_entries, nvec);
-		if (status)
-			goto Exit;
+		nr_entries = pci_alloc_irq_vectors(dev, nvec, nvec,
+				PCI_IRQ_MSIX);
+		if (nr_entries < 0)
+			return nr_entries;
 	}
 
-	for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++)
-		vectors[i] = idx[i] >= 0 ? msix_entries[idx[i]].vector : -1;
-
- Exit:
-	kfree(msix_entries);
-	return status;
+	return 0;
 
- Error:
-	pci_disable_msix(dev);
-	goto Exit;
+out_free_irqs:
+	pci_free_irq_vectors(dev);
+	return -EIO;
 }
 
 /**
- * init_service_irqs - initialize irqs for PCI Express port services
+ * pcie_init_service_irqs - initialize irqs for PCI Express port services
  * @dev: PCI Express port to handle
  * @irqs: Array of irqs to populate
  * @mask: Bitmask of port capabilities returned by get_port_device_capability()
  *
  * Return value: Interrupt mode associated with the port
  */
-static int init_service_irqs(struct pci_dev *dev, int *irqs, int mask)
+static int pcie_init_service_irqs(struct pci_dev *dev, int *irqs, int mask)
 {
-	int i, irq = -1;
+	unsigned flags = PCI_IRQ_LEGACY | PCI_IRQ_MSI;
+	int ret, i;
+	int irq = -1;
+
+	for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++)
+		irqs[i] = -1;
+
+	/* Check if some platforms owns independent irq pins for AER/PME etc.
+	 * Some platforms may own independent AER/PME interrupts and set
+	 * them in the device tree file.
+	 */
+	ret = pcibios_check_service_irqs(dev, irqs, mask);
+	if (ret) {
+		if (dev->irq)
+			irq = dev->irq;
+		for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++)
+			if (irqs[i] == -1 && i != PCIE_PORT_SERVICE_VC_SHIFT)
+				irqs[i] = irq;
+		return 0;
+	}
 
 	/*
 	 * If MSI cannot be used for PCIe PME or hotplug, we have to use
@@ -207,41 +187,25 @@ static int init_service_irqs(struct pci_
 	 */
 	if (((mask & PCIE_PORT_SERVICE_PME) && pcie_pme_no_msi()) ||
 	    ((mask & PCIE_PORT_SERVICE_HP) && pciehp_no_msi())) {
-		if (dev->irq)
-			irq = dev->irq;
-		goto no_msi;
+		flags &= ~PCI_IRQ_MSI;
+	} else {
+		/* Try to use MSI-X if supported */
+		if (!pcie_port_enable_msix(dev, irqs, mask))
+			return 0;
 	}
 
-	/* Try to use MSI-X if supported */
-	if (!pcie_port_enable_msix(dev, irqs, mask))
-		return 0;
-
-	/*
-	 * We're not going to use MSI-X, so try MSI and fall back to INTx.
-	 * If neither MSI/MSI-X nor INTx available, try other interrupt.  On
-	 * some platforms, root port doesn't support MSI/MSI-X/INTx in RC mode.
-	 */
-	if (!pci_enable_msi(dev) || dev->irq)
-		irq = dev->irq;
+	ret = pci_alloc_irq_vectors(dev, 1, 1, flags);
+	if (ret < 0)
+		return -ENODEV;
 
- no_msi:
-	for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++)
-		irqs[i] = irq;
-	irqs[PCIE_PORT_SERVICE_VC_SHIFT] = -1;
+	for (i = 0; i < PCIE_PORT_DEVICE_MAXSERVICES; i++) {
+		if (i != PCIE_PORT_SERVICE_VC_SHIFT)
+			irqs[i] = pci_irq_vector(dev, 0);
+	}
 
-	if (irq < 0)
-		return -ENODEV;
 	return 0;
 }
 
-static void cleanup_service_irqs(struct pci_dev *dev)
-{
-	if (dev->msix_enabled)
-		pci_disable_msix(dev);
-	else if (dev->msi_enabled)
-		pci_disable_msi(dev);
-}
-
 /**
  * get_port_device_capability - discover capabilities of a PCI Express port
  * @dev: PCI Express port to examine
@@ -378,7 +342,7 @@ int pcie_port_device_register(struct pci
 	 * that can be used in the absence of irqs.  Allow them to determine
 	 * if that is to be used.
 	 */
-	status = init_service_irqs(dev, irqs, capabilities);
+	status = pcie_init_service_irqs(dev, irqs, capabilities);
 	if (status) {
 		capabilities &= PCIE_PORT_SERVICE_VC | PCIE_PORT_SERVICE_HP;
 		if (!capabilities)
@@ -401,7 +365,7 @@ int pcie_port_device_register(struct pci
 	return 0;
 
 error_cleanup_irqs:
-	cleanup_service_irqs(dev);
+	pci_free_irq_vectors(dev);
 error_disable:
 	pci_disable_device(dev);
 	return status;
@@ -469,7 +433,7 @@ static int remove_iter(struct device *de
 void pcie_port_device_remove(struct pci_dev *dev)
 {
 	device_for_each_child(&dev->dev, NULL, remove_iter);
-	cleanup_service_irqs(dev);
+	pci_free_irq_vectors(dev);
 	pci_disable_device(dev);
 }
 
@@ -499,7 +463,6 @@ static int pcie_port_probe_service(struc
 	if (status)
 		return status;
 
-	dev_printk(KERN_DEBUG, dev, "service driver %s loaded\n", driver->name);
 	get_device(dev);
 	return 0;
 }
@@ -524,8 +487,6 @@ static int pcie_port_remove_service(stru
 	pciedev = to_pcie_device(dev);
 	driver = to_service_driver(dev->driver);
 	if (driver && driver->remove) {
-		dev_printk(KERN_DEBUG, dev, "unloading service driver %s\n",
-			driver->name);
 		driver->remove(pciedev);
 		put_device(dev);
 	}
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -4677,3 +4677,11 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_IN
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x2031, quirk_no_aersid);
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x2032, quirk_no_aersid);
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x2033, quirk_no_aersid);
+
+/* Freescale PCIe doesn't support MSI in RC mode */
+static void quirk_fsl_no_msi(struct pci_dev *pdev)
+{
+	if (pci_pcie_type(pdev) == PCI_EXP_TYPE_ROOT_PORT)
+		pdev->no_msi = 1;
+}
+DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_FREESCALE, PCI_ANY_ID, quirk_fsl_no_msi);
--- a/include/linux/pci.h
+++ b/include/linux/pci.h
@@ -1825,6 +1825,7 @@ void pcibios_release_device(struct pci_d
 void pcibios_penalize_isa_irq(int irq, int active);
 int pcibios_alloc_irq(struct pci_dev *dev);
 void pcibios_free_irq(struct pci_dev *dev);
+int pcibios_check_service_irqs(struct pci_dev *dev, int *irqs, int mask);
 
 #ifdef CONFIG_HIBERNATE_CALLBACKS
 extern struct dev_pm_ops pcibios_pm_ops;