From 080b47def5e5e28b2509c5bb92160d1df730f27b Mon Sep 17 00:00:00 2001 From: Logan Gunthorpe Date: Mon, 6 Mar 2017 18:30:54 -0600 Subject: MicroSemi Switchtec management interface driver Microsemi's "Switchtec" line of PCI switch devices is already well supported by the kernel with standard PCI switch drivers. However, the Switchtec device advertises a special management endpoint with a separate PCI function address and class code. This endpoint enables some additional functionality which includes: * Packet and Byte Counters * Switch Firmware Upgrades * Event and Error logs * Querying port link status * Custom user firmware commands Add a switchtec kernel module which provides PCI driver that exposes a char device. The char device provides userspace access to this interface through read, write and (optionally) poll calls. A userspace tool and library which utilizes this interface is available at [1]. This tool takes inspiration (and borrows some code) from nvme-cli [2]. The tool is largely complete at this time but additional features may be added in the future. [1] https://github.com/sbates130272/switchtec-user [2] https://github.com/linux-nvme/nvme-cli [Dan Carpenter : don't invert error codes] [Christophe JAILLET : fix switchtec_dev_open() error handling] Tested-by: Krishna Dhulipala Signed-off-by: Logan Gunthorpe Signed-off-by: Stephen Bates Signed-off-by: Bjorn Helgaas Reviewed-by: Wei Zhang Reviewed-by: Jens Axboe Reviewed-by: Greg Kroah-Hartman --- MAINTAINERS | 8 + drivers/pci/Kconfig | 1 + drivers/pci/Makefile | 1 + drivers/pci/switch/Kconfig | 13 + drivers/pci/switch/Makefile | 1 + drivers/pci/switch/switchtec.c | 1006 ++++++++++++++++++++++++++++++++++++++++ 6 files changed, 1030 insertions(+) create mode 100644 drivers/pci/switch/Kconfig create mode 100644 drivers/pci/switch/Makefile create mode 100644 drivers/pci/switch/switchtec.c diff --git a/MAINTAINERS b/MAINTAINERS index c265a5fe4848..16bd75d30ac1 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9655,6 +9655,14 @@ S: Maintained F: Documentation/devicetree/bindings/pci/aardvark-pci.txt F: drivers/pci/host/pci-aardvark.c +PCI DRIVER FOR MICROSEMI SWITCHTEC +M: Kurt Schwemmer +M: Stephen Bates +M: Logan Gunthorpe +L: linux-pci@vger.kernel.org +S: Maintained +F: drivers/pci/switch/switchtec* + PCI DRIVER FOR NVIDIA TEGRA M: Thierry Reding L: linux-tegra@vger.kernel.org diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig index df141420c902..c0bb4cbd1edb 100644 --- a/drivers/pci/Kconfig +++ b/drivers/pci/Kconfig @@ -134,3 +134,4 @@ config PCI_HYPERV source "drivers/pci/hotplug/Kconfig" source "drivers/pci/dwc/Kconfig" source "drivers/pci/host/Kconfig" +source "drivers/pci/switch/Kconfig" diff --git a/drivers/pci/Makefile b/drivers/pci/Makefile index 8db5079f09a7..15b46dd5074b 100644 --- a/drivers/pci/Makefile +++ b/drivers/pci/Makefile @@ -68,3 +68,4 @@ ccflags-$(CONFIG_PCI_DEBUG) := -DDEBUG # PCI host controller drivers obj-y += host/ +obj-y += switch/ diff --git a/drivers/pci/switch/Kconfig b/drivers/pci/switch/Kconfig new file mode 100644 index 000000000000..4c49648e0646 --- /dev/null +++ b/drivers/pci/switch/Kconfig @@ -0,0 +1,13 @@ +menu "PCI switch controller drivers" + depends on PCI + +config PCI_SW_SWITCHTEC + tristate "MicroSemi Switchtec PCIe Switch Management Driver" + help + Enables support for the management interface for the MicroSemi + Switchtec series of PCIe switches. Supports userspace access + to submit MRPC commands to the switch via /dev/switchtecX + devices. See for more + information. + +endmenu diff --git a/drivers/pci/switch/Makefile b/drivers/pci/switch/Makefile new file mode 100644 index 000000000000..37d8cfb03f3f --- /dev/null +++ b/drivers/pci/switch/Makefile @@ -0,0 +1 @@ +obj-$(CONFIG_PCI_SW_SWITCHTEC) += switchtec.o diff --git a/drivers/pci/switch/switchtec.c b/drivers/pci/switch/switchtec.c new file mode 100644 index 000000000000..0bbe90df968d --- /dev/null +++ b/drivers/pci/switch/switchtec.c @@ -0,0 +1,1006 @@ +/* + * Microsemi Switchtec(tm) PCIe Management Driver + * Copyright (c) 2017, Microsemi Corporation + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +MODULE_DESCRIPTION("Microsemi Switchtec(tm) PCIe Management Driver"); +MODULE_VERSION("0.1"); +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Microsemi Corporation"); + +static int max_devices = 16; +module_param(max_devices, int, 0644); +MODULE_PARM_DESC(max_devices, "max number of switchtec device instances"); + +static dev_t switchtec_devt; +static struct class *switchtec_class; +static DEFINE_IDA(switchtec_minor_ida); + +#define MICROSEMI_VENDOR_ID 0x11f8 +#define MICROSEMI_NTB_CLASSCODE 0x068000 +#define MICROSEMI_MGMT_CLASSCODE 0x058000 + +#define SWITCHTEC_MRPC_PAYLOAD_SIZE 1024 +#define SWITCHTEC_MAX_PFF_CSR 48 + +#define SWITCHTEC_EVENT_OCCURRED BIT(0) +#define SWITCHTEC_EVENT_CLEAR BIT(0) +#define SWITCHTEC_EVENT_EN_LOG BIT(1) +#define SWITCHTEC_EVENT_EN_CLI BIT(2) +#define SWITCHTEC_EVENT_EN_IRQ BIT(3) +#define SWITCHTEC_EVENT_FATAL BIT(4) + +enum { + SWITCHTEC_GAS_MRPC_OFFSET = 0x0000, + SWITCHTEC_GAS_TOP_CFG_OFFSET = 0x1000, + SWITCHTEC_GAS_SW_EVENT_OFFSET = 0x1800, + SWITCHTEC_GAS_SYS_INFO_OFFSET = 0x2000, + SWITCHTEC_GAS_FLASH_INFO_OFFSET = 0x2200, + SWITCHTEC_GAS_PART_CFG_OFFSET = 0x4000, + SWITCHTEC_GAS_NTB_OFFSET = 0x10000, + SWITCHTEC_GAS_PFF_CSR_OFFSET = 0x134000, +}; + +struct mrpc_regs { + u8 input_data[SWITCHTEC_MRPC_PAYLOAD_SIZE]; + u8 output_data[SWITCHTEC_MRPC_PAYLOAD_SIZE]; + u32 cmd; + u32 status; + u32 ret_value; +} __packed; + +enum mrpc_status { + SWITCHTEC_MRPC_STATUS_INPROGRESS = 1, + SWITCHTEC_MRPC_STATUS_DONE = 2, + SWITCHTEC_MRPC_STATUS_ERROR = 0xFF, + SWITCHTEC_MRPC_STATUS_INTERRUPTED = 0x100, +}; + +struct sw_event_regs { + u64 event_report_ctrl; + u64 reserved1; + u64 part_event_bitmap; + u64 reserved2; + u32 global_summary; + u32 reserved3[3]; + u32 stack_error_event_hdr; + u32 stack_error_event_data; + u32 reserved4[4]; + u32 ppu_error_event_hdr; + u32 ppu_error_event_data; + u32 reserved5[4]; + u32 isp_error_event_hdr; + u32 isp_error_event_data; + u32 reserved6[4]; + u32 sys_reset_event_hdr; + u32 reserved7[5]; + u32 fw_exception_hdr; + u32 reserved8[5]; + u32 fw_nmi_hdr; + u32 reserved9[5]; + u32 fw_non_fatal_hdr; + u32 reserved10[5]; + u32 fw_fatal_hdr; + u32 reserved11[5]; + u32 twi_mrpc_comp_hdr; + u32 twi_mrpc_comp_data; + u32 reserved12[4]; + u32 twi_mrpc_comp_async_hdr; + u32 twi_mrpc_comp_async_data; + u32 reserved13[4]; + u32 cli_mrpc_comp_hdr; + u32 cli_mrpc_comp_data; + u32 reserved14[4]; + u32 cli_mrpc_comp_async_hdr; + u32 cli_mrpc_comp_async_data; + u32 reserved15[4]; + u32 gpio_interrupt_hdr; + u32 gpio_interrupt_data; + u32 reserved16[4]; +} __packed; + +struct sys_info_regs { + u32 device_id; + u32 device_version; + u32 firmware_version; + u32 reserved1; + u32 vendor_table_revision; + u32 table_format_version; + u32 partition_id; + u32 cfg_file_fmt_version; + u32 reserved2[58]; + char vendor_id[8]; + char product_id[16]; + char product_revision[4]; + char component_vendor[8]; + u16 component_id; + u8 component_revision; +} __packed; + +struct flash_info_regs { + u32 flash_part_map_upd_idx; + + struct active_partition_info { + u32 address; + u32 build_version; + u32 build_string; + } active_img; + + struct active_partition_info active_cfg; + struct active_partition_info inactive_img; + struct active_partition_info inactive_cfg; + + u32 flash_length; + + struct partition_info { + u32 address; + u32 length; + } cfg0; + + struct partition_info cfg1; + struct partition_info img0; + struct partition_info img1; + struct partition_info nvlog; + struct partition_info vendor[8]; +}; + +struct ntb_info_regs { + u8 partition_count; + u8 partition_id; + u16 reserved1; + u64 ep_map; + u16 requester_id; +} __packed; + +struct part_cfg_regs { + u32 status; + u32 state; + u32 port_cnt; + u32 usp_port_mode; + u32 usp_pff_inst_id; + u32 vep_pff_inst_id; + u32 dsp_pff_inst_id[47]; + u32 reserved1[11]; + u16 vep_vector_number; + u16 usp_vector_number; + u32 port_event_bitmap; + u32 reserved2[3]; + u32 part_event_summary; + u32 reserved3[3]; + u32 part_reset_hdr; + u32 part_reset_data[5]; + u32 mrpc_comp_hdr; + u32 mrpc_comp_data[5]; + u32 mrpc_comp_async_hdr; + u32 mrpc_comp_async_data[5]; + u32 dyn_binding_hdr; + u32 dyn_binding_data[5]; + u32 reserved4[159]; +} __packed; + +enum { + SWITCHTEC_PART_CFG_EVENT_RESET = 1 << 0, + SWITCHTEC_PART_CFG_EVENT_MRPC_CMP = 1 << 1, + SWITCHTEC_PART_CFG_EVENT_MRPC_ASYNC_CMP = 1 << 2, + SWITCHTEC_PART_CFG_EVENT_DYN_PART_CMP = 1 << 3, +}; + +struct pff_csr_regs { + u16 vendor_id; + u16 device_id; + u32 pci_cfg_header[15]; + u32 pci_cap_region[48]; + u32 pcie_cap_region[448]; + u32 indirect_gas_window[128]; + u32 indirect_gas_window_off; + u32 reserved[127]; + u32 pff_event_summary; + u32 reserved2[3]; + u32 aer_in_p2p_hdr; + u32 aer_in_p2p_data[5]; + u32 aer_in_vep_hdr; + u32 aer_in_vep_data[5]; + u32 dpc_hdr; + u32 dpc_data[5]; + u32 cts_hdr; + u32 cts_data[5]; + u32 reserved3[6]; + u32 hotplug_hdr; + u32 hotplug_data[5]; + u32 ier_hdr; + u32 ier_data[5]; + u32 threshold_hdr; + u32 threshold_data[5]; + u32 power_mgmt_hdr; + u32 power_mgmt_data[5]; + u32 tlp_throttling_hdr; + u32 tlp_throttling_data[5]; + u32 force_speed_hdr; + u32 force_speed_data[5]; + u32 credit_timeout_hdr; + u32 credit_timeout_data[5]; + u32 link_state_hdr; + u32 link_state_data[5]; + u32 reserved4[174]; +} __packed; + +struct switchtec_dev { + struct pci_dev *pdev; + struct device dev; + struct cdev cdev; + + int partition; + int partition_count; + int pff_csr_count; + char pff_local[SWITCHTEC_MAX_PFF_CSR]; + + void __iomem *mmio; + struct mrpc_regs __iomem *mmio_mrpc; + struct sw_event_regs __iomem *mmio_sw_event; + struct sys_info_regs __iomem *mmio_sys_info; + struct flash_info_regs __iomem *mmio_flash_info; + struct ntb_info_regs __iomem *mmio_ntb; + struct part_cfg_regs __iomem *mmio_part_cfg; + struct part_cfg_regs __iomem *mmio_part_cfg_all; + struct pff_csr_regs __iomem *mmio_pff_csr; + + /* + * The mrpc mutex must be held when accessing the other + * mrpc_ fields, alive flag and stuser->state field + */ + struct mutex mrpc_mutex; + struct list_head mrpc_queue; + int mrpc_busy; + struct work_struct mrpc_work; + struct delayed_work mrpc_timeout; + bool alive; + + wait_queue_head_t event_wq; + atomic_t event_cnt; +}; + +static struct switchtec_dev *to_stdev(struct device *dev) +{ + return container_of(dev, struct switchtec_dev, dev); +} + +enum mrpc_state { + MRPC_IDLE = 0, + MRPC_QUEUED, + MRPC_RUNNING, + MRPC_DONE, +}; + +struct switchtec_user { + struct switchtec_dev *stdev; + + enum mrpc_state state; + + struct completion comp; + struct kref kref; + struct list_head list; + + u32 cmd; + u32 status; + u32 return_code; + size_t data_len; + size_t read_len; + unsigned char data[SWITCHTEC_MRPC_PAYLOAD_SIZE]; + int event_cnt; +}; + +static struct switchtec_user *stuser_create(struct switchtec_dev *stdev) +{ + struct switchtec_user *stuser; + + stuser = kzalloc(sizeof(*stuser), GFP_KERNEL); + if (!stuser) + return ERR_PTR(-ENOMEM); + + get_device(&stdev->dev); + stuser->stdev = stdev; + kref_init(&stuser->kref); + INIT_LIST_HEAD(&stuser->list); + init_completion(&stuser->comp); + stuser->event_cnt = atomic_read(&stdev->event_cnt); + + dev_dbg(&stdev->dev, "%s: %p\n", __func__, stuser); + + return stuser; +} + +static void stuser_free(struct kref *kref) +{ + struct switchtec_user *stuser; + + stuser = container_of(kref, struct switchtec_user, kref); + + dev_dbg(&stuser->stdev->dev, "%s: %p\n", __func__, stuser); + + put_device(&stuser->stdev->dev); + kfree(stuser); +} + +static void stuser_put(struct switchtec_user *stuser) +{ + kref_put(&stuser->kref, stuser_free); +} + +static void stuser_set_state(struct switchtec_user *stuser, + enum mrpc_state state) +{ + /* requires the mrpc_mutex to already be held when called */ + + const char * const state_names[] = { + [MRPC_IDLE] = "IDLE", + [MRPC_QUEUED] = "QUEUED", + [MRPC_RUNNING] = "RUNNING", + [MRPC_DONE] = "DONE", + }; + + stuser->state = state; + + dev_dbg(&stuser->stdev->dev, "stuser state %p -> %s", + stuser, state_names[state]); +} + +static void mrpc_complete_cmd(struct switchtec_dev *stdev); + +static void mrpc_cmd_submit(struct switchtec_dev *stdev) +{ + /* requires the mrpc_mutex to already be held when called */ + + struct switchtec_user *stuser; + + if (stdev->mrpc_busy) + return; + + if (list_empty(&stdev->mrpc_queue)) + return; + + stuser = list_entry(stdev->mrpc_queue.next, struct switchtec_user, + list); + + stuser_set_state(stuser, MRPC_RUNNING); + stdev->mrpc_busy = 1; + memcpy_toio(&stdev->mmio_mrpc->input_data, + stuser->data, stuser->data_len); + iowrite32(stuser->cmd, &stdev->mmio_mrpc->cmd); + + stuser->status = ioread32(&stdev->mmio_mrpc->status); + if (stuser->status != SWITCHTEC_MRPC_STATUS_INPROGRESS) + mrpc_complete_cmd(stdev); + + schedule_delayed_work(&stdev->mrpc_timeout, + msecs_to_jiffies(500)); +} + +static int mrpc_queue_cmd(struct switchtec_user *stuser) +{ + /* requires the mrpc_mutex to already be held when called */ + + struct switchtec_dev *stdev = stuser->stdev; + + kref_get(&stuser->kref); + stuser->read_len = sizeof(stuser->data); + stuser_set_state(stuser, MRPC_QUEUED); + init_completion(&stuser->comp); + list_add_tail(&stuser->list, &stdev->mrpc_queue); + + mrpc_cmd_submit(stdev); + + return 0; +} + +static void mrpc_complete_cmd(struct switchtec_dev *stdev) +{ + /* requires the mrpc_mutex to already be held when called */ + struct switchtec_user *stuser; + + if (list_empty(&stdev->mrpc_queue)) + return; + + stuser = list_entry(stdev->mrpc_queue.next, struct switchtec_user, + list); + + stuser->status = ioread32(&stdev->mmio_mrpc->status); + if (stuser->status == SWITCHTEC_MRPC_STATUS_INPROGRESS) + return; + + stuser_set_state(stuser, MRPC_DONE); + stuser->return_code = 0; + + if (stuser->status != SWITCHTEC_MRPC_STATUS_DONE) + goto out; + + stuser->return_code = ioread32(&stdev->mmio_mrpc->ret_value); + if (stuser->return_code != 0) + goto out; + + memcpy_fromio(stuser->data, &stdev->mmio_mrpc->output_data, + stuser->read_len); + +out: + complete_all(&stuser->comp); + list_del_init(&stuser->list); + stuser_put(stuser); + stdev->mrpc_busy = 0; + + mrpc_cmd_submit(stdev); +} + +static void mrpc_event_work(struct work_struct *work) +{ + struct switchtec_dev *stdev; + + stdev = container_of(work, struct switchtec_dev, mrpc_work); + + dev_dbg(&stdev->dev, "%s\n", __func__); + + mutex_lock(&stdev->mrpc_mutex); + cancel_delayed_work(&stdev->mrpc_timeout); + mrpc_complete_cmd(stdev); + mutex_unlock(&stdev->mrpc_mutex); +} + +static void mrpc_timeout_work(struct work_struct *work) +{ + struct switchtec_dev *stdev; + u32 status; + + stdev = container_of(work, struct switchtec_dev, mrpc_timeout.work); + + dev_dbg(&stdev->dev, "%s\n", __func__); + + mutex_lock(&stdev->mrpc_mutex); + + status = ioread32(&stdev->mmio_mrpc->status); + if (status == SWITCHTEC_MRPC_STATUS_INPROGRESS) { + schedule_delayed_work(&stdev->mrpc_timeout, + msecs_to_jiffies(500)); + goto out; + } + + mrpc_complete_cmd(stdev); + +out: + mutex_unlock(&stdev->mrpc_mutex); +} + +static int switchtec_dev_open(struct inode *inode, struct file *filp) +{ + struct switchtec_dev *stdev; + struct switchtec_user *stuser; + + stdev = container_of(inode->i_cdev, struct switchtec_dev, cdev); + + stuser = stuser_create(stdev); + if (IS_ERR(stuser)) + return PTR_ERR(stuser); + + filp->private_data = stuser; + nonseekable_open(inode, filp); + + dev_dbg(&stdev->dev, "%s: %p\n", __func__, stuser); + + return 0; +} + +static int switchtec_dev_release(struct inode *inode, struct file *filp) +{ + struct switchtec_user *stuser = filp->private_data; + + stuser_put(stuser); + + return 0; +} + +static int lock_mutex_and_test_alive(struct switchtec_dev *stdev) +{ + if (mutex_lock_interruptible(&stdev->mrpc_mutex)) + return -EINTR; + + if (!stdev->alive) { + mutex_unlock(&stdev->mrpc_mutex); + return -ENODEV; + } + + return 0; +} + +static ssize_t switchtec_dev_write(struct file *filp, const char __user *data, + size_t size, loff_t *off) +{ + struct switchtec_user *stuser = filp->private_data; + struct switchtec_dev *stdev = stuser->stdev; + int rc; + + if (size < sizeof(stuser->cmd) || + size > sizeof(stuser->cmd) + sizeof(stuser->data)) + return -EINVAL; + + stuser->data_len = size - sizeof(stuser->cmd); + + rc = lock_mutex_and_test_alive(stdev); + if (rc) + return rc; + + if (stuser->state != MRPC_IDLE) { + rc = -EBADE; + goto out; + } + + rc = copy_from_user(&stuser->cmd, data, sizeof(stuser->cmd)); + if (rc) { + rc = -EFAULT; + goto out; + } + + data += sizeof(stuser->cmd); + rc = copy_from_user(&stuser->data, data, size - sizeof(stuser->cmd)); + if (rc) { + rc = -EFAULT; + goto out; + } + + rc = mrpc_queue_cmd(stuser); + +out: + mutex_unlock(&stdev->mrpc_mutex); + + if (rc) + return rc; + + return size; +} + +static ssize_t switchtec_dev_read(struct file *filp, char __user *data, + size_t size, loff_t *off) +{ + struct switchtec_user *stuser = filp->private_data; + struct switchtec_dev *stdev = stuser->stdev; + int rc; + + if (size < sizeof(stuser->cmd) || + size > sizeof(stuser->cmd) + sizeof(stuser->data)) + return -EINVAL; + + rc = lock_mutex_and_test_alive(stdev); + if (rc) + return rc; + + if (stuser->state == MRPC_IDLE) { + mutex_unlock(&stdev->mrpc_mutex); + return -EBADE; + } + + stuser->read_len = size - sizeof(stuser->return_code); + + mutex_unlock(&stdev->mrpc_mutex); + + if (filp->f_flags & O_NONBLOCK) { + if (!try_wait_for_completion(&stuser->comp)) + return -EAGAIN; + } else { + rc = wait_for_completion_interruptible(&stuser->comp); + if (rc < 0) + return rc; + } + + rc = lock_mutex_and_test_alive(stdev); + if (rc) + return rc; + + if (stuser->state != MRPC_DONE) { + mutex_unlock(&stdev->mrpc_mutex); + return -EBADE; + } + + rc = copy_to_user(data, &stuser->return_code, + sizeof(stuser->return_code)); + if (rc) { + rc = -EFAULT; + goto out; + } + + data += sizeof(stuser->return_code); + rc = copy_to_user(data, &stuser->data, + size - sizeof(stuser->return_code)); + if (rc) { + rc = -EFAULT; + goto out; + } + + stuser_set_state(stuser, MRPC_IDLE); + +out: + mutex_unlock(&stdev->mrpc_mutex); + + if (stuser->status == SWITCHTEC_MRPC_STATUS_DONE) + return size; + else if (stuser->status == SWITCHTEC_MRPC_STATUS_INTERRUPTED) + return -ENXIO; + else + return -EBADMSG; +} + +static unsigned int switchtec_dev_poll(struct file *filp, poll_table *wait) +{ + struct switchtec_user *stuser = filp->private_data; + struct switchtec_dev *stdev = stuser->stdev; + int ret = 0; + + poll_wait(filp, &stuser->comp.wait, wait); + poll_wait(filp, &stdev->event_wq, wait); + + if (lock_mutex_and_test_alive(stdev)) + return POLLIN | POLLRDHUP | POLLOUT | POLLERR | POLLHUP; + + mutex_unlock(&stdev->mrpc_mutex); + + if (try_wait_for_completion(&stuser->comp)) + ret |= POLLIN | POLLRDNORM; + + if (stuser->event_cnt != atomic_read(&stdev->event_cnt)) + ret |= POLLPRI | POLLRDBAND; + + return ret; +} + +static const struct file_operations switchtec_fops = { + .owner = THIS_MODULE, + .open = switchtec_dev_open, + .release = switchtec_dev_release, + .write = switchtec_dev_write, + .read = switchtec_dev_read, + .poll = switchtec_dev_poll, +}; + +static void stdev_release(struct device *dev) +{ + struct switchtec_dev *stdev = to_stdev(dev); + + kfree(stdev); +} + +static void stdev_kill(struct switchtec_dev *stdev) +{ + struct switchtec_user *stuser, *tmpuser; + + pci_clear_master(stdev->pdev); + + cancel_delayed_work_sync(&stdev->mrpc_timeout); + + /* Mark the hardware as unavailable and complete all completions */ + mutex_lock(&stdev->mrpc_mutex); + stdev->alive = false; + + /* Wake up and kill any users waiting on an MRPC request */ + list_for_each_entry_safe(stuser, tmpuser, &stdev->mrpc_queue, list) { + complete_all(&stuser->comp); + list_del_init(&stuser->list); + stuser_put(stuser); + } + + mutex_unlock(&stdev->mrpc_mutex); + + /* Wake up any users waiting on event_wq */ + wake_up_interruptible(&stdev->event_wq); +} + +static struct switchtec_dev *stdev_create(struct pci_dev *pdev) +{ + struct switchtec_dev *stdev; + int minor; + struct device *dev; + struct cdev *cdev; + int rc; + + stdev = kzalloc_node(sizeof(*stdev), GFP_KERNEL, + dev_to_node(&pdev->dev)); + if (!stdev) + return ERR_PTR(-ENOMEM); + + stdev->alive = true; + stdev->pdev = pdev; + INIT_LIST_HEAD(&stdev->mrpc_queue); + mutex_init(&stdev->mrpc_mutex); + stdev->mrpc_busy = 0; + INIT_WORK(&stdev->mrpc_work, mrpc_event_work); + INIT_DELAYED_WORK(&stdev->mrpc_timeout, mrpc_timeout_work); + init_waitqueue_head(&stdev->event_wq); + atomic_set(&stdev->event_cnt, 0); + + dev = &stdev->dev; + device_initialize(dev); + dev->class = switchtec_class; + dev->parent = &pdev->dev; + dev->release = stdev_release; + + minor = ida_simple_get(&switchtec_minor_ida, 0, 0, + GFP_KERNEL); + if (minor < 0) { + rc = minor; + goto err_put; + } + + dev->devt = MKDEV(MAJOR(switchtec_devt), minor); + dev_set_name(dev, "switchtec%d", minor); + + cdev = &stdev->cdev; + cdev_init(cdev, &switchtec_fops); + cdev->owner = THIS_MODULE; + cdev->kobj.parent = &dev->kobj; + + return stdev; + +err_put: + put_device(&stdev->dev); + return ERR_PTR(rc); +} + +static irqreturn_t switchtec_event_isr(int irq, void *dev) +{ + struct switchtec_dev *stdev = dev; + u32 reg; + irqreturn_t ret = IRQ_NONE; + + reg = ioread32(&stdev->mmio_part_cfg->mrpc_comp_hdr); + if (reg & SWITCHTEC_EVENT_OCCURRED) { + dev_dbg(&stdev->dev, "%s: mrpc comp\n", __func__); + ret = IRQ_HANDLED; + schedule_work(&stdev->mrpc_work); + iowrite32(reg, &stdev->mmio_part_cfg->mrpc_comp_hdr); + } + + return ret; +} + +static int switchtec_init_isr(struct switchtec_dev *stdev) +{ + int nvecs; + int event_irq; + + nvecs = pci_alloc_irq_vectors(stdev->pdev, 1, 4, + PCI_IRQ_MSIX | PCI_IRQ_MSI); + if (nvecs < 0) + return nvecs; + + event_irq = ioread32(&stdev->mmio_part_cfg->vep_vector_number); + if (event_irq < 0 || event_irq >= nvecs) + return -EFAULT; + + event_irq = pci_irq_vector(stdev->pdev, event_irq); + if (event_irq < 0) + return event_irq; + + return devm_request_irq(&stdev->pdev->dev, event_irq, + switchtec_event_isr, 0, + KBUILD_MODNAME, stdev); +} + +static void init_pff(struct switchtec_dev *stdev) +{ + int i; + u32 reg; + struct part_cfg_regs *pcfg = stdev->mmio_part_cfg; + + for (i = 0; i < SWITCHTEC_MAX_PFF_CSR; i++) { + reg = ioread16(&stdev->mmio_pff_csr[i].vendor_id); + if (reg != MICROSEMI_VENDOR_ID) + break; + } + + stdev->pff_csr_count = i; + + reg = ioread32(&pcfg->usp_pff_inst_id); + if (reg < SWITCHTEC_MAX_PFF_CSR) + stdev->pff_local[reg] = 1; + + reg = ioread32(&pcfg->vep_pff_inst_id); + if (reg < SWITCHTEC_MAX_PFF_CSR) + stdev->pff_local[reg] = 1; + + for (i = 0; i < ARRAY_SIZE(pcfg->dsp_pff_inst_id); i++) { + reg = ioread32(&pcfg->dsp_pff_inst_id[i]); + if (reg < SWITCHTEC_MAX_PFF_CSR) + stdev->pff_local[reg] = 1; + } +} + +static int switchtec_init_pci(struct switchtec_dev *stdev, + struct pci_dev *pdev) +{ + int rc; + + rc = pcim_enable_device(pdev); + if (rc) + return rc; + + rc = pcim_iomap_regions(pdev, 0x1, KBUILD_MODNAME); + if (rc) + return rc; + + pci_set_master(pdev); + + stdev->mmio = pcim_iomap_table(pdev)[0]; + stdev->mmio_mrpc = stdev->mmio + SWITCHTEC_GAS_MRPC_OFFSET; + stdev->mmio_sw_event = stdev->mmio + SWITCHTEC_GAS_SW_EVENT_OFFSET; + stdev->mmio_sys_info = stdev->mmio + SWITCHTEC_GAS_SYS_INFO_OFFSET; + stdev->mmio_flash_info = stdev->mmio + SWITCHTEC_GAS_FLASH_INFO_OFFSET; + stdev->mmio_ntb = stdev->mmio + SWITCHTEC_GAS_NTB_OFFSET; + stdev->partition = ioread8(&stdev->mmio_ntb->partition_id); + stdev->partition_count = ioread8(&stdev->mmio_ntb->partition_count); + stdev->mmio_part_cfg_all = stdev->mmio + SWITCHTEC_GAS_PART_CFG_OFFSET; + stdev->mmio_part_cfg = &stdev->mmio_part_cfg_all[stdev->partition]; + stdev->mmio_pff_csr = stdev->mmio + SWITCHTEC_GAS_PFF_CSR_OFFSET; + + init_pff(stdev); + + pci_set_drvdata(pdev, stdev); + + return 0; +} + +static int switchtec_pci_probe(struct pci_dev *pdev, + const struct pci_device_id *id) +{ + struct switchtec_dev *stdev; + int rc; + + stdev = stdev_create(pdev); + if (IS_ERR(stdev)) + return PTR_ERR(stdev); + + rc = switchtec_init_pci(stdev, pdev); + if (rc) + goto err_put; + + rc = switchtec_init_isr(stdev); + if (rc) { + dev_err(&stdev->dev, "failed to init isr.\n"); + goto err_put; + } + + iowrite32(SWITCHTEC_EVENT_CLEAR | + SWITCHTEC_EVENT_EN_IRQ, + &stdev->mmio_part_cfg->mrpc_comp_hdr); + + rc = cdev_add(&stdev->cdev, stdev->dev.devt, 1); + if (rc) + goto err_put; + + rc = device_add(&stdev->dev); + if (rc) + goto err_devadd; + + dev_info(&stdev->dev, "Management device registered.\n"); + + return 0; + +err_devadd: + cdev_del(&stdev->cdev); + stdev_kill(stdev); +err_put: + ida_simple_remove(&switchtec_minor_ida, MINOR(stdev->dev.devt)); + put_device(&stdev->dev); + return rc; +} + +static void switchtec_pci_remove(struct pci_dev *pdev) +{ + struct switchtec_dev *stdev = pci_get_drvdata(pdev); + + pci_set_drvdata(pdev, NULL); + + device_del(&stdev->dev); + cdev_del(&stdev->cdev); + ida_simple_remove(&switchtec_minor_ida, MINOR(stdev->dev.devt)); + dev_info(&stdev->dev, "unregistered.\n"); + + stdev_kill(stdev); + put_device(&stdev->dev); +} + +#define SWITCHTEC_PCI_DEVICE(device_id) \ + { \ + .vendor = MICROSEMI_VENDOR_ID, \ + .device = device_id, \ + .subvendor = PCI_ANY_ID, \ + .subdevice = PCI_ANY_ID, \ + .class = MICROSEMI_MGMT_CLASSCODE, \ + .class_mask = 0xFFFFFFFF, \ + }, \ + { \ + .vendor = MICROSEMI_VENDOR_ID, \ + .device = device_id, \ + .subvendor = PCI_ANY_ID, \ + .subdevice = PCI_ANY_ID, \ + .class = MICROSEMI_NTB_CLASSCODE, \ + .class_mask = 0xFFFFFFFF, \ + } + +static const struct pci_device_id switchtec_pci_tbl[] = { + SWITCHTEC_PCI_DEVICE(0x8531), //PFX 24xG3 + SWITCHTEC_PCI_DEVICE(0x8532), //PFX 32xG3 + SWITCHTEC_PCI_DEVICE(0x8533), //PFX 48xG3 + SWITCHTEC_PCI_DEVICE(0x8534), //PFX 64xG3 + SWITCHTEC_PCI_DEVICE(0x8535), //PFX 80xG3 + SWITCHTEC_PCI_DEVICE(0x8536), //PFX 96xG3 + SWITCHTEC_PCI_DEVICE(0x8543), //PSX 48xG3 + SWITCHTEC_PCI_DEVICE(0x8544), //PSX 64xG3 + SWITCHTEC_PCI_DEVICE(0x8545), //PSX 80xG3 + SWITCHTEC_PCI_DEVICE(0x8546), //PSX 96xG3 + {0} +}; +MODULE_DEVICE_TABLE(pci, switchtec_pci_tbl); + +static struct pci_driver switchtec_pci_driver = { + .name = KBUILD_MODNAME, + .id_table = switchtec_pci_tbl, + .probe = switchtec_pci_probe, + .remove = switchtec_pci_remove, +}; + +static int __init switchtec_init(void) +{ + int rc; + + rc = alloc_chrdev_region(&switchtec_devt, 0, max_devices, + "switchtec"); + if (rc) + return rc; + + switchtec_class = class_create(THIS_MODULE, "switchtec"); + if (IS_ERR(switchtec_class)) { + rc = PTR_ERR(switchtec_class); + goto err_create_class; + } + + rc = pci_register_driver(&switchtec_pci_driver); + if (rc) + goto err_pci_register; + + pr_info(KBUILD_MODNAME ": loaded.\n"); + + return 0; + +err_pci_register: + class_destroy(switchtec_class); + +err_create_class: + unregister_chrdev_region(switchtec_devt, max_devices); + + return rc; +} +module_init(switchtec_init); + +static void __exit switchtec_exit(void) +{ + pci_unregister_driver(&switchtec_pci_driver); + class_destroy(switchtec_class); + unregister_chrdev_region(switchtec_devt, max_devices); + ida_destroy(&switchtec_minor_ida); + + pr_info(KBUILD_MODNAME ": unloaded.\n"); +} +module_exit(switchtec_exit); -- cgit v1.2.3-59-g8ed1b From fda78d7a0ead144f4b2cdb582dcba47911f4952c Mon Sep 17 00:00:00 2001 From: Prarit Bhargava Date: Thu, 26 Jan 2017 14:07:47 -0500 Subject: PCI/MSI: Stop disabling MSI/MSI-X in pci_device_shutdown() The pci_bus_type .shutdown method, pci_device_shutdown(), is called from device_shutdown() in the kernel restart and shutdown paths. Previously, pci_device_shutdown() called pci_msi_shutdown() and pci_msix_shutdown(). This disables MSI and MSI-X, which causes the device to fall back to raising interrupts via INTx. But the driver is still bound to the device, it doesn't know about this change, and it likely doesn't have an INTx handler, so these INTx interrupts cause "nobody cared" warnings like this: irq 16: nobody cared (try booting with the "irqpoll" option) CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.8.2-1.el7_UNSUPPORTED.x86_64 #1 Hardware name: Hewlett-Packard HP Z820 Workstation/158B, BIOS J63 v03.90 06/ ... The MSI disabling code was added by d52877c7b1af ("pci/irq: let pci_device_shutdown to call pci_msi_shutdown v2") because a driver left MSI enabled and kdump failed because the kexeced kernel wasn't prepared to receive the MSI interrupts. Subsequent commits 1851617cd2da ("PCI/MSI: Disable MSI at enumeration even if kernel doesn't support MSI") and e80e7edc55ba ("PCI/MSI: Initialize MSI capability for all architectures") changed the kexeced kernel to disable all MSIs itself so it no longer depends on the crashed kernel to clean up after itself. Stop disabling MSI/MSI-X in pci_device_shutdown(). This resolves the "nobody cared" unhandled IRQ issue above. It also allows PCI serial devices, which may rely on the MSI interrupts, to continue outputting messages during reboot/shutdown. [bhelgaas: changelog, drop pci_msi_shutdown() and pci_msix_shutdown() calls altogether] Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=187351 Signed-off-by: Prarit Bhargava Signed-off-by: Bjorn Helgaas CC: Alex Williamson CC: David Arcari CC: Myron Stowe CC: Lukas Wunner CC: Keith Busch CC: Mika Westerberg --- drivers/pci/pci-driver.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index afa72717a979..8ec136164e93 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -461,8 +461,6 @@ static void pci_device_shutdown(struct device *dev) if (drv && drv->shutdown) drv->shutdown(pci_dev); - pci_msi_shutdown(pci_dev); - pci_msix_shutdown(pci_dev); /* * If this is a kexec reboot, turn off Bus Master bit on the -- cgit v1.2.3-59-g8ed1b From 688769f643bfce894f14dc7141bfc6c010f52750 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 9 Mar 2017 15:45:14 -0600 Subject: PCI/MSI: Make pci_msi_shutdown() and pci_msix_shutdown() static pci_msi_shutdown() and pci_msix_shutdown() are used only in drivers/pci/msi.c, so make them static. Signed-off-by: Bjorn Helgaas --- drivers/pci/msi.c | 4 ++-- include/linux/pci.h | 4 ---- 2 files changed, 2 insertions(+), 6 deletions(-) diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c index d571bc330686..4d062c3bf5f0 100644 --- a/drivers/pci/msi.c +++ b/drivers/pci/msi.c @@ -882,7 +882,7 @@ int pci_msi_vec_count(struct pci_dev *dev) } EXPORT_SYMBOL(pci_msi_vec_count); -void pci_msi_shutdown(struct pci_dev *dev) +static void pci_msi_shutdown(struct pci_dev *dev) { struct msi_desc *desc; u32 mask; @@ -994,7 +994,7 @@ int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec) } EXPORT_SYMBOL(pci_enable_msix); -void pci_msix_shutdown(struct pci_dev *dev) +static void pci_msix_shutdown(struct pci_dev *dev) { struct msi_desc *entry; diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..10917c122974 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1297,11 +1297,9 @@ struct msix_entry { #ifdef CONFIG_PCI_MSI int pci_msi_vec_count(struct pci_dev *dev); -void pci_msi_shutdown(struct pci_dev *dev); void pci_disable_msi(struct pci_dev *dev); int pci_msix_vec_count(struct pci_dev *dev); int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec); -void pci_msix_shutdown(struct pci_dev *dev); void pci_disable_msix(struct pci_dev *dev); void pci_restore_msi_state(struct pci_dev *dev); int pci_msi_enabled(void); @@ -1327,13 +1325,11 @@ int pci_irq_get_node(struct pci_dev *pdev, int vec); #else static inline int pci_msi_vec_count(struct pci_dev *dev) { return -ENOSYS; } -static inline void pci_msi_shutdown(struct pci_dev *dev) { } static inline void pci_disable_msi(struct pci_dev *dev) { } static inline int pci_msix_vec_count(struct pci_dev *dev) { return -ENOSYS; } static inline int pci_enable_msix(struct pci_dev *dev, struct msix_entry *entries, int nvec) { return -ENOSYS; } -static inline void pci_msix_shutdown(struct pci_dev *dev) { } static inline void pci_disable_msix(struct pci_dev *dev) { } static inline void pci_restore_msi_state(struct pci_dev *dev) { } static inline int pci_msi_enabled(void) { return 0; } -- cgit v1.2.3-59-g8ed1b From 50b2b540c00941ced618df9deafccc30826b13b5 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Tue, 14 Mar 2017 15:21:58 +0200 Subject: PCI/PM: Don't sleep at all when d3_delay or d3cold_delay is zero msleep() still sleeps 1 jiffy even when told to sleep for zero milliseconds. That can end up being 1-2 milliseconds or more. In the cases of d3_delay and d3cold_delay, that unnecessarily increases suspend and/or resume latencies. Do not sleep at all for the respective cases if d3_delay is zero or d3cold_delay is zero. Signed-off-by: Adrian Hunter Signed-off-by: Bjorn Helgaas Acked-by: Rafael J. Wysocki --- drivers/pci/pci.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..9779483f81fd 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -66,7 +66,8 @@ static void pci_dev_d3_sleep(struct pci_dev *dev) if (delay < pci_pm_d3_delay) delay = pci_pm_d3_delay; - msleep(delay); + if (delay) + msleep(delay); } #ifdef CONFIG_PCI_DOMAINS @@ -827,7 +828,8 @@ static void __pci_start_power_transition(struct pci_dev *dev, pci_power_t state) * because have already delayed for the bridge. */ if (dev->runtime_d3cold) { - msleep(dev->d3cold_delay); + if (dev->d3cold_delay) + msleep(dev->d3cold_delay); /* * When powering on a bridge from D3cold, the * whole hierarchy may be powered on into -- cgit v1.2.3-59-g8ed1b From f21a8b1b68379c116a6d55821f5f6024ce4f3a48 Mon Sep 17 00:00:00 2001 From: Thomas Petazzoni Date: Tue, 28 Feb 2017 15:31:14 +0100 Subject: PCI: aardvark: Move to MSI handling using generic MSI support The MSI support introduced with the initial Aardvark driver was based on the msi_controller structure and the of_pci_msi_chip_add() / of_pci_find_msi_chip_by_node() API, which are being deprecated in favor of the generic MSI support. Update the Aardvark driver to use the generic MSI support. Signed-off-by: Thomas Petazzoni Signed-off-by: Bjorn Helgaas Acked-by: Marc Zyngier --- drivers/pci/host/pci-aardvark.c | 173 ++++++++++++++++------------------------ 1 file changed, 67 insertions(+), 106 deletions(-) diff --git a/drivers/pci/host/pci-aardvark.c b/drivers/pci/host/pci-aardvark.c index 4fce494271cc..37d0bcd31f8a 100644 --- a/drivers/pci/host/pci-aardvark.c +++ b/drivers/pci/host/pci-aardvark.c @@ -200,10 +200,12 @@ struct advk_pcie { struct list_head resources; struct irq_domain *irq_domain; struct irq_chip irq_chip; - struct msi_controller msi; struct irq_domain *msi_domain; + struct irq_domain *msi_inner_domain; + struct irq_chip msi_bottom_irq_chip; struct irq_chip msi_irq_chip; - DECLARE_BITMAP(msi_irq_in_use, MSI_IRQ_NUM); + struct msi_domain_info msi_domain_info; + DECLARE_BITMAP(msi_used, MSI_IRQ_NUM); struct mutex msi_used_lock; u16 msi_msg; int root_bus_nr; @@ -545,94 +547,64 @@ static struct pci_ops advk_pcie_ops = { .write = advk_pcie_wr_conf, }; -static int advk_pcie_alloc_msi(struct advk_pcie *pcie) +static void advk_msi_irq_compose_msi_msg(struct irq_data *data, + struct msi_msg *msg) { - int hwirq; + struct advk_pcie *pcie = irq_data_get_irq_chip_data(data); + phys_addr_t msi_msg = virt_to_phys(&pcie->msi_msg); - mutex_lock(&pcie->msi_used_lock); - hwirq = find_first_zero_bit(pcie->msi_irq_in_use, MSI_IRQ_NUM); - if (hwirq >= MSI_IRQ_NUM) - hwirq = -ENOSPC; - else - set_bit(hwirq, pcie->msi_irq_in_use); - mutex_unlock(&pcie->msi_used_lock); - - return hwirq; + msg->address_lo = lower_32_bits(msi_msg); + msg->address_hi = upper_32_bits(msi_msg); + msg->data = data->irq; } -static void advk_pcie_free_msi(struct advk_pcie *pcie, int hwirq) +static int advk_msi_set_affinity(struct irq_data *irq_data, + const struct cpumask *mask, bool force) { - struct device *dev = &pcie->pdev->dev; - - mutex_lock(&pcie->msi_used_lock); - if (!test_bit(hwirq, pcie->msi_irq_in_use)) - dev_err(dev, "trying to free unused MSI#%d\n", hwirq); - else - clear_bit(hwirq, pcie->msi_irq_in_use); - mutex_unlock(&pcie->msi_used_lock); + return -EINVAL; } -static int advk_pcie_setup_msi_irq(struct msi_controller *chip, - struct pci_dev *pdev, - struct msi_desc *desc) +static int advk_msi_irq_domain_alloc(struct irq_domain *domain, + unsigned int virq, + unsigned int nr_irqs, void *args) { - struct advk_pcie *pcie = pdev->bus->sysdata; - struct msi_msg msg; - int virq, hwirq; - phys_addr_t msi_msg_phys; - - /* We support MSI, but not MSI-X */ - if (desc->msi_attrib.is_msix) - return -EINVAL; - - hwirq = advk_pcie_alloc_msi(pcie); - if (hwirq < 0) - return hwirq; + struct advk_pcie *pcie = domain->host_data; + int hwirq, i; - virq = irq_create_mapping(pcie->msi_domain, hwirq); - if (!virq) { - advk_pcie_free_msi(pcie, hwirq); - return -EINVAL; + mutex_lock(&pcie->msi_used_lock); + hwirq = bitmap_find_next_zero_area(pcie->msi_used, MSI_IRQ_NUM, + 0, nr_irqs, 0); + if (hwirq >= MSI_IRQ_NUM) { + mutex_unlock(&pcie->msi_used_lock); + return -ENOSPC; } - irq_set_msi_desc(virq, desc); - - msi_msg_phys = virt_to_phys(&pcie->msi_msg); - - msg.address_lo = lower_32_bits(msi_msg_phys); - msg.address_hi = upper_32_bits(msi_msg_phys); - msg.data = virq; - - pci_write_msi_msg(virq, &msg); - - return 0; -} + bitmap_set(pcie->msi_used, hwirq, nr_irqs); + mutex_unlock(&pcie->msi_used_lock); -static void advk_pcie_teardown_msi_irq(struct msi_controller *chip, - unsigned int irq) -{ - struct irq_data *d = irq_get_irq_data(irq); - struct msi_desc *msi = irq_data_get_msi_desc(d); - struct advk_pcie *pcie = msi_desc_to_pci_sysdata(msi); - unsigned long hwirq = d->hwirq; + for (i = 0; i < nr_irqs; i++) + irq_domain_set_info(domain, virq + i, hwirq + i, + &pcie->msi_bottom_irq_chip, + domain->host_data, handle_simple_irq, + NULL, NULL); - irq_dispose_mapping(irq); - advk_pcie_free_msi(pcie, hwirq); + return hwirq; } -static int advk_pcie_msi_map(struct irq_domain *domain, - unsigned int virq, irq_hw_number_t hw) +static void advk_msi_irq_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 advk_pcie *pcie = domain->host_data; - irq_set_chip_and_handler(virq, &pcie->msi_irq_chip, - handle_simple_irq); - - return 0; + mutex_lock(&pcie->msi_used_lock); + bitmap_clear(pcie->msi_used, d->hwirq, nr_irqs); + mutex_unlock(&pcie->msi_used_lock); } -static const struct irq_domain_ops advk_pcie_msi_irq_ops = { - .map = advk_pcie_msi_map, +static const struct irq_domain_ops advk_msi_domain_ops = { + .alloc = advk_msi_irq_domain_alloc, + .free = advk_msi_irq_domain_free, }; static void advk_pcie_irq_mask(struct irq_data *d) @@ -680,30 +652,25 @@ static int advk_pcie_init_msi_irq_domain(struct advk_pcie *pcie) { struct device *dev = &pcie->pdev->dev; struct device_node *node = dev->of_node; - struct irq_chip *msi_irq_chip; - struct msi_controller *msi; + struct irq_chip *bottom_ic, *msi_ic; + struct msi_domain_info *msi_di; phys_addr_t msi_msg_phys; - int ret; - msi_irq_chip = &pcie->msi_irq_chip; + mutex_init(&pcie->msi_used_lock); - msi_irq_chip->name = devm_kasprintf(dev, GFP_KERNEL, "%s-msi", - dev_name(dev)); - if (!msi_irq_chip->name) - return -ENOMEM; + bottom_ic = &pcie->msi_bottom_irq_chip; - msi_irq_chip->irq_enable = pci_msi_unmask_irq; - msi_irq_chip->irq_disable = pci_msi_mask_irq; - msi_irq_chip->irq_mask = pci_msi_mask_irq; - msi_irq_chip->irq_unmask = pci_msi_unmask_irq; + bottom_ic->name = "MSI"; + bottom_ic->irq_compose_msi_msg = advk_msi_irq_compose_msi_msg; + bottom_ic->irq_set_affinity = advk_msi_set_affinity; - msi = &pcie->msi; + msi_ic = &pcie->msi_irq_chip; + msi_ic->name = "advk-MSI"; - msi->setup_irq = advk_pcie_setup_msi_irq; - msi->teardown_irq = advk_pcie_teardown_msi_irq; - msi->of_node = node; - - mutex_init(&pcie->msi_used_lock); + msi_di = &pcie->msi_domain_info; + msi_di->flags = MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS | + MSI_FLAG_MULTI_PCI_MSI; + msi_di->chip = msi_ic; msi_msg_phys = virt_to_phys(&pcie->msi_msg); @@ -712,16 +679,18 @@ static int advk_pcie_init_msi_irq_domain(struct advk_pcie *pcie) advk_writel(pcie, upper_32_bits(msi_msg_phys), PCIE_MSI_ADDR_HIGH_REG); - pcie->msi_domain = + pcie->msi_inner_domain = irq_domain_add_linear(NULL, MSI_IRQ_NUM, - &advk_pcie_msi_irq_ops, pcie); - if (!pcie->msi_domain) + &advk_msi_domain_ops, pcie); + if (!pcie->msi_inner_domain) return -ENOMEM; - ret = of_pci_msi_chip_add(msi); - if (ret < 0) { - irq_domain_remove(pcie->msi_domain); - return ret; + pcie->msi_domain = + pci_msi_create_irq_domain(of_node_to_fwnode(node), + msi_di, pcie->msi_inner_domain); + if (!pcie->msi_domain) { + irq_domain_remove(pcie->msi_inner_domain); + return -ENOMEM; } return 0; @@ -729,8 +698,8 @@ static int advk_pcie_init_msi_irq_domain(struct advk_pcie *pcie) static void advk_pcie_remove_msi_irq_domain(struct advk_pcie *pcie) { - of_pci_msi_chip_remove(&pcie->msi); irq_domain_remove(pcie->msi_domain); + irq_domain_remove(pcie->msi_inner_domain); } static int advk_pcie_init_irq_domain(struct advk_pcie *pcie) @@ -917,8 +886,6 @@ static int advk_pcie_probe(struct platform_device *pdev) struct advk_pcie *pcie; struct resource *res; struct pci_bus *bus, *child; - struct msi_controller *msi; - struct device_node *msi_node; int ret, irq; pcie = devm_kzalloc(dev, sizeof(struct advk_pcie), GFP_KERNEL); @@ -962,14 +929,8 @@ static int advk_pcie_probe(struct platform_device *pdev) return ret; } - msi_node = of_parse_phandle(dev->of_node, "msi-parent", 0); - if (msi_node) - msi = of_pci_find_msi_chip_by_node(msi_node); - else - msi = NULL; - - bus = pci_scan_root_bus_msi(dev, 0, &advk_pcie_ops, - pcie, &pcie->resources, &pcie->msi); + bus = pci_scan_root_bus(dev, 0, &advk_pcie_ops, + pcie, &pcie->resources); if (!bus) { advk_pcie_remove_msi_irq_domain(pcie); advk_pcie_remove_irq_domain(pcie); -- cgit v1.2.3-59-g8ed1b From 393bf9b38ad3f29c6663e8e76fce4f034e6ef673 Mon Sep 17 00:00:00 2001 From: Thomas Petazzoni Date: Tue, 28 Feb 2017 15:31:15 +0100 Subject: PCI: mvebu: Remove useless MSI enabling code Since commit fcc392d501bd2 ("irqchip/armada-370-xp: Use the generic MSI infrastructure"), the irqchip driver used on Armada 370, XP, 375, 38x, 39x for the MPIC interrupt controller has been converted to use the generic MSI infrastructure. Since this commit, it is no longer registering an msi_controller structure with the of_pci_msi_chip_add() function. Therefore, having the PCI driver used on the same platform calling of_pci_find_msi_chip_by_node() is pretty useless. The MSI resolution is now done in the generic interrupt resolution code, since the MSI controller is an irq domain attached to the interrupt controller node, which is pointed to by the msi-parent DT property in the PCIe controller node. Signed-off-by: Thomas Petazzoni Signed-off-by: Bjorn Helgaas Acked-by: Marc Zyngier --- drivers/pci/host/pci-mvebu.c | 17 ----------------- 1 file changed, 17 deletions(-) diff --git a/drivers/pci/host/pci-mvebu.c b/drivers/pci/host/pci-mvebu.c index cd7d51988738..4e4e591d19de 100644 --- a/drivers/pci/host/pci-mvebu.c +++ b/drivers/pci/host/pci-mvebu.c @@ -1005,22 +1005,6 @@ static int mvebu_get_tgt_attr(struct device_node *np, int devfn, return -ENOENT; } -static void mvebu_pcie_msi_enable(struct mvebu_pcie *pcie) -{ - struct device_node *msi_node; - - msi_node = of_parse_phandle(pcie->pdev->dev.of_node, - "msi-parent", 0); - if (!msi_node) - return; - - pcie->msi = of_pci_find_msi_chip_by_node(msi_node); - of_node_put(msi_node); - - if (pcie->msi) - pcie->msi->dev = &pcie->pdev->dev; -} - #ifdef CONFIG_PM_SLEEP static int mvebu_pcie_suspend(struct device *dev) { @@ -1298,7 +1282,6 @@ static int mvebu_pcie_probe(struct platform_device *pdev) for (i = 0; i < (IO_SPACE_LIMIT - SZ_64K); i += SZ_64K) pci_ioremap_io(i, pcie->io.start + i); - mvebu_pcie_msi_enable(pcie); mvebu_pcie_enable(pcie); platform_set_drvdata(pdev, pcie); -- cgit v1.2.3-59-g8ed1b From a26356ab9392e0c5f8ad87d76c42e7c58c036d24 Mon Sep 17 00:00:00 2001 From: Thomas Petazzoni Date: Tue, 28 Feb 2017 15:31:16 +0100 Subject: of/pci: Remove unused MSI controller helpers All users of the small MSI controller API have been migrated to use the generic MSI infrastructure instead. We no longer need a global chained list of msi_controller. Instead, MSI controllers are now represented as IRQ domains attached to OF nodes, and the resolution between a device requesting an MSI and the corresponding MSI controller is done by the generic interrupt resolution logic. Therefore, this API is now completely useless, and can be removed from the kernel. Signed-off-by: Thomas Petazzoni Signed-off-by: Bjorn Helgaas Acked-by: Marc Zyngier Acked-by: Rob Herring --- drivers/of/of_pci.c | 45 --------------------------------------------- include/linux/of_pci.h | 11 ----------- 2 files changed, 56 deletions(-) diff --git a/drivers/of/of_pci.c b/drivers/of/of_pci.c index 0ee42c3e66a1..c9d4d3a7b0fe 100644 --- a/drivers/of/of_pci.c +++ b/drivers/of/of_pci.c @@ -285,51 +285,6 @@ parse_failed: EXPORT_SYMBOL_GPL(of_pci_get_host_bridge_resources); #endif /* CONFIG_OF_ADDRESS */ -#ifdef CONFIG_PCI_MSI - -static LIST_HEAD(of_pci_msi_chip_list); -static DEFINE_MUTEX(of_pci_msi_chip_mutex); - -int of_pci_msi_chip_add(struct msi_controller *chip) -{ - if (!of_property_read_bool(chip->of_node, "msi-controller")) - return -EINVAL; - - mutex_lock(&of_pci_msi_chip_mutex); - list_add(&chip->list, &of_pci_msi_chip_list); - mutex_unlock(&of_pci_msi_chip_mutex); - - return 0; -} -EXPORT_SYMBOL_GPL(of_pci_msi_chip_add); - -void of_pci_msi_chip_remove(struct msi_controller *chip) -{ - mutex_lock(&of_pci_msi_chip_mutex); - list_del(&chip->list); - mutex_unlock(&of_pci_msi_chip_mutex); -} -EXPORT_SYMBOL_GPL(of_pci_msi_chip_remove); - -struct msi_controller *of_pci_find_msi_chip_by_node(struct device_node *of_node) -{ - struct msi_controller *c; - - mutex_lock(&of_pci_msi_chip_mutex); - list_for_each_entry(c, &of_pci_msi_chip_list, list) { - if (c->of_node == of_node) { - mutex_unlock(&of_pci_msi_chip_mutex); - return c; - } - } - mutex_unlock(&of_pci_msi_chip_mutex); - - return NULL; -} -EXPORT_SYMBOL_GPL(of_pci_find_msi_chip_by_node); - -#endif /* CONFIG_PCI_MSI */ - /** * of_pci_map_rid - Translate a requester ID through a downstream mapping. * @np: root complex device node. diff --git a/include/linux/of_pci.h b/include/linux/of_pci.h index 0e0974eceb80..518c8d20647a 100644 --- a/include/linux/of_pci.h +++ b/include/linux/of_pci.h @@ -85,15 +85,4 @@ static inline int of_pci_get_host_bridge_resources(struct device_node *dev, } #endif -#if defined(CONFIG_OF) && defined(CONFIG_PCI_MSI) -int of_pci_msi_chip_add(struct msi_controller *chip); -void of_pci_msi_chip_remove(struct msi_controller *chip); -struct msi_controller *of_pci_find_msi_chip_by_node(struct device_node *of_node); -#else -static inline int of_pci_msi_chip_add(struct msi_controller *chip) { return -EINVAL; } -static inline void of_pci_msi_chip_remove(struct msi_controller *chip) { } -static inline struct msi_controller * -of_pci_find_msi_chip_by_node(struct device_node *of_node) { return NULL; } -#endif - #endif -- cgit v1.2.3-59-g8ed1b From e584b06c4730f1b6d9ccb47fc8b7d79411eb9012 Mon Sep 17 00:00:00 2001 From: Jon Mason Date: Wed, 1 Mar 2017 10:53:13 -0500 Subject: PCI: iproc: Add PCI_DOMAIN dependency to PCI Kconfig 2+ PCI devices fail to be discovered due to each bus having the same PCI domain. This is because the domain defined in the device tree file is not being added due to PCI_DOMAIN not being enabled. So, every PCI bus has a domain of zero. When PCI_DOMAIN is selected by the Kconfig, it picks up the domain defined in the device tree file and everything works as expected. Since both PCIE_IPROC_PLATFORM and PCIE_IPROC_BCMA need PCI_DOMAIN, move it to PCIE_IPROC so it will be automatically selected for both. Signed-off-by: Jon Mason Signed-off-by: Bjorn Helgaas --- drivers/pci/host/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig index f7c1d4d5c665..a35d4588cfb3 100644 --- a/drivers/pci/host/Kconfig +++ b/drivers/pci/host/Kconfig @@ -95,6 +95,7 @@ config PCI_VERSATILE config PCIE_IPROC tristate + select PCI_DOMAINS help This enables the iProc PCIe core controller support for Broadcom's iProc family of SoCs. An appropriate bus interface driver needs @@ -115,7 +116,6 @@ config PCIE_IPROC_BCMA depends on ARM && (ARCH_BCM_IPROC || COMPILE_TEST) select PCIE_IPROC select BCMA - select PCI_DOMAINS default ARCH_BCM_5301X help Say Y here if you want to use the Broadcom iProc PCIe controller -- cgit v1.2.3-59-g8ed1b From 41334f54a43ab00cbb294e6a08d0f57068f43025 Mon Sep 17 00:00:00 2001 From: Marc Gonzalez Date: Fri, 17 Mar 2017 15:53:19 +0100 Subject: PCI: Include pci.h for struct pci_ops definition struct pci_ecam_ops embeds a struct pci_ops. Explicitly request the definition for struct pci_ops, otherwise gcc might complain: include/linux/pci-ecam.h:29:19: error: field 'pci_ops' has incomplete type Signed-off-by: Marc Gonzalez Signed-off-by: Bjorn Helgaas --- include/linux/pci-ecam.h | 1 + 1 file changed, 1 insertion(+) diff --git a/include/linux/pci-ecam.h b/include/linux/pci-ecam.h index f0d2b9451270..b8f11d783a11 100644 --- a/include/linux/pci-ecam.h +++ b/include/linux/pci-ecam.h @@ -16,6 +16,7 @@ #ifndef DRIVERS_PCI_ECAM_H #define DRIVERS_PCI_ECAM_H +#include #include #include -- cgit v1.2.3-59-g8ed1b From 4bb6669147d9037447baeb633deca8a64e1a55dd Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 23 Mar 2017 12:29:56 -0500 Subject: PCI/MSI: Use dev_printk() when possible Use dev_printk() when possible. This makes messages more consistent with other device-related messages and, in some cases, adds useful information. This changes messages like this: Unable to allocate affinity masks, ignoring to this: pci 0000:01:00.0: can't allocate MSI affinity masks for 4 vectors Signed-off-by: Bjorn Helgaas --- drivers/pci/msi.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c index 4d062c3bf5f0..c4f68e45eadd 100644 --- a/drivers/pci/msi.c +++ b/drivers/pci/msi.c @@ -541,7 +541,8 @@ msi_setup_entry(struct pci_dev *dev, int nvec, const struct irq_affinity *affd) if (affd) { masks = irq_create_affinity_masks(nvec, affd); if (!masks) - pr_err("Unable to allocate affinity masks, ignoring\n"); + dev_err(&dev->dev, "can't allocate MSI affinity masks for %d vectors\n", + nvec); } /* MSI Entry Initialization */ @@ -681,7 +682,8 @@ static int msix_setup_entries(struct pci_dev *dev, void __iomem *base, if (affd) { masks = irq_create_affinity_masks(nvec, affd); if (!masks) - pr_err("Unable to allocate affinity masks, ignoring\n"); + dev_err(&dev->dev, "can't allocate MSI-X affinity masks for %d vectors\n", + nvec); } for (i = 0, curmsk = masks; i < nvec; i++) { -- cgit v1.2.3-59-g8ed1b From 5fcaa0051dfcc5fe77aee95061945d80e465a027 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 9 Mar 2017 18:46:13 -0800 Subject: PCI: rockchip: Handle regulator_get_current_limit() failure correctly regulator_get_current_limit() can return negative error codes. We saved the return value in an unsigned "curr", and a subsequent check interpreted a negative error code as a positive (invalid) current limit. Save the return code as a signed value, which avoids messages like this, seen on Samsung Chromebook Plus: rockchip-pcie f8000000.pcie: invalid power supply [bhelgaas: changelog] Fixes: 4816c4c7b82b ("PCI: rockchip: Provide captured slot power limit and scale") Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas Acked-by: Shawn Lin --- drivers/pci/host/pcie-rockchip.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index 26ddd3535272..d785f64ec03b 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -425,7 +425,8 @@ static struct pci_ops rockchip_pcie_ops = { static void rockchip_pcie_set_power_limit(struct rockchip_pcie *rockchip) { - u32 status, curr, scale, power; + int curr; + u32 status, scale, power; if (IS_ERR(rockchip->vpcie3v3)) return; -- cgit v1.2.3-59-g8ed1b From 73edd2b180a18024605c599074596a9e22d745d6 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 23 Mar 2017 17:21:26 -0500 Subject: PCI: rockchip: Unindent rockchip_pcie_set_power_limit() If regulator_get_current_limit() returns 0 or error, return early so the body of the function doesn't have to be indented as the body of an "if" statement. No functional change intended. Signed-off-by: Bjorn Helgaas --- drivers/pci/host/pcie-rockchip.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index d785f64ec03b..7f76ff6af0ba 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -438,24 +438,25 @@ static void rockchip_pcie_set_power_limit(struct rockchip_pcie *rockchip) * to the actual power supply. */ curr = regulator_get_current_limit(rockchip->vpcie3v3); - if (curr > 0) { - scale = 3; /* 0.001x */ - curr = curr / 1000; /* convert to mA */ - power = (curr * 3300) / 1000; /* milliwatt */ - while (power > PCIE_RC_CONFIG_DCR_CSPL_LIMIT) { - if (!scale) { - dev_warn(rockchip->dev, "invalid power supply\n"); - return; - } - scale--; - power = power / 10; - } + if (curr <= 0) + return; - status = rockchip_pcie_read(rockchip, PCIE_RC_CONFIG_DCR); - status |= (power << PCIE_RC_CONFIG_DCR_CSPL_SHIFT) | - (scale << PCIE_RC_CONFIG_DCR_CPLS_SHIFT); - rockchip_pcie_write(rockchip, status, PCIE_RC_CONFIG_DCR); + scale = 3; /* 0.001x */ + curr = curr / 1000; /* convert to mA */ + power = (curr * 3300) / 1000; /* milliwatt */ + while (power > PCIE_RC_CONFIG_DCR_CSPL_LIMIT) { + if (!scale) { + dev_warn(rockchip->dev, "invalid power supply\n"); + return; + } + scale--; + power = power / 10; } + + status = rockchip_pcie_read(rockchip, PCIE_RC_CONFIG_DCR); + status |= (power << PCIE_RC_CONFIG_DCR_CSPL_SHIFT) | + (scale << PCIE_RC_CONFIG_DCR_CPLS_SHIFT); + rockchip_pcie_write(rockchip, status, PCIE_RC_CONFIG_DCR); } /** -- cgit v1.2.3-59-g8ed1b From deb518f6ae4d5da22fb1e8d6c6a690253ffa700d Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 9 Mar 2017 18:46:14 -0800 Subject: PCI: rockchip: Make 'return 0' more obvious in probe() There's no way to get here with 'err != 0'. Just return 0 to be more obvious and prevent future changes from accidentally erroring out here without going through the right error paths. Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas --- drivers/pci/host/pcie-rockchip.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index 7f76ff6af0ba..b4dfe3bd15c3 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -1399,7 +1399,7 @@ static int rockchip_pcie_probe(struct platform_device *pdev) pcie_bus_configure_settings(child); pci_bus_add_devices(bus); - return err; + return 0; err_free_res: pci_free_resource_list(&res); -- cgit v1.2.3-59-g8ed1b From d3a78d8bf759d8848339dcc367c4c1678b57a08b Mon Sep 17 00:00:00 2001 From: Long Li Date: Thu, 23 Mar 2017 14:58:10 -0700 Subject: PCI: hv: Properly handle PCI bus remove hv_pci_devices_present() is called in hv_pci_remove() when we remove a PCI device from the host, e.g., by disabling SR-IOV on a device. In hv_pci_remove(), the bus is already removed before the call, so we don't need to rescan the bus in the workqueue scheduled from hv_pci_devices_present(). By introducing bus state hv_pcibus_removed, we can avoid this situation. Reported-by: Xiaofeng Wang Signed-off-by: Long Li Signed-off-by: Bjorn Helgaas Acked-by: K. Y. Srinivasan --- drivers/pci/host/pci-hyperv.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c index ada98569b78e..39fafda4deff 100644 --- a/drivers/pci/host/pci-hyperv.c +++ b/drivers/pci/host/pci-hyperv.c @@ -350,6 +350,7 @@ enum hv_pcibus_state { hv_pcibus_init = 0, hv_pcibus_probed, hv_pcibus_installed, + hv_pcibus_removed, hv_pcibus_maximum }; @@ -1504,13 +1505,24 @@ static void pci_devices_present_work(struct work_struct *work) put_pcichild(hpdev, hv_pcidev_ref_initial); } - /* Tell the core to rescan bus because there may have been changes. */ - if (hbus->state == hv_pcibus_installed) { + switch(hbus->state) { + case hv_pcibus_installed: + /* + * Tell the core to rescan bus + * because there may have been changes. + */ pci_lock_rescan_remove(); pci_scan_child_bus(hbus->pci_bus); pci_unlock_rescan_remove(); - } else { + break; + + case hv_pcibus_init: + case hv_pcibus_probed: survey_child_resources(hbus); + break; + + default: + break; } up(&hbus->enum_sem); @@ -2185,6 +2197,7 @@ static int hv_pci_probe(struct hv_device *hdev, hbus = kzalloc(sizeof(*hbus), GFP_KERNEL); if (!hbus) return -ENOMEM; + hbus->state = hv_pcibus_init; /* * The PCI bus "domain" is what is called "segment" in ACPI and @@ -2348,6 +2361,7 @@ static int hv_pci_remove(struct hv_device *hdev) pci_stop_root_bus(hbus->pci_bus); pci_remove_root_bus(hbus->pci_bus); pci_unlock_rescan_remove(); + hbus->state = hv_pcibus_removed; } hv_pci_bus_exit(hdev); -- cgit v1.2.3-59-g8ed1b From 414428c5da1c71986727c2fa5cdf1ed071e398d7 Mon Sep 17 00:00:00 2001 From: Long Li Date: Thu, 23 Mar 2017 14:58:32 -0700 Subject: PCI: hv: Lock PCI bus on device eject A PCI_EJECT message can arrive at the same time we are calling pci_scan_child_bus() in the workqueue for the previous PCI_BUS_RELATIONS message or in create_root_hv_pci_bus(). In this case we could potentially modify the bus from multiple places. Properly lock the bus access. Thanks Dexuan Cui for pointing out the race condition in create_root_hv_pci_bus(). Reported-by: Xiaofeng Wang Signed-off-by: Long Li Signed-off-by: Bjorn Helgaas Acked-by: K. Y. Srinivasan --- drivers/pci/host/pci-hyperv.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c index 39fafda4deff..a1b3c1957d43 100644 --- a/drivers/pci/host/pci-hyperv.c +++ b/drivers/pci/host/pci-hyperv.c @@ -1209,9 +1209,11 @@ static int create_root_hv_pci_bus(struct hv_pcibus_device *hbus) hbus->pci_bus->msi = &hbus->msi_chip; hbus->pci_bus->msi->dev = &hbus->hdev->device; + pci_lock_rescan_remove(); pci_scan_child_bus(hbus->pci_bus); pci_bus_assign_resources(hbus->pci_bus); pci_bus_add_devices(hbus->pci_bus); + pci_unlock_rescan_remove(); hbus->state = hv_pcibus_installed; return 0; } @@ -1612,8 +1614,10 @@ static void hv_eject_device_work(struct work_struct *work) pdev = pci_get_domain_bus_and_slot(hpdev->hbus->sysdata.domain, 0, wslot); if (pdev) { + pci_lock_rescan_remove(); pci_stop_and_remove_bus_device(pdev); pci_dev_put(pdev); + pci_unlock_rescan_remove(); } spin_lock_irqsave(&hpdev->hbus->device_list_lock, flags); -- cgit v1.2.3-59-g8ed1b From 1f12d3c11f4952295070f56fba6095b896399007 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Sun, 12 Mar 2017 23:23:52 +0100 Subject: PCI: Add DT bindings for Faraday Technology PCI Host Bridge Add device tree bindings for the Faraday technology PCI Host Bridge. This IP is found in the Storlink/Storm/Cortina Gemini SoC platform. Signed-off-by: Linus Walleij Signed-off-by: Bjorn Helgaas Acked-by: Rob Herring CC: Janos Laube CC: Paulius Zaleckas CC: Hans Ulli Kroll CC: Florian Fainelli CC: devicetree@vger.kernel.org CC: Feng-Hsin Chiang CC: Greentime Hu --- .../devicetree/bindings/pci/faraday,ftpci100.txt | 129 +++++++++++++++++++++ 1 file changed, 129 insertions(+) create mode 100644 Documentation/devicetree/bindings/pci/faraday,ftpci100.txt diff --git a/Documentation/devicetree/bindings/pci/faraday,ftpci100.txt b/Documentation/devicetree/bindings/pci/faraday,ftpci100.txt new file mode 100644 index 000000000000..35d4a979bb7b --- /dev/null +++ b/Documentation/devicetree/bindings/pci/faraday,ftpci100.txt @@ -0,0 +1,129 @@ +Faraday Technology FTPCI100 PCI Host Bridge + +This PCI bridge is found inside that Cortina Systems Gemini SoC platform and +is a generic IP block from Faraday Technology. It exists in two variants: +plain and dual PCI. The plain version embeds a cascading interrupt controller +into the host bridge. The dual version routes the interrupts to the host +chips interrupt controller. + +The host controller appear on the PCI bus with vendor ID 0x159b (Faraday +Technology) and product ID 0x4321. + +Mandatory properties: + +- compatible: ranging from specific to generic, should be one of + "cortina,gemini-pci", "faraday,ftpci100" + "cortina,gemini-pci-dual", "faraday,ftpci100-dual" + "faraday,ftpci100" + "faraday,ftpci100-dual" +- reg: memory base and size for the host bridge +- #address-cells: set to <3> +- #size-cells: set to <2> +- #interrupt-cells: set to <1> +- bus-range: set to <0x00 0xff> +- device_type, set to "pci" +- ranges: see pci.txt +- interrupt-map-mask: see pci.txt +- interrupt-map: see pci.txt +- dma-ranges: three ranges for the inbound memory region. The ranges must + be aligned to a 1MB boundary, and may be 1MB, 2MB, 4MB, 8MB, 16MB, 32MB, 64MB, + 128MB, 256MB, 512MB, 1GB or 2GB in size. The memory should be marked as + pre-fetchable. + +Mandatory subnodes: +- For "faraday,ftpci100" a node representing the interrupt-controller inside the + host bridge is mandatory. It has the following mandatory properties: + - interrupt: see interrupt-controller/interrupts.txt + - interrupt-parent: see interrupt-controller/interrupts.txt + - interrupt-controller: see interrupt-controller/interrupts.txt + - #address-cells: set to <0> + - #interrupt-cells: set to <1> + +I/O space considerations: + +The plain variant has 128MiB of non-prefetchable memory space, whereas the +"dual" variant has 64MiB. Take this into account when describing the ranges. + +Interrupt map considerations: + +The "dual" variant will get INT A, B, C, D from the system interrupt controller +and should point to respective interrupt in that controller in its +interrupt-map. + +The code which is the only documentation of how the Faraday PCI (the non-dual +variant) interrupts assigns the default interrupt mapping/swizzling has +typically been like this, doing the swizzling on the interrupt controller side +rather than in the interconnect: + +interrupt-map-mask = <0xf800 0 0 7>; +interrupt-map = + <0x4800 0 0 1 &pci_intc 0>, /* Slot 9 */ + <0x4800 0 0 2 &pci_intc 1>, + <0x4800 0 0 3 &pci_intc 2>, + <0x4800 0 0 4 &pci_intc 3>, + <0x5000 0 0 1 &pci_intc 1>, /* Slot 10 */ + <0x5000 0 0 2 &pci_intc 2>, + <0x5000 0 0 3 &pci_intc 3>, + <0x5000 0 0 4 &pci_intc 0>, + <0x5800 0 0 1 &pci_intc 2>, /* Slot 11 */ + <0x5800 0 0 2 &pci_intc 3>, + <0x5800 0 0 3 &pci_intc 0>, + <0x5800 0 0 4 &pci_intc 1>, + <0x6000 0 0 1 &pci_intc 3>, /* Slot 12 */ + <0x6000 0 0 2 &pci_intc 0>, + <0x6000 0 0 3 &pci_intc 1>, + <0x6000 0 0 4 &pci_intc 2>; + +Example: + +pci@50000000 { + compatible = "cortina,gemini-pci", "faraday,ftpci100"; + reg = <0x50000000 0x100>; + interrupts = <8 IRQ_TYPE_LEVEL_HIGH>, /* PCI A */ + <26 IRQ_TYPE_LEVEL_HIGH>, /* PCI B */ + <27 IRQ_TYPE_LEVEL_HIGH>, /* PCI C */ + <28 IRQ_TYPE_LEVEL_HIGH>; /* PCI D */ + #address-cells = <3>; + #size-cells = <2>; + #interrupt-cells = <1>; + + bus-range = <0x00 0xff>; + ranges = /* 1MiB I/O space 0x50000000-0x500fffff */ + <0x01000000 0 0 0x50000000 0 0x00100000>, + /* 128MiB non-prefetchable memory 0x58000000-0x5fffffff */ + <0x02000000 0 0x58000000 0x58000000 0 0x08000000>; + + /* DMA ranges */ + dma-ranges = + /* 128MiB at 0x00000000-0x07ffffff */ + <0x02000000 0 0x00000000 0x00000000 0 0x08000000>, + /* 64MiB at 0x00000000-0x03ffffff */ + <0x02000000 0 0x00000000 0x00000000 0 0x04000000>, + /* 64MiB at 0x00000000-0x03ffffff */ + <0x02000000 0 0x00000000 0x00000000 0 0x04000000>; + + interrupt-map-mask = <0xf800 0 0 7>; + interrupt-map = + <0x4800 0 0 1 &pci_intc 0>, /* Slot 9 */ + <0x4800 0 0 2 &pci_intc 1>, + <0x4800 0 0 3 &pci_intc 2>, + <0x4800 0 0 4 &pci_intc 3>, + <0x5000 0 0 1 &pci_intc 1>, /* Slot 10 */ + <0x5000 0 0 2 &pci_intc 2>, + <0x5000 0 0 3 &pci_intc 3>, + <0x5000 0 0 4 &pci_intc 0>, + <0x5800 0 0 1 &pci_intc 2>, /* Slot 11 */ + <0x5800 0 0 2 &pci_intc 3>, + <0x5800 0 0 3 &pci_intc 0>, + <0x5800 0 0 4 &pci_intc 1>, + <0x6000 0 0 1 &pci_intc 3>, /* Slot 12 */ + <0x6000 0 0 2 &pci_intc 0>, + <0x6000 0 0 3 &pci_intc 0>, + <0x6000 0 0 4 &pci_intc 0>; + pci_intc: interrupt-controller { + interrupt-parent = <&intcon>; + interrupt-controller; + #address-cells = <0>; + #interrupt-cells = <1>; + }; +}; -- cgit v1.2.3-59-g8ed1b From d3c68e0a7e34af2b86f5408c4d6e794a38bbd9e1 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Sun, 12 Mar 2017 23:24:03 +0100 Subject: PCI: faraday: Add Faraday Technology FTPCI100 PCI Host Bridge driver Add a host bridge driver for the Faraday Technology FPPCI100 host bridge, used for Cortina Systems Gemini SoC (SL3516) PCI Host Bridge. This code is inspired by the out-of-tree OpenWRT patch and then extensively rewritten for device tree and using the modern helpers to cut down and modernize the code to all new PCI frameworks. A driver exists in U-Boot as well. Tested on the ITian Square One SQ201 NAS with the following result in the boot log (trimmed to relevant parts): OF: PCI: host bridge /soc/pci@50000000 ranges: OF: PCI: IO 0x50000000..0x500fffff -> 0x00000000 OF: PCI: MEM 0x58000000..0x5fffffff -> 0x58000000 ftpci100 50000000.pci: PCI host bridge to bus 0000:00 pci_bus 0000:00: root bus resource [bus 00-ff] pci_bus 0000:00: root bus resource [io 0x0000-0xfffff] pci_bus 0000:00: root bus resource [mem 0x58000000-0x5fffffff] ftpci100 50000000.pci: DMA MEM1 BASE: 0x0000000000000000 -> 0x0000000007ffffff config 00070000 ftpci100 50000000.pci: DMA MEM2 BASE: 0x0000000000000000 -> 0x0000000003ffffff config 00060000 ftpci100 50000000.pci: DMA MEM3 BASE: 0x0000000000000000 -> 0x0000000003ffffff config 00060000 PCI: bus0: Fast back to back transfers disabled pci 0000:00:00.0: of_irq_parse_pci() failed with rc=-22 pci 0000:00:0c.0: BAR 0: assigned [mem 0x58000000-0x58007fff] pci 0000:00:09.2: BAR 0: assigned [mem 0x58008000-0x580080ff] pci 0000:00:09.0: BAR 4: assigned [io 0x1000-0x101f] pci 0000:00:09.1: BAR 4: assigned [io 0x1020-0x103f] pci 0000:00:09.0: enabling device (0140 -> 0141) pci 0000:00:09.0: HCRESET not completed yet! pci 0000:00:09.1: enabling device (0140 -> 0141) pci 0000:00:09.1: HCRESET not completed yet! pci 0000:00:09.2: enabling device (0140 -> 0142) rt61pci 0000:00:0c.0: enabling device (0140 -> 0142) ieee80211 phy0: rt2x00_set_chip: Info - Chipset detected - rt: 2561, rf: 0003, rev: 000c ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver ehci-pci: EHCI PCI platform driver ehci-pci 0000:00:09.2: EHCI Host Controller ehci-pci 0000:00:09.2: new USB bus registered, assigned bus number 1 ehci-pci 0000:00:09.2: irq 125, io mem 0x58008000 ehci-pci 0000:00:09.2: USB 2.0 started, EHCI 1.00 hub 1-0:1.0: USB hub found hub 1-0:1.0: 4 ports detected uhci_hcd: USB Universal Host Controller Interface driver uhci_hcd 0000:00:09.0: UHCI Host Controller uhci_hcd 0000:00:09.0: new USB bus registered, assigned bus number 2 uhci_hcd 0000:00:09.0: HCRESET not completed yet! uhci_hcd 0000:00:09.0: irq 123, io base 0x00001000 hub 2-0:1.0: USB hub found hub 2-0:1.0: config failed, hub doesn't have any ports! (err -19) uhci_hcd 0000:00:09.1: UHCI Host Controller uhci_hcd 0000:00:09.1: new USB bus registered, assigned bus number 3 uhci_hcd 0000:00:09.1: HCRESET not completed yet! uhci_hcd 0000:00:09.1: irq 124, io base 0x00001020 hub 3-0:1.0: USB hub found hub 3-0:1.0: config failed, hub doesn't have any ports! (err -19) scsi 0:0:0:0: Direct-Access USB Flash Disk 1.00 PQ: 0 ANSI: 2 sd 0:0:0:0: [sda] 7900336 512-byte logical blocks: (4.04 GB/3.77 GiB) sd 0:0:0:0: [sda] Write Protect is off sd 0:0:0:0: [sda] No Caching mode page found sd 0:0:0:0: [sda] Assuming drive cache: write through sda: sda1 sda2 sda3 sd 0:0:0:0: [sda] Attached SCSI removable disk ieee80211 phy0: rt2x00lib_request_firmware: Info - Loading firmware file 'rt2561s.bin' ieee80211 phy0: rt2x00lib_request_firmware: Info - Firmware detected - version: 0.8 IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready $ lspci 00:00.0 Class 0600: 159b:4321 00:09.2 Class 0c03: 1106:3104 00:09.0 Class 0c03: 1106:3038 00:09.1 Class 0c03: 1106:3038 00:0c.0 Class 0280: 1814:0301 $ cat /proc/interrupts CPU0 123: 0 PCI 0 Edge uhci_hcd:usb2 124: 0 PCI 1 Edge uhci_hcd:usb3 125: 159 PCI 2 Edge ehci_hcd:usb1 126: 1082 PCI 3 Edge rt61pci $ cat /proc/iomem 50000000-500000ff : /soc/pci@50000000 58000000-5fffffff : Gemini PCI MEM 58000000-58007fff : 0000:00:0c.0 58000000-58007fff : 0000:00:0c.0 58008000-580080ff : 0000:00:09.2 58008000-580080ff : ehci_hcd The EHCI USB hub works fine; I can mount and manage files and the IRQs just keep ticking up. I can issue iwlist wlan0 scanning and see all the WLANs here. I don't have wpa_supplicant so have not tried connecting to them. [bhelgaas: fold in %pap change from Arnd Bergmann ] Signed-off-by: Linus Walleij Signed-off-by: Bjorn Helgaas CC: Janos Laube CC: Paulius Zaleckas CC: Hans Ulli Kroll CC: Florian Fainelli CC: Feng-Hsin Chiang CC: Greentime Hu --- drivers/pci/host/Kconfig | 6 + drivers/pci/host/Makefile | 1 + drivers/pci/host/pci-ftpci100.c | 562 ++++++++++++++++++++++++++++++++++++++++ 3 files changed, 569 insertions(+) create mode 100644 drivers/pci/host/pci-ftpci100.c diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig index f7c1d4d5c665..aa23b70d61eb 100644 --- a/drivers/pci/host/Kconfig +++ b/drivers/pci/host/Kconfig @@ -27,6 +27,12 @@ config PCIE_XILINX_NWL or End Point. The current option selection will only support root port enabling. +config PCI_FTPCI100 + bool "Faraday Technology FTPCI100 PCI controller" + depends on OF + depends on ARM + default ARCH_GEMINI + config PCI_TEGRA bool "NVIDIA Tegra PCIe controller" depends on ARCH_TEGRA diff --git a/drivers/pci/host/Makefile b/drivers/pci/host/Makefile index 4d3686676cc3..cab879578003 100644 --- a/drivers/pci/host/Makefile +++ b/drivers/pci/host/Makefile @@ -1,3 +1,4 @@ +obj-$(CONFIG_PCI_FTPCI100) += pci-ftpci100.o obj-$(CONFIG_PCI_HYPERV) += pci-hyperv.o obj-$(CONFIG_PCI_MVEBU) += pci-mvebu.o obj-$(CONFIG_PCI_AARDVARK) += pci-aardvark.o diff --git a/drivers/pci/host/pci-ftpci100.c b/drivers/pci/host/pci-ftpci100.c new file mode 100644 index 000000000000..c0f29d1de341 --- /dev/null +++ b/drivers/pci/host/pci-ftpci100.c @@ -0,0 +1,562 @@ +/* + * Support for Faraday Technology FTPC100 PCI Controller + * + * Copyright (C) 2017 Linus Walleij + * + * Based on the out-of-tree OpenWRT patch for Cortina Gemini: + * Copyright (C) 2009 Janos Laube + * Copyright (C) 2009 Paulius Zaleckas + * Based on SL2312 PCI controller code + * Storlink (C) 2003 + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* + * Special configuration registers directly in the first few words + * in I/O space. + */ +#define PCI_IOSIZE 0x00 +#define PCI_PROT 0x04 /* AHB protection */ +#define PCI_CTRL 0x08 /* PCI control signal */ +#define PCI_SOFTRST 0x10 /* Soft reset counter and response error enable */ +#define PCI_CONFIG 0x28 /* PCI configuration command register */ +#define PCI_DATA 0x2C + +#define FARADAY_PCI_PMC 0x40 /* Power management control */ +#define FARADAY_PCI_PMCSR 0x44 /* Power management status */ +#define FARADAY_PCI_CTRL1 0x48 /* Control register 1 */ +#define FARADAY_PCI_CTRL2 0x4C /* Control register 2 */ +#define FARADAY_PCI_MEM1_BASE_SIZE 0x50 /* Memory base and size #1 */ +#define FARADAY_PCI_MEM2_BASE_SIZE 0x54 /* Memory base and size #2 */ +#define FARADAY_PCI_MEM3_BASE_SIZE 0x58 /* Memory base and size #3 */ + +/* Bits 31..28 gives INTD..INTA status */ +#define PCI_CTRL2_INTSTS_SHIFT 28 +#define PCI_CTRL2_INTMASK_CMDERR BIT(27) +#define PCI_CTRL2_INTMASK_PARERR BIT(26) +/* Bits 25..22 masks INTD..INTA */ +#define PCI_CTRL2_INTMASK_SHIFT 22 +#define PCI_CTRL2_INTMASK_MABRT_RX BIT(21) +#define PCI_CTRL2_INTMASK_TABRT_RX BIT(20) +#define PCI_CTRL2_INTMASK_TABRT_TX BIT(19) +#define PCI_CTRL2_INTMASK_RETRY4 BIT(18) +#define PCI_CTRL2_INTMASK_SERR_RX BIT(17) +#define PCI_CTRL2_INTMASK_PERR_RX BIT(16) +/* Bit 15 reserved */ +#define PCI_CTRL2_MSTPRI_REQ6 BIT(14) +#define PCI_CTRL2_MSTPRI_REQ5 BIT(13) +#define PCI_CTRL2_MSTPRI_REQ4 BIT(12) +#define PCI_CTRL2_MSTPRI_REQ3 BIT(11) +#define PCI_CTRL2_MSTPRI_REQ2 BIT(10) +#define PCI_CTRL2_MSTPRI_REQ1 BIT(9) +#define PCI_CTRL2_MSTPRI_REQ0 BIT(8) +/* Bits 7..4 reserved */ +/* Bits 3..0 TRDYW */ + +/* + * Memory configs: + * Bit 31..20 defines the PCI side memory base + * Bit 19..16 (4 bits) defines the size per below + */ +#define FARADAY_PCI_MEMBASE_MASK 0xfff00000 +#define FARADAY_PCI_MEMSIZE_1MB 0x0 +#define FARADAY_PCI_MEMSIZE_2MB 0x1 +#define FARADAY_PCI_MEMSIZE_4MB 0x2 +#define FARADAY_PCI_MEMSIZE_8MB 0x3 +#define FARADAY_PCI_MEMSIZE_16MB 0x4 +#define FARADAY_PCI_MEMSIZE_32MB 0x5 +#define FARADAY_PCI_MEMSIZE_64MB 0x6 +#define FARADAY_PCI_MEMSIZE_128MB 0x7 +#define FARADAY_PCI_MEMSIZE_256MB 0x8 +#define FARADAY_PCI_MEMSIZE_512MB 0x9 +#define FARADAY_PCI_MEMSIZE_1GB 0xa +#define FARADAY_PCI_MEMSIZE_2GB 0xb +#define FARADAY_PCI_MEMSIZE_SHIFT 16 + +/* + * The DMA base is set to 0x0 for all memory segments, it reflects the + * fact that the memory of the host system starts at 0x0. + */ +#define FARADAY_PCI_DMA_MEM1_BASE 0x00000000 +#define FARADAY_PCI_DMA_MEM2_BASE 0x00000000 +#define FARADAY_PCI_DMA_MEM3_BASE 0x00000000 + +/* Defines for PCI configuration command register */ +#define PCI_CONF_ENABLE BIT(31) +#define PCI_CONF_WHERE(r) ((r) & 0xFC) +#define PCI_CONF_BUS(b) (((b) & 0xFF) << 16) +#define PCI_CONF_DEVICE(d) (((d) & 0x1F) << 11) +#define PCI_CONF_FUNCTION(f) (((f) & 0x07) << 8) + +/** + * struct faraday_pci_variant - encodes IP block differences + * @cascaded_irq: this host has cascaded IRQs from an interrupt controller + * embedded in the host bridge. + */ +struct faraday_pci_variant { + bool cascaded_irq; +}; + +struct faraday_pci { + struct device *dev; + void __iomem *base; + struct irq_domain *irqdomain; + struct pci_bus *bus; +}; + +static int faraday_res_to_memcfg(resource_size_t mem_base, + resource_size_t mem_size, u32 *val) +{ + u32 outval; + + switch (mem_size) { + case SZ_1M: + outval = FARADAY_PCI_MEMSIZE_1MB; + break; + case SZ_2M: + outval = FARADAY_PCI_MEMSIZE_2MB; + break; + case SZ_4M: + outval = FARADAY_PCI_MEMSIZE_4MB; + break; + case SZ_8M: + outval = FARADAY_PCI_MEMSIZE_8MB; + break; + case SZ_16M: + outval = FARADAY_PCI_MEMSIZE_16MB; + break; + case SZ_32M: + outval = FARADAY_PCI_MEMSIZE_32MB; + break; + case SZ_64M: + outval = FARADAY_PCI_MEMSIZE_64MB; + break; + case SZ_128M: + outval = FARADAY_PCI_MEMSIZE_128MB; + break; + case SZ_256M: + outval = FARADAY_PCI_MEMSIZE_256MB; + break; + case SZ_512M: + outval = FARADAY_PCI_MEMSIZE_512MB; + break; + case SZ_1G: + outval = FARADAY_PCI_MEMSIZE_1GB; + break; + case SZ_2G: + outval = FARADAY_PCI_MEMSIZE_2GB; + break; + default: + return -EINVAL; + } + outval <<= FARADAY_PCI_MEMSIZE_SHIFT; + + /* This is probably not good */ + if (mem_base & ~(FARADAY_PCI_MEMBASE_MASK)) + pr_warn("truncated PCI memory base\n"); + /* Translate to bridge side address space */ + outval |= (mem_base & FARADAY_PCI_MEMBASE_MASK); + pr_debug("Translated pci base @%pap, size %pap to config %08x\n", + &mem_base, &mem_size, outval); + + *val = outval; + return 0; +} + +static int faraday_pci_read_config(struct pci_bus *bus, unsigned int fn, + int config, int size, u32 *value) +{ + struct faraday_pci *p = bus->sysdata; + + writel(PCI_CONF_BUS(bus->number) | + PCI_CONF_DEVICE(PCI_SLOT(fn)) | + PCI_CONF_FUNCTION(PCI_FUNC(fn)) | + PCI_CONF_WHERE(config) | + PCI_CONF_ENABLE, + p->base + PCI_CONFIG); + + *value = readl(p->base + PCI_DATA); + + if (size == 1) + *value = (*value >> (8 * (config & 3))) & 0xFF; + else if (size == 2) + *value = (*value >> (8 * (config & 3))) & 0xFFFF; + + dev_dbg(&bus->dev, + "[read] slt: %.2d, fnc: %d, cnf: 0x%.2X, val (%d bytes): 0x%.8X\n", + PCI_SLOT(fn), PCI_FUNC(fn), config, size, *value); + + return PCIBIOS_SUCCESSFUL; +} + +static int faraday_pci_write_config(struct pci_bus *bus, unsigned int fn, + int config, int size, u32 value) +{ + struct faraday_pci *p = bus->sysdata; + int ret = PCIBIOS_SUCCESSFUL; + + dev_dbg(&bus->dev, + "[write] slt: %.2d, fnc: %d, cnf: 0x%.2X, val (%d bytes): 0x%.8X\n", + PCI_SLOT(fn), PCI_FUNC(fn), config, size, value); + + writel(PCI_CONF_BUS(bus->number) | + PCI_CONF_DEVICE(PCI_SLOT(fn)) | + PCI_CONF_FUNCTION(PCI_FUNC(fn)) | + PCI_CONF_WHERE(config) | + PCI_CONF_ENABLE, + p->base + PCI_CONFIG); + + switch (size) { + case 4: + writel(value, p->base + PCI_DATA); + break; + case 2: + writew(value, p->base + PCI_DATA + (config & 3)); + break; + case 1: + writeb(value, p->base + PCI_DATA + (config & 3)); + break; + default: + ret = PCIBIOS_BAD_REGISTER_NUMBER; + } + + return ret; +} + +static struct pci_ops faraday_pci_ops = { + .read = faraday_pci_read_config, + .write = faraday_pci_write_config, +}; + +static void faraday_pci_ack_irq(struct irq_data *d) +{ + struct faraday_pci *p = irq_data_get_irq_chip_data(d); + unsigned int reg; + + faraday_pci_read_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, ®); + reg &= ~(0xF << PCI_CTRL2_INTSTS_SHIFT); + reg |= BIT(irqd_to_hwirq(d) + PCI_CTRL2_INTSTS_SHIFT); + faraday_pci_write_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, reg); +} + +static void faraday_pci_mask_irq(struct irq_data *d) +{ + struct faraday_pci *p = irq_data_get_irq_chip_data(d); + unsigned int reg; + + faraday_pci_read_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, ®); + reg &= ~((0xF << PCI_CTRL2_INTSTS_SHIFT) + | BIT(irqd_to_hwirq(d) + PCI_CTRL2_INTMASK_SHIFT)); + faraday_pci_write_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, reg); +} + +static void faraday_pci_unmask_irq(struct irq_data *d) +{ + struct faraday_pci *p = irq_data_get_irq_chip_data(d); + unsigned int reg; + + faraday_pci_read_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, ®); + reg &= ~(0xF << PCI_CTRL2_INTSTS_SHIFT); + reg |= BIT(irqd_to_hwirq(d) + PCI_CTRL2_INTMASK_SHIFT); + faraday_pci_write_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, reg); +} + +static void faraday_pci_irq_handler(struct irq_desc *desc) +{ + struct faraday_pci *p = irq_desc_get_handler_data(desc); + struct irq_chip *irqchip = irq_desc_get_chip(desc); + unsigned int irq_stat, reg, i; + + faraday_pci_read_config(p->bus, 0, FARADAY_PCI_CTRL2, 4, ®); + irq_stat = reg >> PCI_CTRL2_INTSTS_SHIFT; + + chained_irq_enter(irqchip, desc); + + for (i = 0; i < 4; i++) { + if ((irq_stat & BIT(i)) == 0) + continue; + generic_handle_irq(irq_find_mapping(p->irqdomain, i)); + } + + chained_irq_exit(irqchip, desc); +} + +static struct irq_chip faraday_pci_irq_chip = { + .name = "PCI", + .irq_ack = faraday_pci_ack_irq, + .irq_mask = faraday_pci_mask_irq, + .irq_unmask = faraday_pci_unmask_irq, +}; + +static int faraday_pci_irq_map(struct irq_domain *domain, unsigned int irq, + irq_hw_number_t hwirq) +{ + irq_set_chip_and_handler(irq, &faraday_pci_irq_chip, handle_level_irq); + irq_set_chip_data(irq, domain->host_data); + + return 0; +} + +static const struct irq_domain_ops faraday_pci_irqdomain_ops = { + .map = faraday_pci_irq_map, +}; + +static int faraday_pci_setup_cascaded_irq(struct faraday_pci *p) +{ + struct device_node *intc = of_get_next_child(p->dev->of_node, NULL); + int irq; + int i; + + if (!intc) { + dev_err(p->dev, "missing child interrupt-controller node\n"); + return -EINVAL; + } + + /* All PCI IRQs cascade off this one */ + irq = of_irq_get(intc, 0); + if (!irq) { + dev_err(p->dev, "failed to get parent IRQ\n"); + return -EINVAL; + } + + p->irqdomain = irq_domain_add_linear(intc, 4, + &faraday_pci_irqdomain_ops, p); + if (!p->irqdomain) { + dev_err(p->dev, "failed to create Gemini PCI IRQ domain\n"); + return -EINVAL; + } + + irq_set_chained_handler_and_data(irq, faraday_pci_irq_handler, p); + + for (i = 0; i < 4; i++) + irq_create_mapping(p->irqdomain, i); + + return 0; +} + +static int pci_dma_range_parser_init(struct of_pci_range_parser *parser, + struct device_node *node) +{ + const int na = 3, ns = 2; + int rlen; + + parser->node = node; + parser->pna = of_n_addr_cells(node); + parser->np = parser->pna + na + ns; + + parser->range = of_get_property(node, "dma-ranges", &rlen); + if (!parser->range) + return -ENOENT; + parser->end = parser->range + rlen / sizeof(__be32); + + return 0; +} + +static int faraday_pci_parse_map_dma_ranges(struct faraday_pci *p, + struct device_node *np) +{ + struct of_pci_range range; + struct of_pci_range_parser parser; + struct device *dev = p->dev; + u32 confreg[3] = { + FARADAY_PCI_MEM1_BASE_SIZE, + FARADAY_PCI_MEM2_BASE_SIZE, + FARADAY_PCI_MEM3_BASE_SIZE, + }; + int i = 0; + u32 val; + + if (pci_dma_range_parser_init(&parser, np)) { + dev_err(dev, "missing dma-ranges property\n"); + return -EINVAL; + } + + /* + * Get the dma-ranges from the device tree + */ + for_each_of_pci_range(&parser, &range) { + u64 end = range.pci_addr + range.size - 1; + int ret; + + ret = faraday_res_to_memcfg(range.pci_addr, range.size, &val); + if (ret) { + dev_err(dev, + "DMA range %d: illegal MEM resource size\n", i); + return -EINVAL; + } + + dev_info(dev, "DMA MEM%d BASE: 0x%016llx -> 0x%016llx config %08x\n", + i + 1, range.pci_addr, end, val); + if (i <= 2) { + faraday_pci_write_config(p->bus, 0, confreg[i], + 4, val); + } else { + dev_err(dev, "ignore extraneous dma-range %d\n", i); + break; + } + + i++; + } + + return 0; +} + +static int faraday_pci_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + const struct faraday_pci_variant *variant = + of_device_get_match_data(dev); + struct resource *regs; + resource_size_t io_base; + struct resource_entry *win; + struct faraday_pci *p; + struct resource *mem; + struct resource *io; + struct pci_host_bridge *host; + int ret; + u32 val; + LIST_HEAD(res); + + host = pci_alloc_host_bridge(sizeof(*p)); + if (!host) + return -ENOMEM; + + host->dev.parent = dev; + host->ops = &faraday_pci_ops; + host->busnr = 0; + host->msi = NULL; + p = pci_host_bridge_priv(host); + host->sysdata = p; + p->dev = dev; + + regs = platform_get_resource(pdev, IORESOURCE_MEM, 0); + p->base = devm_ioremap_resource(dev, regs); + if (IS_ERR(p->base)) + return PTR_ERR(p->base); + + ret = of_pci_get_host_bridge_resources(dev->of_node, 0, 0xff, + &res, &io_base); + if (ret) + return ret; + + ret = devm_request_pci_bus_resources(dev, &res); + if (ret) + return ret; + + /* Get the I/O and memory ranges from DT */ + resource_list_for_each_entry(win, &res) { + switch (resource_type(win->res)) { + case IORESOURCE_IO: + io = win->res; + io->name = "Gemini PCI I/O"; + if (!faraday_res_to_memcfg(io->start - win->offset, + resource_size(io), &val)) { + /* setup I/O space size */ + writel(val, p->base + PCI_IOSIZE); + } else { + dev_err(dev, "illegal IO mem size\n"); + return -EINVAL; + } + ret = pci_remap_iospace(io, io_base); + if (ret) { + dev_warn(dev, "error %d: failed to map resource %pR\n", + ret, io); + continue; + } + break; + case IORESOURCE_MEM: + mem = win->res; + mem->name = "Gemini PCI MEM"; + break; + case IORESOURCE_BUS: + break; + default: + break; + } + } + + /* Setup hostbridge */ + val = readl(p->base + PCI_CTRL); + val |= PCI_COMMAND_IO; + val |= PCI_COMMAND_MEMORY; + val |= PCI_COMMAND_MASTER; + writel(val, p->base + PCI_CTRL); + + list_splice_init(&res, &host->windows); + ret = pci_register_host_bridge(host); + if (ret) { + dev_err(dev, "failed to register host: %d\n", ret); + return ret; + } + p->bus = host->bus; + + /* Mask and clear all interrupts */ + faraday_pci_write_config(p->bus, 0, FARADAY_PCI_CTRL2 + 2, 2, 0xF000); + if (variant->cascaded_irq) { + ret = faraday_pci_setup_cascaded_irq(p); + if (ret) { + dev_err(dev, "failed to setup cascaded IRQ\n"); + return ret; + } + } + + ret = faraday_pci_parse_map_dma_ranges(p, dev->of_node); + if (ret) + return ret; + + pci_scan_child_bus(p->bus); + pci_fixup_irqs(pci_common_swizzle, of_irq_parse_and_map_pci); + pci_bus_assign_resources(p->bus); + pci_bus_add_devices(p->bus); + pci_free_resource_list(&res); + + return 0; +} + +/* + * We encode bridge variants here, we have at least two so it doesn't + * hurt to have infrastructure to encompass future variants as well. + */ +const struct faraday_pci_variant faraday_regular = { + .cascaded_irq = true, +}; + +const struct faraday_pci_variant faraday_dual = { + .cascaded_irq = false, +}; + +static const struct of_device_id faraday_pci_of_match[] = { + { + .compatible = "faraday,ftpci100", + .data = &faraday_regular, + }, + { + .compatible = "faraday,ftpci100-dual", + .data = &faraday_dual, + }, + {}, +}; + +static struct platform_driver faraday_pci_driver = { + .driver = { + .name = "ftpci100", + .of_match_table = of_match_ptr(faraday_pci_of_match), + }, + .probe = faraday_pci_probe, +}; +builtin_platform_driver(faraday_pci_driver); -- cgit v1.2.3-59-g8ed1b From b8e82c1bdd2871dc9c704502acc2ca634b51eeda Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 17 Jan 2017 21:40:52 +0000 Subject: PCI: mvebu: Avoid changing the SCC bit in the Link Status register It seems on later Armada 38x, the slot clock configuration bit is not read-only, but can be written. This means that our RW1C protection ends up clearing this bit when the link control register is written. Adjust the mask so that we only avoid writing '1' bits to the RW1C bits of this register (bits 15 and 14 of the link status) rather than masking out all the status register bits. Signed-off-by: Russell King Signed-off-by: Bjorn Helgaas --- drivers/pci/host/pci-mvebu.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/pci/host/pci-mvebu.c b/drivers/pci/host/pci-mvebu.c index cd7d51988738..266447ee8867 100644 --- a/drivers/pci/host/pci-mvebu.c +++ b/drivers/pci/host/pci-mvebu.c @@ -752,10 +752,11 @@ static int mvebu_sw_pci_bridge_write(struct mvebu_pcie_port *port, * If the mask is 0xffff0000, then we only want to write * the link control register, rather than clearing the * RW1C bits in the link status register. Mask out the - * status register bits. + * RW1C status register bits. */ if (mask == 0xffff0000) - value &= 0xffff; + value &= ~((PCI_EXP_LNKSTA_LABS | + PCI_EXP_LNKSTA_LBMS) << 16); mvebu_writel(port, value, PCIE_CAP_PCIEXP + PCI_EXP_LNKCTL); break; -- cgit v1.2.3-59-g8ed1b From d3881e5015421a578bc328136471fcf1d02ac389 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Tue, 7 Feb 2017 14:32:33 -0500 Subject: PCI: Export PCI device config accessors Replace the inline PCI device config read and write accessors with exported functions. This is preparing for these functions to make use of private data. Tested-by: Krishna Dhulipala Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig Reviewed-by: Wei Zhang --- drivers/pci/access.c | 38 ++++++++++++++++++++++++++++++++++++++ include/linux/pci.h | 32 ++++++-------------------------- 2 files changed, 44 insertions(+), 26 deletions(-) diff --git a/drivers/pci/access.c b/drivers/pci/access.c index 8b7382705bf2..c70e3113df7a 100644 --- a/drivers/pci/access.c +++ b/drivers/pci/access.c @@ -890,3 +890,41 @@ int pcie_capability_clear_and_set_dword(struct pci_dev *dev, int pos, return ret; } EXPORT_SYMBOL(pcie_capability_clear_and_set_dword); + +int pci_read_config_byte(const struct pci_dev *dev, int where, u8 *val) +{ + return pci_bus_read_config_byte(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_read_config_byte); + +int pci_read_config_word(const struct pci_dev *dev, int where, u16 *val) +{ + return pci_bus_read_config_word(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_read_config_word); + +int pci_read_config_dword(const struct pci_dev *dev, int where, + u32 *val) +{ + return pci_bus_read_config_dword(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_read_config_dword); + +int pci_write_config_byte(const struct pci_dev *dev, int where, u8 val) +{ + return pci_bus_write_config_byte(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_write_config_byte); + +int pci_write_config_word(const struct pci_dev *dev, int where, u16 val) +{ + return pci_bus_write_config_word(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_write_config_word); + +int pci_write_config_dword(const struct pci_dev *dev, int where, + u32 val) +{ + return pci_bus_write_config_dword(dev->bus, dev->devfn, where, val); +} +EXPORT_SYMBOL(pci_write_config_dword); diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..d705f3088ff9 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -940,32 +940,12 @@ int pci_generic_config_write32(struct pci_bus *bus, unsigned int devfn, struct pci_ops *pci_bus_set_ops(struct pci_bus *bus, struct pci_ops *ops); -static inline int pci_read_config_byte(const struct pci_dev *dev, int where, u8 *val) -{ - return pci_bus_read_config_byte(dev->bus, dev->devfn, where, val); -} -static inline int pci_read_config_word(const struct pci_dev *dev, int where, u16 *val) -{ - return pci_bus_read_config_word(dev->bus, dev->devfn, where, val); -} -static inline int pci_read_config_dword(const struct pci_dev *dev, int where, - u32 *val) -{ - return pci_bus_read_config_dword(dev->bus, dev->devfn, where, val); -} -static inline int pci_write_config_byte(const struct pci_dev *dev, int where, u8 val) -{ - return pci_bus_write_config_byte(dev->bus, dev->devfn, where, val); -} -static inline int pci_write_config_word(const struct pci_dev *dev, int where, u16 val) -{ - return pci_bus_write_config_word(dev->bus, dev->devfn, where, val); -} -static inline int pci_write_config_dword(const struct pci_dev *dev, int where, - u32 val) -{ - return pci_bus_write_config_dword(dev->bus, dev->devfn, where, val); -} +int pci_read_config_byte(const struct pci_dev *dev, int where, u8 *val); +int pci_read_config_word(const struct pci_dev *dev, int where, u16 *val); +int pci_read_config_dword(const struct pci_dev *dev, int where, u32 *val); +int pci_write_config_byte(const struct pci_dev *dev, int where, u8 val); +int pci_write_config_word(const struct pci_dev *dev, int where, u16 val); +int pci_write_config_dword(const struct pci_dev *dev, int where, u32 val); int pcie_capability_read_word(struct pci_dev *dev, int pos, u16 *val); int pcie_capability_read_dword(struct pci_dev *dev, int pos, u32 *val); -- cgit v1.2.3-59-g8ed1b From 89ee9f7680031d7df91a1a27abac69e034c2e892 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Wed, 29 Mar 2017 22:48:59 -0500 Subject: PCI: Add device disconnected state Add a new state to pci_dev to be set when it is unexpectedly disconnected. The PCI driver tear down functions can observe this new device state so they may skip operations that will fail. The pciehp and pcie-dpc drivers are aware when the link is down, so these set the flag when their handlers detect the device is disconnected. Tested-by: Krishna Dhulipala Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig Reviewed-by: Wei Zhang --- drivers/pci/hotplug/pciehp_pci.c | 6 ++++++ drivers/pci/pci.h | 14 ++++++++++++++ drivers/pci/pcie/pcie-dpc.c | 5 +++++ include/linux/pci.h | 2 ++ 4 files changed, 27 insertions(+) diff --git a/drivers/pci/hotplug/pciehp_pci.c b/drivers/pci/hotplug/pciehp_pci.c index 9e69403be632..19f30a9f461d 100644 --- a/drivers/pci/hotplug/pciehp_pci.c +++ b/drivers/pci/hotplug/pciehp_pci.c @@ -109,6 +109,12 @@ int pciehp_unconfigure_device(struct slot *p_slot) break; } } + if (!presence) { + pci_dev_set_disconnected(dev, NULL); + if (pci_has_subordinate(dev)) + pci_walk_bus(dev->subordinate, + pci_dev_set_disconnected, NULL); + } pci_stop_and_remove_bus_device(dev); /* * Ensure that no new Requests will be generated from diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 8dd38e69d6f2..245719c3e409 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -274,6 +274,20 @@ struct pci_sriov { resource_size_t barsz[PCI_SRIOV_NUM_BARS]; /* VF BAR size */ }; +/* pci_dev priv_flags */ +#define PCI_DEV_DISCONNECTED 0 + +static inline int pci_dev_set_disconnected(struct pci_dev *dev, void *unused) +{ + set_bit(PCI_DEV_DISCONNECTED, &dev->priv_flags); + return 0; +} + +static inline bool pci_dev_is_disconnected(const struct pci_dev *dev) +{ + return test_bit(PCI_DEV_DISCONNECTED, &dev->priv_flags); +} + #ifdef CONFIG_PCI_ATS void pci_restore_ats_state(struct pci_dev *dev); #else diff --git a/drivers/pci/pcie/pcie-dpc.c b/drivers/pci/pcie/pcie-dpc.c index d4d70ef4a2d7..77d2ca99d2ec 100644 --- a/drivers/pci/pcie/pcie-dpc.c +++ b/drivers/pci/pcie/pcie-dpc.c @@ -14,6 +14,7 @@ #include #include #include +#include "../pci.h" struct dpc_dev { struct pcie_device *dev; @@ -66,6 +67,10 @@ static void interrupt_event_handler(struct work_struct *work) list_for_each_entry_safe_reverse(dev, temp, &parent->devices, bus_list) { pci_dev_get(dev); + pci_dev_set_disconnected(dev, NULL); + if (pci_has_subordinate(dev)) + pci_walk_bus(dev->subordinate, + pci_dev_set_disconnected, NULL); pci_stop_and_remove_bus_device(dev); pci_dev_put(dev); } diff --git a/include/linux/pci.h b/include/linux/pci.h index d705f3088ff9..2887933329a2 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -396,6 +396,8 @@ struct pci_dev { phys_addr_t rom; /* Physical address of ROM if it's not from the BAR */ size_t romlen; /* Length of ROM if it's not from the BAR */ char *driver_override; /* Driver name to force a match */ + + unsigned long priv_flags; /* Private flags for the pci driver */ }; static inline struct pci_dev *pci_physfn(struct pci_dev *dev) -- cgit v1.2.3-59-g8ed1b From 4b103883473964bbb7b0b6934999b283ee234a8a Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Wed, 29 Mar 2017 22:49:06 -0500 Subject: PCI: Don't attempt config access to disconnected devices If we've detected the PCI device is disconnected, there is no need to attempt to access its config space since we know the operation will fail. Make all the config reads and writes return -ENODEV error immediately when in such a state. If a caller requests a config read to a disconnected device, return a data value of all 1's. This is the same as what hardware is expected to return when accessing a removed device, but software can do this faster without relying on hardware. Tested-by: Krishna Dhulipala Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig Reviewed-by: Wei Zhang --- drivers/pci/access.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/drivers/pci/access.c b/drivers/pci/access.c index c70e3113df7a..1c8051003e67 100644 --- a/drivers/pci/access.c +++ b/drivers/pci/access.c @@ -893,12 +893,20 @@ EXPORT_SYMBOL(pcie_capability_clear_and_set_dword); int pci_read_config_byte(const struct pci_dev *dev, int where, u8 *val) { + if (pci_dev_is_disconnected(dev)) { + *val = ~0; + return -ENODEV; + } return pci_bus_read_config_byte(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_read_config_byte); int pci_read_config_word(const struct pci_dev *dev, int where, u16 *val) { + if (pci_dev_is_disconnected(dev)) { + *val = ~0; + return -ENODEV; + } return pci_bus_read_config_word(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_read_config_word); @@ -906,18 +914,26 @@ EXPORT_SYMBOL(pci_read_config_word); int pci_read_config_dword(const struct pci_dev *dev, int where, u32 *val) { + if (pci_dev_is_disconnected(dev)) { + *val = ~0; + return -ENODEV; + } return pci_bus_read_config_dword(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_read_config_dword); int pci_write_config_byte(const struct pci_dev *dev, int where, u8 val) { + if (pci_dev_is_disconnected(dev)) + return -ENODEV; return pci_bus_write_config_byte(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_write_config_byte); int pci_write_config_word(const struct pci_dev *dev, int where, u16 val) { + if (pci_dev_is_disconnected(dev)) + return -ENODEV; return pci_bus_write_config_word(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_write_config_word); @@ -925,6 +941,8 @@ EXPORT_SYMBOL(pci_write_config_word); int pci_write_config_dword(const struct pci_dev *dev, int where, u32 val) { + if (pci_dev_is_disconnected(dev)) + return -ENODEV; return pci_bus_write_config_dword(dev->bus, dev->devfn, where, val); } EXPORT_SYMBOL(pci_write_config_dword); -- cgit v1.2.3-59-g8ed1b From 0170591bb06788034ae72bfa56e42f283ab3de5a Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Wed, 29 Mar 2017 22:49:11 -0500 Subject: PCI/MSI: Skip disabling disconnected devices Check the device connected state prior to executing device shutdown operations or writing MSI messages so that tear down on disconnected devices completes quicker. Tested-by: Krishna Dhulipala Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig Reviewed-by: Wei Zhang --- drivers/pci/msi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c index d571bc330686..33e766a9dd91 100644 --- a/drivers/pci/msi.c +++ b/drivers/pci/msi.c @@ -298,7 +298,7 @@ void __pci_write_msi_msg(struct msi_desc *entry, struct msi_msg *msg) { struct pci_dev *dev = msi_desc_to_pci_dev(entry); - if (dev->current_state != PCI_D0) { + if (dev->current_state != PCI_D0 || pci_dev_is_disconnected(dev)) { /* Don't touch the hardware now */ } else if (entry->msi_attrib.is_msix) { void __iomem *base = pci_msix_desc_addr(entry); @@ -1001,6 +1001,11 @@ void pci_msix_shutdown(struct pci_dev *dev) if (!pci_msi_enable || !dev || !dev->msix_enabled) return; + if (pci_dev_is_disconnected(dev)) { + dev->msix_enabled = 0; + return; + } + /* Return the device with MSI-X masked as initial states */ for_each_pci_msi_entry(entry, dev) { /* Keep cached states to be restored */ -- cgit v1.2.3-59-g8ed1b From fe2bd75b2290bf6acd4b91312697e445668dd773 Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Wed, 29 Mar 2017 22:49:17 -0500 Subject: PCI: Short-circuit pci_device_is_present() for disconnected devices If the PCI device is disconnected, return false immediately from pci_device_is_present(). pci_device_is_present() uses the bus accessors, so the early return in the device accessors doesn't help here. Tested-by: Krishna Dhulipala Signed-off-by: Keith Busch Signed-off-by: Bjorn Helgaas Reviewed-by: Christoph Hellwig Reviewed-by: Wei Zhang --- drivers/pci/pci.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..8ab0a0d3cddb 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -4932,6 +4932,8 @@ bool pci_device_is_present(struct pci_dev *pdev) { u32 v; + if (pci_dev_is_disconnected(pdev)) + return false; return pci_bus_read_dev_vendor_id(pdev->bus, pdev->devfn, &v, 0); } EXPORT_SYMBOL_GPL(pci_device_is_present); -- cgit v1.2.3-59-g8ed1b From b77d537d00d08fcf0bf641cd3491dd7df0ad1475 Mon Sep 17 00:00:00 2001 From: Manish Jaggi Date: Thu, 30 Mar 2017 18:47:14 -0500 Subject: PCI: Apply Cavium ACS quirk only to CN81xx/CN83xx/CN88xx devices Only apply the Cavium ACS quirk to devices with ID in the range 0xa000-0xa0ff. These are the on-chip PCI devices for CN81xx/CN83xx/CN88xx. Fixes: b404bcfbf035 ("PCI: Add ACS quirk for all Cavium devices") Reported-by: Alex Williamson Signed-off-by: Manish Jaggi Acked-by: David Daney Acked-by: Alex Williamson --- drivers/pci/quirks.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f754453fe754..178242d0d9df 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -4094,6 +4094,9 @@ static int pci_quirk_cavium_acs(struct pci_dev *dev, u16 acs_flags) acs_flags &= ~(PCI_ACS_SV | PCI_ACS_TB | PCI_ACS_RR | PCI_ACS_CR | PCI_ACS_UF | PCI_ACS_DT); + if (!((dev->device >= 0xa000) && (dev->device <= 0xa0ff))) + return -ENOTTY; + return acs_flags ? 0 : 1; } -- cgit v1.2.3-59-g8ed1b From f65fd1aa4f9881d5540192d11f7b8ed2fec936db Mon Sep 17 00:00:00 2001 From: Sasha Neftin Date: Mon, 3 Apr 2017 16:02:50 -0500 Subject: PCI: Avoid FLR for Intel 82579 NICs Per Intel Specification Update 335553-002 (see link below), some 82579 network adapters advertise a Function Level Reset (FLR) capability, but they can hang when an FLR is triggered. To reproduce the problem, attach the device to a VM, then detach and try to attach again. Add a quirk to prevent the use of FLR on these devices. [bhelgaas: changelog, comments] Link: http://www.intel.com/content/dam/www/public/us/en/documents/specification-updates/82579lm-82579v-gigabit-network-connection-spec-update.pdf Signed-off-by: Sasha Neftin Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 6 ++++++ drivers/pci/quirks.c | 8 ++++++++ include/linux/pci.h | 2 ++ 3 files changed, 16 insertions(+) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..bef14777bb30 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -3781,6 +3781,9 @@ static int pcie_flr(struct pci_dev *dev, int probe) if (!(cap & PCI_EXP_DEVCAP_FLR)) return -ENOTTY; + if (dev->dev_flags & PCI_DEV_FLAGS_NO_FLR_RESET) + return -ENOTTY; + if (probe) return 0; @@ -3801,6 +3804,9 @@ static int pci_af_flr(struct pci_dev *dev, int probe) if (!pos) return -ENOTTY; + if (dev->dev_flags & PCI_DEV_FLAGS_NO_FLR_RESET) + return -ENOTTY; + pci_read_config_byte(dev, pos + PCI_AF_CAP, &cap); if (!(cap & PCI_AF_CAP_TP) || !(cap & PCI_AF_CAP_FLR)) return -ENOTTY; diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f754453fe754..823271b13d12 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -4633,3 +4633,11 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x2030, quirk_no_aersid); 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); + +/* FLR may cause some 82579 devices to hang. */ +static void quirk_intel_no_flr(struct pci_dev *dev) +{ + dev->dev_flags |= PCI_DEV_FLAGS_NO_FLR_RESET; +} +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x1502, quirk_intel_no_flr); +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, 0x1503, quirk_intel_no_flr); diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..22cad2c66d59 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -178,6 +178,8 @@ enum pci_dev_flags { PCI_DEV_FLAGS_NO_PM_RESET = (__force pci_dev_flags_t) (1 << 7), /* Get VPD from function 0 VPD */ PCI_DEV_FLAGS_VPD_REF_F0 = (__force pci_dev_flags_t) (1 << 8), + /* Do not use FLR even if device advertises PCI_AF_CAP */ + PCI_DEV_FLAGS_NO_FLR_RESET = (__force pci_dev_flags_t) (1 << 10), }; enum pci_irq_reroute_variant { -- cgit v1.2.3-59-g8ed1b From 55021718b9e9d21efd47117dd0e6c1cfebd654f8 Mon Sep 17 00:00:00 2001 From: Shawn Lin Date: Mon, 20 Mar 2017 17:39:40 +0800 Subject: PCI: rockchip: Advertise 128-byte Read Completion Boundary support Rockchip Root Ports support either 64 or 128 byte Read Completion Boundary (RCB). Set the RCB bit in the Link Control register to indicate this. A 128 byte RCB significantly improves performance of NVMe with libaio. [bhelgaas: changelog] Signed-off-by: Shawn Lin Signed-off-by: Bjorn Helgaas Cc: Brian Norris Cc: Jeffy Chen --- drivers/pci/host/pcie-rockchip.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index b4dfe3bd15c3..a7467212ea18 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -601,6 +601,11 @@ static int rockchip_pcie_init_port(struct rockchip_pcie *rockchip) status |= PCI_EXP_LNKCTL_CCC; rockchip_pcie_write(rockchip, status, PCIE_RC_CONFIG_LCS); + /* Set RC's RCB to 128 */ + status = rockchip_pcie_read(rockchip, PCIE_RC_CONFIG_LCS); + status |= PCI_EXP_LNKCTL_RCB; + rockchip_pcie_write(rockchip, status, PCIE_RC_CONFIG_LCS); + /* Enable Gen1 training */ rockchip_pcie_write(rockchip, PCIE_CLIENT_LINK_TRAIN_ENABLE, PCIE_CLIENT_CONFIG); -- cgit v1.2.3-59-g8ed1b From dbe4a09e8bbcf88809a8394d6a359d8cebd22a86 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Thu, 16 Mar 2017 14:34:59 -0500 Subject: PCI: dwc: Unindent dw_handle_msi_irq() loop Use "continue" to skip rest of the loop when possible to save an indent level. No functional change intended. Suggested-by: walter harms Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pcie-designware-host.c | 22 +++++++++++----------- 1 file changed, 11 insertions(+), 11 deletions(-) diff --git a/drivers/pci/dwc/pcie-designware-host.c b/drivers/pci/dwc/pcie-designware-host.c index 5ba334938b52..6cdd41f06dea 100644 --- a/drivers/pci/dwc/pcie-designware-host.c +++ b/drivers/pci/dwc/pcie-designware-host.c @@ -63,17 +63,17 @@ irqreturn_t dw_handle_msi_irq(struct pcie_port *pp) for (i = 0; i < MAX_MSI_CTRLS; i++) { dw_pcie_rd_own_conf(pp, PCIE_MSI_INTR0_STATUS + i * 12, 4, (u32 *)&val); - if (val) { - ret = IRQ_HANDLED; - pos = 0; - while ((pos = find_next_bit(&val, 32, pos)) != 32) { - irq = irq_find_mapping(pp->irq_domain, - i * 32 + pos); - dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_STATUS + - i * 12, 4, 1 << pos); - generic_handle_irq(irq); - pos++; - } + if (!val) + continue; + + ret = IRQ_HANDLED; + pos = 0; + while ((pos = find_next_bit(&val, 32, pos)) != 32) { + irq = irq_find_mapping(pp->irq_domain, i * 32 + pos); + dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_STATUS + i * 12, + 4, 1 << pos); + generic_handle_irq(irq); + pos++; } } -- cgit v1.2.3-59-g8ed1b From 1b497e6493c49bbb55c89f53562f7f853495e90d Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Thu, 16 Mar 2017 14:34:51 -0500 Subject: PCI: dwc: Fix uninitialized variable in dw_handle_msi_irq() The bug is that "val" is unsigned long but we only initialize 32 bits of it. Then we test "if (val)" and that might be true not because we set the bits but because some were never initialized. Fixes: f342d940ee0e ("PCI: exynos: Add support for MSI") Signed-off-by: Dan Carpenter Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pcie-designware-host.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/pci/dwc/pcie-designware-host.c b/drivers/pci/dwc/pcie-designware-host.c index 6cdd41f06dea..aece2962defc 100644 --- a/drivers/pci/dwc/pcie-designware-host.c +++ b/drivers/pci/dwc/pcie-designware-host.c @@ -56,19 +56,20 @@ static struct irq_chip dw_msi_irq_chip = { /* MSI int handler */ irqreturn_t dw_handle_msi_irq(struct pcie_port *pp) { - unsigned long val; + u32 val; int i, pos, irq; irqreturn_t ret = IRQ_NONE; for (i = 0; i < MAX_MSI_CTRLS; i++) { dw_pcie_rd_own_conf(pp, PCIE_MSI_INTR0_STATUS + i * 12, 4, - (u32 *)&val); + &val); if (!val) continue; ret = IRQ_HANDLED; pos = 0; - while ((pos = find_next_bit(&val, 32, pos)) != 32) { + while ((pos = find_next_bit((unsigned long *) &val, 32, + pos)) != 32) { irq = irq_find_mapping(pp->irq_domain, i * 32 + pos); dw_pcie_wr_own_conf(pp, PCIE_MSI_INTR0_STATUS + i * 12, 4, 1 << pos); -- cgit v1.2.3-59-g8ed1b From a660083eb06c5bb0ad049377dbd2522e4b1551d6 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:22 +0530 Subject: PCI: dwc: designware: Add new *ops* for CPU addr fixup Some platforms (like dra7xx) require only the least 28 bits of the corresponding 32 bit CPU address to be programmed in the address translation unit. This modified address is stored in io_base/mem_base/ cfg0_base/cfg1_base in dra7xx_pcie_host_init(). While this is okay for host mode where the address range is fixed, device mode requires different addresses to be programmed based on the host buffer address. Add a new ops to get the least 28 bits of the corresponding 32 bit CPU address and invoke it before programming the address translation unit. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Joao Pinto --- drivers/pci/dwc/pcie-designware.c | 3 +++ drivers/pci/dwc/pcie-designware.h | 1 + 2 files changed, 4 insertions(+) diff --git a/drivers/pci/dwc/pcie-designware.c b/drivers/pci/dwc/pcie-designware.c index 7e1fb7d6643c..3eaf3ccdc118 100644 --- a/drivers/pci/dwc/pcie-designware.c +++ b/drivers/pci/dwc/pcie-designware.c @@ -97,6 +97,9 @@ void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, int type, { u32 retries, val; + if (pci->ops->cpu_addr_fixup) + cpu_addr = pci->ops->cpu_addr_fixup(cpu_addr); + if (pci->iatu_unroll_enabled) { dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_LOWER_BASE, lower_32_bits(cpu_addr)); diff --git a/drivers/pci/dwc/pcie-designware.h b/drivers/pci/dwc/pcie-designware.h index cd3b8713fe50..8f3dcb2b099b 100644 --- a/drivers/pci/dwc/pcie-designware.h +++ b/drivers/pci/dwc/pcie-designware.h @@ -143,6 +143,7 @@ struct pcie_port { }; struct dw_pcie_ops { + u64 (*cpu_addr_fixup)(u64 cpu_addr); u32 (*readl_dbi)(struct dw_pcie *pcie, u32 reg); void (*writel_dbi)(struct dw_pcie *pcie, u32 reg, u32 val); int (*link_up)(struct dw_pcie *pcie); -- cgit v1.2.3-59-g8ed1b From 2ed6cc71e6f726c131fb8675eb64cc0519c26ae8 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:23 +0530 Subject: PCI: dwc: dra7xx: Populate cpu_addr_fixup ops Populate cpu_addr_fixup ops to extract the least 28 bits of the corresponding CPU address. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Joao Pinto --- drivers/pci/dwc/pci-dra7xx.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/pci/dwc/pci-dra7xx.c b/drivers/pci/dwc/pci-dra7xx.c index 0984baff07e3..07c45ec07662 100644 --- a/drivers/pci/dwc/pci-dra7xx.c +++ b/drivers/pci/dwc/pci-dra7xx.c @@ -88,6 +88,11 @@ static inline void dra7xx_pcie_writel(struct dra7xx_pcie *pcie, u32 offset, writel(value, pcie->base + offset); } +static u64 dra7xx_pcie_cpu_addr_fixup(u64 pci_addr) +{ + return pci_addr & DRA7XX_CPU_TO_BUS_ADDR; +} + static int dra7xx_pcie_link_up(struct dw_pcie *pci) { struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); @@ -152,11 +157,6 @@ static void dra7xx_pcie_host_init(struct pcie_port *pp) struct dw_pcie *pci = to_dw_pcie_from_pp(pp); struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); - pp->io_base &= DRA7XX_CPU_TO_BUS_ADDR; - pp->mem_base &= DRA7XX_CPU_TO_BUS_ADDR; - pp->cfg0_base &= DRA7XX_CPU_TO_BUS_ADDR; - pp->cfg1_base &= DRA7XX_CPU_TO_BUS_ADDR; - dw_pcie_setup_rc(pp); dra7xx_pcie_establish_link(dra7xx); @@ -329,6 +329,7 @@ static int __init dra7xx_add_pcie_port(struct dra7xx_pcie *dra7xx, } static const struct dw_pcie_ops dw_pcie_ops = { + .cpu_addr_fixup = dra7xx_pcie_cpu_addr_fixup, .link_up = dra7xx_pcie_link_up, }; -- cgit v1.2.3-59-g8ed1b From 62c5549fc4eab7ce7dc328b5c325e7a98b5b16c0 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:24 +0530 Subject: PCI: dwc: artpec6: Populate cpu_addr_fixup ops Populate cpu_addr_fixup ops to extract the least 28 bits of the corresponding CPU address. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Niklas Cassel Acked-by: Joao Pinto --- drivers/pci/dwc/pcie-artpec6.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/pci/dwc/pcie-artpec6.c b/drivers/pci/dwc/pcie-artpec6.c index 6d23683c0892..5b3b3afc0edb 100644 --- a/drivers/pci/dwc/pcie-artpec6.c +++ b/drivers/pci/dwc/pcie-artpec6.c @@ -78,6 +78,11 @@ static void artpec6_pcie_writel(struct artpec6_pcie *artpec6_pcie, u32 offset, u regmap_write(artpec6_pcie->regmap, offset, val); } +static u64 artpec6_pcie_cpu_addr_fixup(u64 pci_addr) +{ + return pci_addr & ARTPEC6_CPU_TO_BUS_ADDR; +} + static int artpec6_pcie_establish_link(struct artpec6_pcie *artpec6_pcie) { struct dw_pcie *pci = artpec6_pcie->pci; @@ -142,11 +147,6 @@ static int artpec6_pcie_establish_link(struct artpec6_pcie *artpec6_pcie) */ dw_pcie_writel_dbi(pci, MISC_CONTROL_1_OFF, DBI_RO_WR_EN); - pp->io_base &= ARTPEC6_CPU_TO_BUS_ADDR; - pp->mem_base &= ARTPEC6_CPU_TO_BUS_ADDR; - pp->cfg0_base &= ARTPEC6_CPU_TO_BUS_ADDR; - pp->cfg1_base &= ARTPEC6_CPU_TO_BUS_ADDR; - /* setup root complex */ dw_pcie_setup_rc(pp); @@ -235,6 +235,7 @@ static int artpec6_add_pcie_port(struct artpec6_pcie *artpec6_pcie, } static const struct dw_pcie_ops dw_pcie_ops = { + .cpu_addr_fixup = artpec6_pcie_cpu_addr_fixup, }; static int artpec6_pcie_probe(struct platform_device *pdev) -- cgit v1.2.3-59-g8ed1b From b50b2db266d8a8c303e8d88590c6416dfe576c6c Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:25 +0530 Subject: PCI: dwc: all: Modify dbi accessors to take dbi_base as argument dwc has 2 dbi address space labeled dbics and dbics2. The existing helper to access dbi address space can access only dbics. However dbics2 has to be accessed for programming the BAR registers in the case of EP mode. This is in preparation for adding EP mode support to dwc driver. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Niklas Cassel Cc: Jingoo Han Cc: Joao Pinto --- drivers/pci/dwc/pci-exynos.c | 10 ++++++---- drivers/pci/dwc/pcie-designware.c | 13 +++++++------ drivers/pci/dwc/pcie-designware.h | 20 ++++++++++++++++---- 3 files changed, 29 insertions(+), 14 deletions(-) diff --git a/drivers/pci/dwc/pci-exynos.c b/drivers/pci/dwc/pci-exynos.c index 44f774c12fb2..c2dafad48d50 100644 --- a/drivers/pci/dwc/pci-exynos.c +++ b/drivers/pci/dwc/pci-exynos.c @@ -521,23 +521,25 @@ static void exynos_pcie_enable_interrupts(struct exynos_pcie *ep) exynos_pcie_msi_init(ep); } -static u32 exynos_pcie_readl_dbi(struct dw_pcie *pci, u32 reg) +static u32 exynos_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, + u32 reg) { struct exynos_pcie *ep = to_exynos_pcie(pci); u32 val; exynos_pcie_sideband_dbi_r_mode(ep, true); - val = readl(pci->dbi_base + reg); + val = readl(base + reg); exynos_pcie_sideband_dbi_r_mode(ep, false); return val; } -static void exynos_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val) +static void exynos_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, + u32 reg, u32 val) { struct exynos_pcie *ep = to_exynos_pcie(pci); exynos_pcie_sideband_dbi_w_mode(ep, true); - writel(val, pci->dbi_base + reg); + writel(val, base + reg); exynos_pcie_sideband_dbi_w_mode(ep, false); } diff --git a/drivers/pci/dwc/pcie-designware.c b/drivers/pci/dwc/pcie-designware.c index 3eaf3ccdc118..ea403e2240cf 100644 --- a/drivers/pci/dwc/pcie-designware.c +++ b/drivers/pci/dwc/pcie-designware.c @@ -61,20 +61,21 @@ int dw_pcie_write(void __iomem *addr, int size, u32 val) return PCIBIOS_SUCCESSFUL; } -u32 dw_pcie_readl_dbi(struct dw_pcie *pci, u32 reg) +u32 __dw_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg) { if (pci->ops->readl_dbi) - return pci->ops->readl_dbi(pci, reg); + return pci->ops->readl_dbi(pci, base, reg); - return readl(pci->dbi_base + reg); + return readl(base + reg); } -void dw_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val) +void __dw_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + u32 val) { if (pci->ops->writel_dbi) - pci->ops->writel_dbi(pci, reg, val); + pci->ops->writel_dbi(pci, base, reg, val); else - writel(val, pci->dbi_base + reg); + writel(val, base + reg); } static u32 dw_pcie_readl_unroll(struct dw_pcie *pci, u32 index, u32 reg) diff --git a/drivers/pci/dwc/pcie-designware.h b/drivers/pci/dwc/pcie-designware.h index 8f3dcb2b099b..09b334a12892 100644 --- a/drivers/pci/dwc/pcie-designware.h +++ b/drivers/pci/dwc/pcie-designware.h @@ -144,8 +144,9 @@ struct pcie_port { struct dw_pcie_ops { u64 (*cpu_addr_fixup)(u64 cpu_addr); - u32 (*readl_dbi)(struct dw_pcie *pcie, u32 reg); - void (*writel_dbi)(struct dw_pcie *pcie, u32 reg, u32 val); + u32 (*readl_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg); + void (*writel_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, + u32 val); int (*link_up)(struct dw_pcie *pcie); }; @@ -163,8 +164,9 @@ struct dw_pcie { int dw_pcie_read(void __iomem *addr, int size, u32 *val); int dw_pcie_write(void __iomem *addr, int size, u32 val); -u32 dw_pcie_readl_dbi(struct dw_pcie *pci, u32 reg); -void dw_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val); +u32 __dw_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg); +void __dw_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + u32 val); int dw_pcie_link_up(struct dw_pcie *pci); int dw_pcie_wait_for_link(struct dw_pcie *pci); void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, @@ -172,6 +174,16 @@ void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, u32 size); void dw_pcie_setup(struct dw_pcie *pci); +static inline void dw_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val) +{ + __dw_pcie_writel_dbi(pci, pci->dbi_base, reg, val); +} + +static inline u32 dw_pcie_readl_dbi(struct dw_pcie *pci, u32 reg) +{ + return __dw_pcie_readl_dbi(pci, pci->dbi_base, reg); +} + #ifdef CONFIG_PCIE_DW_HOST irqreturn_t dw_handle_msi_irq(struct pcie_port *pp); void dw_pcie_msi_init(struct pcie_port *pp); -- cgit v1.2.3-59-g8ed1b From a509d7d9af5ebf86ffbefa98e49761d813fb1d40 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:26 +0530 Subject: PCI: dwc: all: Modify dbi accessors to access data of 4/2/1 bytes Previously dbi accessors can be used to access data of size 4 bytes. But there might be situations (like accessing MSI_MESSAGE_CONTROL in order to set/get the number of required MSI interrupts in EP mode) where dbi accessors must be used to access data of size 2. This is in preparation for adding endpoint mode support to designware driver. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Niklas Cassel Cc: Jingoo Han Cc: Joao Pinto --- drivers/pci/dwc/pci-exynos.c | 16 ++++++++-------- drivers/pci/dwc/pcie-designware.c | 34 ++++++++++++++++++++++++---------- drivers/pci/dwc/pcie-designware.h | 18 ++++++++++-------- 3 files changed, 42 insertions(+), 26 deletions(-) diff --git a/drivers/pci/dwc/pci-exynos.c b/drivers/pci/dwc/pci-exynos.c index c2dafad48d50..546082ad5a3f 100644 --- a/drivers/pci/dwc/pci-exynos.c +++ b/drivers/pci/dwc/pci-exynos.c @@ -521,25 +521,25 @@ static void exynos_pcie_enable_interrupts(struct exynos_pcie *ep) exynos_pcie_msi_init(ep); } -static u32 exynos_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, - u32 reg) +static u32 exynos_pcie_read_dbi(struct dw_pcie *pci, void __iomem *base, + u32 reg, size_t size) { struct exynos_pcie *ep = to_exynos_pcie(pci); u32 val; exynos_pcie_sideband_dbi_r_mode(ep, true); - val = readl(base + reg); + dw_pcie_read(base + reg, size, &val); exynos_pcie_sideband_dbi_r_mode(ep, false); return val; } -static void exynos_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, - u32 reg, u32 val) +static void exynos_pcie_write_dbi(struct dw_pcie *pci, void __iomem *base, + u32 reg, size_t size, u32 val) { struct exynos_pcie *ep = to_exynos_pcie(pci); exynos_pcie_sideband_dbi_w_mode(ep, true); - writel(val, base + reg); + dw_pcie_write(base + reg, size, val); exynos_pcie_sideband_dbi_w_mode(ep, false); } @@ -646,8 +646,8 @@ static int __init exynos_add_pcie_port(struct exynos_pcie *ep, } static const struct dw_pcie_ops dw_pcie_ops = { - .readl_dbi = exynos_pcie_readl_dbi, - .writel_dbi = exynos_pcie_writel_dbi, + .read_dbi = exynos_pcie_read_dbi, + .write_dbi = exynos_pcie_write_dbi, .link_up = exynos_pcie_link_up, }; diff --git a/drivers/pci/dwc/pcie-designware.c b/drivers/pci/dwc/pcie-designware.c index ea403e2240cf..734acac1926d 100644 --- a/drivers/pci/dwc/pcie-designware.c +++ b/drivers/pci/dwc/pcie-designware.c @@ -61,21 +61,35 @@ int dw_pcie_write(void __iomem *addr, int size, u32 val) return PCIBIOS_SUCCESSFUL; } -u32 __dw_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg) +u32 __dw_pcie_read_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + size_t size) { - if (pci->ops->readl_dbi) - return pci->ops->readl_dbi(pci, base, reg); + int ret; + u32 val; + + if (pci->ops->read_dbi) + return pci->ops->read_dbi(pci, base, reg, size); - return readl(base + reg); + ret = dw_pcie_read(base + reg, size, &val); + if (ret) + dev_err(pci->dev, "read DBI address failed\n"); + + return val; } -void __dw_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, - u32 val) +void __dw_pcie_write_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + size_t size, u32 val) { - if (pci->ops->writel_dbi) - pci->ops->writel_dbi(pci, base, reg, val); - else - writel(val, base + reg); + int ret; + + if (pci->ops->write_dbi) { + pci->ops->write_dbi(pci, base, reg, size, val); + return; + } + + ret = dw_pcie_write(base + reg, size, val); + if (ret) + dev_err(pci->dev, "write DBI address failed\n"); } static u32 dw_pcie_readl_unroll(struct dw_pcie *pci, u32 index, u32 reg) diff --git a/drivers/pci/dwc/pcie-designware.h b/drivers/pci/dwc/pcie-designware.h index 09b334a12892..bfaf2b850a88 100644 --- a/drivers/pci/dwc/pcie-designware.h +++ b/drivers/pci/dwc/pcie-designware.h @@ -144,9 +144,10 @@ struct pcie_port { struct dw_pcie_ops { u64 (*cpu_addr_fixup)(u64 cpu_addr); - u32 (*readl_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg); - void (*writel_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, - u32 val); + u32 (*read_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, + size_t size); + void (*write_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, + size_t size, u32 val); int (*link_up)(struct dw_pcie *pcie); }; @@ -164,9 +165,10 @@ struct dw_pcie { int dw_pcie_read(void __iomem *addr, int size, u32 *val); int dw_pcie_write(void __iomem *addr, int size, u32 val); -u32 __dw_pcie_readl_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg); -void __dw_pcie_writel_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, - u32 val); +u32 __dw_pcie_read_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + size_t size); +void __dw_pcie_write_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, + size_t size, u32 val); int dw_pcie_link_up(struct dw_pcie *pci); int dw_pcie_wait_for_link(struct dw_pcie *pci); void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, @@ -176,12 +178,12 @@ void dw_pcie_setup(struct dw_pcie *pci); static inline void dw_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val) { - __dw_pcie_writel_dbi(pci, pci->dbi_base, reg, val); + __dw_pcie_write_dbi(pci, pci->dbi_base, reg, 0x4, val); } static inline u32 dw_pcie_readl_dbi(struct dw_pcie *pci, u32 reg) { - return __dw_pcie_readl_dbi(pci, pci->dbi_base, reg); + return __dw_pcie_read_dbi(pci, pci->dbi_base, reg, 0x4); } #ifdef CONFIG_PCIE_DW_HOST -- cgit v1.2.3-59-g8ed1b From edd45e3968299f9b4635bdfeca1edab842d81eac Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 13 Mar 2017 19:13:27 +0530 Subject: PCI: dwc: designware: Move _unroll configurations to a separate function No functional change. Rename dw_pcie_writel_unroll/dw_pcie_readl_unroll to dw_pcie_writel_ob_unroll/dw_pcie_readl_ob_unroll respectively as these functions are used to perform only outbound configurations. Also move these _unroll configurations to a separate function. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pcie-designware.c | 97 +++++++++++++++++++++++---------------- 1 file changed, 58 insertions(+), 39 deletions(-) diff --git a/drivers/pci/dwc/pcie-designware.c b/drivers/pci/dwc/pcie-designware.c index 734acac1926d..54de468a745e 100644 --- a/drivers/pci/dwc/pcie-designware.c +++ b/drivers/pci/dwc/pcie-designware.c @@ -92,21 +92,56 @@ void __dw_pcie_write_dbi(struct dw_pcie *pci, void __iomem *base, u32 reg, dev_err(pci->dev, "write DBI address failed\n"); } -static u32 dw_pcie_readl_unroll(struct dw_pcie *pci, u32 index, u32 reg) +static u32 dw_pcie_readl_ob_unroll(struct dw_pcie *pci, u32 index, u32 reg) { u32 offset = PCIE_GET_ATU_OUTB_UNR_REG_OFFSET(index); return dw_pcie_readl_dbi(pci, offset + reg); } -static void dw_pcie_writel_unroll(struct dw_pcie *pci, u32 index, u32 reg, - u32 val) +static void dw_pcie_writel_ob_unroll(struct dw_pcie *pci, u32 index, u32 reg, + u32 val) { u32 offset = PCIE_GET_ATU_OUTB_UNR_REG_OFFSET(index); dw_pcie_writel_dbi(pci, offset + reg, val); } +void dw_pcie_prog_outbound_atu_unroll(struct dw_pcie *pci, int index, int type, + u64 cpu_addr, u64 pci_addr, u32 size) +{ + u32 retries, val; + + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_LOWER_BASE, + lower_32_bits(cpu_addr)); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_UPPER_BASE, + upper_32_bits(cpu_addr)); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_LIMIT, + lower_32_bits(cpu_addr + size - 1)); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_LOWER_TARGET, + lower_32_bits(pci_addr)); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_UPPER_TARGET, + upper_32_bits(pci_addr)); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL1, + type); + dw_pcie_writel_ob_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL2, + PCIE_ATU_ENABLE); + + /* + * Make sure ATU enable takes effect before any subsequent config + * and I/O accesses. + */ + for (retries = 0; retries < LINK_WAIT_MAX_IATU_RETRIES; retries++) { + val = dw_pcie_readl_ob_unroll(pci, index, + PCIE_ATU_UNR_REGION_CTRL2); + if (val & PCIE_ATU_ENABLE) + return; + + usleep_range(LINK_WAIT_IATU_MIN, LINK_WAIT_IATU_MAX); + } + dev_err(pci->dev, "outbound iATU is not being enabled\n"); +} + void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, int type, u64 cpu_addr, u64 pci_addr, u32 size) { @@ -116,54 +151,38 @@ void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, int type, cpu_addr = pci->ops->cpu_addr_fixup(cpu_addr); if (pci->iatu_unroll_enabled) { - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_LOWER_BASE, - lower_32_bits(cpu_addr)); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_UPPER_BASE, - upper_32_bits(cpu_addr)); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_LIMIT, - lower_32_bits(cpu_addr + size - 1)); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_LOWER_TARGET, - lower_32_bits(pci_addr)); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_UPPER_TARGET, - upper_32_bits(pci_addr)); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL1, - type); - dw_pcie_writel_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL2, - PCIE_ATU_ENABLE); - } else { - dw_pcie_writel_dbi(pci, PCIE_ATU_VIEWPORT, - PCIE_ATU_REGION_OUTBOUND | index); - dw_pcie_writel_dbi(pci, PCIE_ATU_LOWER_BASE, - lower_32_bits(cpu_addr)); - dw_pcie_writel_dbi(pci, PCIE_ATU_UPPER_BASE, - upper_32_bits(cpu_addr)); - dw_pcie_writel_dbi(pci, PCIE_ATU_LIMIT, - lower_32_bits(cpu_addr + size - 1)); - dw_pcie_writel_dbi(pci, PCIE_ATU_LOWER_TARGET, - lower_32_bits(pci_addr)); - dw_pcie_writel_dbi(pci, PCIE_ATU_UPPER_TARGET, - upper_32_bits(pci_addr)); - dw_pcie_writel_dbi(pci, PCIE_ATU_CR1, type); - dw_pcie_writel_dbi(pci, PCIE_ATU_CR2, PCIE_ATU_ENABLE); + dw_pcie_prog_outbound_atu_unroll(pci, index, type, cpu_addr, + pci_addr, size); + return; } + dw_pcie_writel_dbi(pci, PCIE_ATU_VIEWPORT, + PCIE_ATU_REGION_OUTBOUND | index); + dw_pcie_writel_dbi(pci, PCIE_ATU_LOWER_BASE, + lower_32_bits(cpu_addr)); + dw_pcie_writel_dbi(pci, PCIE_ATU_UPPER_BASE, + upper_32_bits(cpu_addr)); + dw_pcie_writel_dbi(pci, PCIE_ATU_LIMIT, + lower_32_bits(cpu_addr + size - 1)); + dw_pcie_writel_dbi(pci, PCIE_ATU_LOWER_TARGET, + lower_32_bits(pci_addr)); + dw_pcie_writel_dbi(pci, PCIE_ATU_UPPER_TARGET, + upper_32_bits(pci_addr)); + dw_pcie_writel_dbi(pci, PCIE_ATU_CR1, type); + dw_pcie_writel_dbi(pci, PCIE_ATU_CR2, PCIE_ATU_ENABLE); + /* * Make sure ATU enable takes effect before any subsequent config * and I/O accesses. */ for (retries = 0; retries < LINK_WAIT_MAX_IATU_RETRIES; retries++) { - if (pci->iatu_unroll_enabled) - val = dw_pcie_readl_unroll(pci, index, - PCIE_ATU_UNR_REGION_CTRL2); - else - val = dw_pcie_readl_dbi(pci, PCIE_ATU_CR2); - + val = dw_pcie_readl_dbi(pci, PCIE_ATU_CR2); if (val == PCIE_ATU_ENABLE) return; usleep_range(LINK_WAIT_IATU_MIN, LINK_WAIT_IATU_MAX); } - dev_err(pci->dev, "iATU is not being enabled\n"); + dev_err(pci->dev, "outbound iATU is not being enabled\n"); } int dw_pcie_wait_for_link(struct dw_pcie *pci) -- cgit v1.2.3-59-g8ed1b From d4c7d1a089d6fddf38c4fcdc91f15791bfb59a95 Mon Sep 17 00:00:00 2001 From: Keerthy Date: Mon, 13 Mar 2017 19:13:28 +0530 Subject: PCI: dwc: dra7xx: Push request_irq() call to the bottom of probe Currently devm_request_irq() is being called before base, PCI fields of dra7xx_pcie structure are populated. It is called even before pm_runtime_enable() and pm_runtime_get_sync() are called. This will lead to exceptions if in case an interrupt is triggered before the all of the above are done. Hence push the devm_request_irq() call to the end of the probe. Signed-off-by: Keerthy Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pci-dra7xx.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/pci/dwc/pci-dra7xx.c b/drivers/pci/dwc/pci-dra7xx.c index 07c45ec07662..7c9ed6a6675c 100644 --- a/drivers/pci/dwc/pci-dra7xx.c +++ b/drivers/pci/dwc/pci-dra7xx.c @@ -410,13 +410,6 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) return -EINVAL; } - ret = devm_request_irq(dev, irq, dra7xx_pcie_irq_handler, - IRQF_SHARED, "dra7xx-pcie-main", dra7xx); - if (ret) { - dev_err(dev, "failed to request irq\n"); - return ret; - } - res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "ti_conf"); base = devm_ioremap_nocache(dev, res->start, resource_size(res)); if (!base) @@ -478,6 +471,13 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) if (ret < 0) goto err_gpio; + ret = devm_request_irq(dev, irq, dra7xx_pcie_irq_handler, + IRQF_SHARED, "dra7xx-pcie-main", dra7xx); + if (ret) { + dev_err(dev, "failed to request irq\n"); + goto err_gpio; + } + return 0; err_gpio: -- cgit v1.2.3-59-g8ed1b From 0b131b139467db844874f31b46bb638b541d9d3f Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Mon, 27 Mar 2017 17:46:14 -0700 Subject: PCI: Fix typo pci_cfg_access_lock() comment There is no pci_cfg_access_unlocked(). I think the author meant pci_cfg_access_unlock(). Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas --- drivers/pci/access.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/access.c b/drivers/pci/access.c index 8b7382705bf2..5b7842b03571 100644 --- a/drivers/pci/access.c +++ b/drivers/pci/access.c @@ -629,7 +629,7 @@ void pci_vpd_release(struct pci_dev *dev) * * When access is locked, any userspace reads or writes to config * space and concurrent lock requests will sleep until access is - * allowed via pci_cfg_access_unlocked again. + * allowed via pci_cfg_access_unlock() again. */ void pci_cfg_access_lock(struct pci_dev *dev) { -- cgit v1.2.3-59-g8ed1b From 433fcf6b7b31f1f233dd50aeb9d066a0f6ed4b9d Mon Sep 17 00:00:00 2001 From: "K. Y. Srinivasan" Date: Fri, 24 Mar 2017 11:07:21 -0700 Subject: PCI: hv: Specify CPU_AFFINITY_ALL for MSI affinity when >= 32 CPUs When we have 32 or more CPUs in the affinity mask, we should use a special constant to specify that to the host. Fix this issue. Signed-off-by: K. Y. Srinivasan Signed-off-by: Bjorn Helgaas Reviewed-by: Long Li Cc: --- drivers/pci/host/pci-hyperv.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c index a1b3c1957d43..8c952f694c2f 100644 --- a/drivers/pci/host/pci-hyperv.c +++ b/drivers/pci/host/pci-hyperv.c @@ -72,6 +72,7 @@ enum { PCI_PROTOCOL_VERSION_CURRENT = PCI_PROTOCOL_VERSION_1_1 }; +#define CPU_AFFINITY_ALL -1ULL #define PCI_CONFIG_MMIO_LENGTH 0x2000 #define CFG_PAGE_OFFSET 0x1000 #define CFG_PAGE_SIZE (PCI_CONFIG_MMIO_LENGTH - CFG_PAGE_OFFSET) @@ -898,9 +899,13 @@ static void hv_compose_msi_msg(struct irq_data *data, struct msi_msg *msg) * processors because Hyper-V only supports 64 in a guest. */ affinity = irq_data_get_affinity_mask(data); - for_each_cpu_and(cpu, affinity, cpu_online_mask) { - int_pkt->int_desc.cpu_mask |= - (1ULL << vmbus_cpu_number_to_vp_number(cpu)); + if (cpumask_weight(affinity) >= 32) { + int_pkt->int_desc.cpu_mask = CPU_AFFINITY_ALL; + } else { + for_each_cpu_and(cpu, affinity, cpu_online_mask) { + int_pkt->int_desc.cpu_mask |= + (1ULL << vmbus_cpu_number_to_vp_number(cpu)); + } } ret = vmbus_sendpacket(hpdev->hbus->hdev->channel, int_pkt, -- cgit v1.2.3-59-g8ed1b From 59c58ceeea9cdc6144d7b0303753e6bd26d87455 Mon Sep 17 00:00:00 2001 From: "K. Y. Srinivasan" Date: Fri, 24 Mar 2017 11:07:22 -0700 Subject: PCI: hv: Allocate interrupt descriptors with GFP_ATOMIC The memory allocation here needs to be non-blocking. Fix the issue. Signed-off-by: K. Y. Srinivasan Signed-off-by: Bjorn Helgaas Reviewed-by: Long Li Cc: --- drivers/pci/host/pci-hyperv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c index 8c952f694c2f..e73880c5d979 100644 --- a/drivers/pci/host/pci-hyperv.c +++ b/drivers/pci/host/pci-hyperv.c @@ -878,7 +878,7 @@ static void hv_compose_msi_msg(struct irq_data *data, struct msi_msg *msg) hv_int_desc_free(hpdev, int_desc); } - int_desc = kzalloc(sizeof(*int_desc), GFP_KERNEL); + int_desc = kzalloc(sizeof(*int_desc), GFP_ATOMIC); if (!int_desc) goto drop_reference; -- cgit v1.2.3-59-g8ed1b From 9b3fe6796d7c0e0c2b87243ce0c7f4744c54efad Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 28 Mar 2017 08:42:49 -0700 Subject: PCI: imx6: Add code to support i.MX7D Add various bits of code needed to support i.MX7D variant of the IP. Signed-off-by: Andrey Smirnov Signed-off-by: Bjorn Helgaas Reviewed-by: Lucas Stach Acked-by: Lee Jones Acked-by: Rob Herring Cc: yurovsky@gmail.com Cc: Mark Rutland Cc: Fabio Estevam Cc: Dong Aisheng Cc: linux-arm-kernel@lists.infradead.org Cc: devicetree@vger.kernel.org --- .../devicetree/bindings/pci/fsl,imx6q-pcie.txt | 14 ++- drivers/pci/dwc/pci-imx6.c | 120 ++++++++++++++++----- include/linux/mfd/syscon/imx7-iomuxc-gpr.h | 4 + 3 files changed, 112 insertions(+), 26 deletions(-) diff --git a/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.txt b/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.txt index 83aeb1f5a645..e3d5680875b1 100644 --- a/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.txt +++ b/Documentation/devicetree/bindings/pci/fsl,imx6q-pcie.txt @@ -4,7 +4,11 @@ This PCIe host controller is based on the Synopsis Designware PCIe IP and thus inherits all the common properties defined in designware-pcie.txt. Required properties: -- compatible: "fsl,imx6q-pcie", "fsl,imx6sx-pcie", "fsl,imx6qp-pcie" +- compatible: + - "fsl,imx6q-pcie" + - "fsl,imx6sx-pcie", + - "fsl,imx6qp-pcie" + - "fsl,imx7d-pcie" - reg: base address and length of the PCIe controller - interrupts: A list of interrupt outputs of the controller. Must contain an entry for each entry in the interrupt-names property. @@ -34,6 +38,14 @@ Additional required properties for imx6sx-pcie: - clock names: Must include the following additional entries: - "pcie_inbound_axi" +Additional required properties for imx7d-pcie: +- power-domains: Must be set to a phandle pointing to PCIE_PHY power domain +- resets: Must contain phandles to PCIe-related reset lines exposed by SRC + IP block +- reset-names: Must contain the following entires: + - "pciephy" + - "apps" + Example: pcie@0x01000000 { diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index 801e46cd266d..e0c36d6f5a21 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include @@ -27,6 +28,7 @@ #include #include #include +#include #include "pcie-designware.h" @@ -36,6 +38,7 @@ enum imx6_pcie_variants { IMX6Q, IMX6SX, IMX6QP, + IMX7D, }; struct imx6_pcie { @@ -47,6 +50,8 @@ struct imx6_pcie { struct clk *pcie_inbound_axi; struct clk *pcie; struct regmap *iomuxc_gpr; + struct reset_control *pciephy_reset; + struct reset_control *apps_reset; enum imx6_pcie_variants variant; u32 tx_deemph_gen1; u32 tx_deemph_gen2_3p5db; @@ -56,6 +61,11 @@ struct imx6_pcie { int link_gen; }; +/* Parameters for the waiting for PCIe PHY PLL to lock on i.MX7 */ +#define PHY_PLL_LOCK_WAIT_MAX_RETRIES 2000 +#define PHY_PLL_LOCK_WAIT_USLEEP_MIN 50 +#define PHY_PLL_LOCK_WAIT_USLEEP_MAX 200 + /* PCIe Root Complex registers (memory-mapped) */ #define PCIE_RC_LCR 0x7c #define PCIE_RC_LCR_MAX_LINK_SPEEDS_GEN1 0x1 @@ -248,6 +258,10 @@ static int imx6q_pcie_abort_handler(unsigned long addr, static void imx6_pcie_assert_core_reset(struct imx6_pcie *imx6_pcie) { switch (imx6_pcie->variant) { + case IMX7D: + reset_control_assert(imx6_pcie->pciephy_reset); + reset_control_assert(imx6_pcie->apps_reset); + break; case IMX6SX: regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, IMX6SX_GPR12_PCIE_TEST_POWERDOWN, @@ -303,11 +317,32 @@ static int imx6_pcie_enable_ref_clk(struct imx6_pcie *imx6_pcie) regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR1, IMX6Q_GPR1_PCIE_REF_CLK_EN, 1 << 16); break; + case IMX7D: + break; } return ret; } +static void imx7d_pcie_wait_for_phy_pll_lock(struct imx6_pcie *imx6_pcie) +{ + u32 val; + unsigned int retries; + struct device *dev = imx6_pcie->pci->dev; + + for (retries = 0; retries < PHY_PLL_LOCK_WAIT_MAX_RETRIES; retries++) { + regmap_read(imx6_pcie->iomuxc_gpr, IOMUXC_GPR22, &val); + + if (val & IMX7D_GPR22_PCIE_PHY_PLL_LOCKED) + return; + + usleep_range(PHY_PLL_LOCK_WAIT_USLEEP_MIN, + PHY_PLL_LOCK_WAIT_USLEEP_MAX); + } + + dev_err(dev, "PCIe PLL lock timeout\n"); +} + static void imx6_pcie_deassert_core_reset(struct imx6_pcie *imx6_pcie) { struct dw_pcie *pci = imx6_pcie->pci; @@ -351,6 +386,10 @@ static void imx6_pcie_deassert_core_reset(struct imx6_pcie *imx6_pcie) } switch (imx6_pcie->variant) { + case IMX7D: + reset_control_deassert(imx6_pcie->pciephy_reset); + imx7d_pcie_wait_for_phy_pll_lock(imx6_pcie); + break; case IMX6SX: regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR5, IMX6SX_GPR5_PCIE_BTNRST_RESET, 0); @@ -377,35 +416,44 @@ err_pcie_bus: static void imx6_pcie_init_phy(struct imx6_pcie *imx6_pcie) { - if (imx6_pcie->variant == IMX6SX) + switch (imx6_pcie->variant) { + case IMX7D: + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, + IMX7D_GPR12_PCIE_PHY_REFCLK_SEL, 0); + break; + case IMX6SX: regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, IMX6SX_GPR12_PCIE_RX_EQ_MASK, IMX6SX_GPR12_PCIE_RX_EQ_2); + /* FALLTHROUGH */ + default: + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, + IMX6Q_GPR12_PCIE_CTL_2, 0 << 10); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, - IMX6Q_GPR12_PCIE_CTL_2, 0 << 10); + /* configure constant input signal to the pcie ctrl and phy */ + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, + IMX6Q_GPR12_LOS_LEVEL, 9 << 4); + + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, + IMX6Q_GPR8_TX_DEEMPH_GEN1, + imx6_pcie->tx_deemph_gen1 << 0); + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, + IMX6Q_GPR8_TX_DEEMPH_GEN2_3P5DB, + imx6_pcie->tx_deemph_gen2_3p5db << 6); + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, + IMX6Q_GPR8_TX_DEEMPH_GEN2_6DB, + imx6_pcie->tx_deemph_gen2_6db << 12); + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, + IMX6Q_GPR8_TX_SWING_FULL, + imx6_pcie->tx_swing_full << 18); + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, + IMX6Q_GPR8_TX_SWING_LOW, + imx6_pcie->tx_swing_low << 25); + break; + } - /* configure constant input signal to the pcie ctrl and phy */ regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, IMX6Q_GPR12_DEVICE_TYPE, PCI_EXP_TYPE_ROOT_PORT << 12); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, - IMX6Q_GPR12_LOS_LEVEL, 9 << 4); - - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, - IMX6Q_GPR8_TX_DEEMPH_GEN1, - imx6_pcie->tx_deemph_gen1 << 0); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, - IMX6Q_GPR8_TX_DEEMPH_GEN2_3P5DB, - imx6_pcie->tx_deemph_gen2_3p5db << 6); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, - IMX6Q_GPR8_TX_DEEMPH_GEN2_6DB, - imx6_pcie->tx_deemph_gen2_6db << 12); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, - IMX6Q_GPR8_TX_SWING_FULL, - imx6_pcie->tx_swing_full << 18); - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR8, - IMX6Q_GPR8_TX_SWING_LOW, - imx6_pcie->tx_swing_low << 25); } static int imx6_pcie_wait_for_link(struct imx6_pcie *imx6_pcie) @@ -469,8 +517,11 @@ static int imx6_pcie_establish_link(struct imx6_pcie *imx6_pcie) dw_pcie_writel_dbi(pci, PCIE_RC_LCR, tmp); /* Start LTSSM. */ - regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, - IMX6Q_GPR12_PCIE_CTL_2, 1 << 10); + if (imx6_pcie->variant == IMX7D) + reset_control_deassert(imx6_pcie->apps_reset); + else + regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12, + IMX6Q_GPR12_PCIE_CTL_2, 1 << 10); ret = imx6_pcie_wait_for_link(imx6_pcie); if (ret) @@ -653,13 +704,31 @@ static int __init imx6_pcie_probe(struct platform_device *pdev) return PTR_ERR(imx6_pcie->pcie); } - if (imx6_pcie->variant == IMX6SX) { + switch (imx6_pcie->variant) { + case IMX6SX: imx6_pcie->pcie_inbound_axi = devm_clk_get(dev, "pcie_inbound_axi"); if (IS_ERR(imx6_pcie->pcie_inbound_axi)) { dev_err(dev, "pcie_inbound_axi clock missing or invalid\n"); return PTR_ERR(imx6_pcie->pcie_inbound_axi); } + break; + case IMX7D: + imx6_pcie->pciephy_reset = devm_reset_control_get(dev, + "pciephy"); + if (IS_ERR(imx6_pcie->pciephy_reset)) { + dev_err(dev, "Failed to get PCIEPHY reset contol\n"); + return PTR_ERR(imx6_pcie->pciephy_reset); + } + + imx6_pcie->apps_reset = devm_reset_control_get(dev, "apps"); + if (IS_ERR(imx6_pcie->apps_reset)) { + dev_err(dev, "Failed to get PCIE APPS reset contol\n"); + return PTR_ERR(imx6_pcie->apps_reset); + } + break; + default: + break; } /* Grab GPR config register range */ @@ -718,6 +787,7 @@ static const struct of_device_id imx6_pcie_of_match[] = { { .compatible = "fsl,imx6q-pcie", .data = (void *)IMX6Q, }, { .compatible = "fsl,imx6sx-pcie", .data = (void *)IMX6SX, }, { .compatible = "fsl,imx6qp-pcie", .data = (void *)IMX6QP, }, + { .compatible = "fsl,imx7d-pcie", .data = (void *)IMX7D, }, {}, }; diff --git a/include/linux/mfd/syscon/imx7-iomuxc-gpr.h b/include/linux/mfd/syscon/imx7-iomuxc-gpr.h index 4585d6105d68..abbd52466573 100644 --- a/include/linux/mfd/syscon/imx7-iomuxc-gpr.h +++ b/include/linux/mfd/syscon/imx7-iomuxc-gpr.h @@ -44,4 +44,8 @@ #define IMX7D_GPR5_CSI_MUX_CONTROL_MIPI (0x1 << 4) +#define IMX7D_GPR12_PCIE_PHY_REFCLK_SEL BIT(5) + +#define IMX7D_GPR22_PCIE_PHY_PLL_LOCKED BIT(31) + #endif /* __LINUX_IMX7_IOMUXC_GPR_H */ -- cgit v1.2.3-59-g8ed1b From bde4a5a00e761f55be92f62378cf5024ced79ee3 Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 28 Mar 2017 08:42:50 -0700 Subject: PCI: imx6: Allow probe deferral by reset GPIO Some designs implement reset GPIO via a GPIO expander connected to a peripheral bus. One such example would be i.MX7 Sabre board where said GPIO is provided by SPI shift register connected to a bitbanged SPI bus. To support such designs, allow reset GPIO request to defer probing of the driver. Signed-off-by: Andrey Smirnov Signed-off-by: Bjorn Helgaas Reviewed-by: Lucas Stach Cc: yurovsky@gmail.com Cc: Fabio Estevam Cc: Dong Aisheng Cc: linux-arm-kernel@lists.infradead.org --- drivers/pci/dwc/pci-imx6.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-) diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index e0c36d6f5a21..d62ee3553cfe 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -595,8 +595,8 @@ static struct dw_pcie_host_ops imx6_pcie_host_ops = { .host_init = imx6_pcie_host_init, }; -static int __init imx6_add_pcie_port(struct imx6_pcie *imx6_pcie, - struct platform_device *pdev) +static int imx6_add_pcie_port(struct imx6_pcie *imx6_pcie, + struct platform_device *pdev) { struct dw_pcie *pci = imx6_pcie->pci; struct pcie_port *pp = &pci->pp; @@ -636,7 +636,7 @@ static const struct dw_pcie_ops dw_pcie_ops = { .link_up = imx6_pcie_link_up, }; -static int __init imx6_pcie_probe(struct platform_device *pdev) +static int imx6_pcie_probe(struct platform_device *pdev) { struct device *dev = &pdev->dev; struct dw_pcie *pci; @@ -660,10 +660,6 @@ static int __init imx6_pcie_probe(struct platform_device *pdev) imx6_pcie->variant = (enum imx6_pcie_variants)of_device_get_match_data(dev); - /* Added for PCI abort handling */ - hook_fault_code(16 + 6, imx6q_pcie_abort_handler, SIGBUS, 0, - "imprecise external abort"); - dbi_base = platform_get_resource(pdev, IORESOURCE_MEM, 0); pci->dbi_base = devm_ioremap_resource(dev, dbi_base); if (IS_ERR(pci->dbi_base)) @@ -683,6 +679,8 @@ static int __init imx6_pcie_probe(struct platform_device *pdev) dev_err(dev, "unable to get reset gpio\n"); return ret; } + } else if (imx6_pcie->reset_gpio == -EPROBE_DEFER) { + return imx6_pcie->reset_gpio; } /* Fetch clocks */ @@ -796,11 +794,22 @@ static struct platform_driver imx6_pcie_driver = { .name = "imx6q-pcie", .of_match_table = imx6_pcie_of_match, }, + .probe = imx6_pcie_probe, .shutdown = imx6_pcie_shutdown, }; static int __init imx6_pcie_init(void) { - return platform_driver_probe(&imx6_pcie_driver, imx6_pcie_probe); + /* + * Since probe() can be deferred we need to make sure that + * hook_fault_code is not called after __init memory is freed + * by kernel and since imx6q_pcie_abort_handler() is a no-op, + * we can install the handler here without risking it + * accessing some uninitialized driver state. + */ + hook_fault_code(16 + 6, imx6q_pcie_abort_handler, SIGBUS, 0, + "imprecise external abort"); + + return platform_driver_register(&imx6_pcie_driver); } device_initcall(imx6_pcie_init); -- cgit v1.2.3-59-g8ed1b From e6dcd87fff69a9d454104b65569074855cf95b1e Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 28 Mar 2017 08:42:51 -0700 Subject: PCI: imx6: Do not wait for speed change on i.MX7 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As can be seen from [1]: "...the different behavior between iMX6Q PCIe and iMX7D PCIe maybe caused by the different controller version. Regarding to the DOC description, the DIRECT_SPEED_CHANGE should be cleared after the speed change from GEN1 to GEN2. Unfortunately, when GEN1 device is used, the behavior is not documented. So, IC design guys run the simulation and find out the following behaviors: 1. DIRECT_SPEED_CHANGE will be cleared in 7D after speed change from GEN1 to GEN2. This matches doc’s description 2. set MAX link speed(PCIE_CAP_TARGET_LINK_SPEED=0x01) as GEN1 and re-run the simulation, DIRECT_SPEED_CHANGE will not be cleared; remain as 1, this matches your result, but function test is passed, so this bit should not affect the normal PCIe function." imx6_pcie_wait_for_speed_change() will report false failures for Gen1 -> Gen1 speed transition, so avoid doing that check and just rely on imx6_pcie_wait_for_link() only. [1] https://community.nxp.com/message/867943 Signed-off-by: Andrey Smirnov Signed-off-by: Bjorn Helgaas Reviewed-by: Lucas Stach Cc: yurovsky@gmail.com Cc: Fabio Estevam Cc: Dong Aisheng Cc: linux-arm-kernel@lists.infradead.org --- drivers/pci/dwc/pci-imx6.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index d62ee3553cfe..793b008553a0 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -545,10 +545,21 @@ static int imx6_pcie_establish_link(struct imx6_pcie *imx6_pcie) tmp |= PORT_LOGIC_SPEED_CHANGE; dw_pcie_writel_dbi(pci, PCIE_LINK_WIDTH_SPEED_CONTROL, tmp); - ret = imx6_pcie_wait_for_speed_change(imx6_pcie); - if (ret) { - dev_err(dev, "Failed to bring link up!\n"); - goto err_reset_phy; + if (imx6_pcie->variant != IMX7D) { + /* + * On i.MX7, DIRECT_SPEED_CHANGE behaves differently + * from i.MX6 family when no link speed transition + * occurs and we go Gen1 -> yep, Gen1. The difference + * is that, in such case, it will not be cleared by HW + * which will cause the following code to report false + * failure. + */ + + ret = imx6_pcie_wait_for_speed_change(imx6_pcie); + if (ret) { + dev_err(dev, "Failed to bring link up!\n"); + goto err_reset_phy; + } } /* Make sure link training is finished as well! */ -- cgit v1.2.3-59-g8ed1b From 93b226f9c65a951a91617f87ba1f05f14e59f26f Mon Sep 17 00:00:00 2001 From: Andrey Smirnov Date: Tue, 28 Mar 2017 08:42:52 -0700 Subject: PCI: imx6: Do not switch speed if Gen2 is disabled Save a bit of time and avoid going through link speed change procedure in configuration where link max speed is limited to Gen1 in DT. Signed-off-by: Andrey Smirnov Signed-off-by: Bjorn Helgaas Reviewed-by: Lucas Stach Cc: yurovsky@gmail.com Cc: Fabio Estevam Cc: Dong Aisheng Cc: linux-arm-kernel@lists.infradead.org --- drivers/pci/dwc/pci-imx6.c | 52 +++++++++++++++++++++++----------------------- 1 file changed, 26 insertions(+), 26 deletions(-) diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index 793b008553a0..102edcf1e261 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -533,40 +533,40 @@ static int imx6_pcie_establish_link(struct imx6_pcie *imx6_pcie) tmp &= ~PCIE_RC_LCR_MAX_LINK_SPEEDS_MASK; tmp |= PCIE_RC_LCR_MAX_LINK_SPEEDS_GEN2; dw_pcie_writel_dbi(pci, PCIE_RC_LCR, tmp); - } else { - dev_info(dev, "Link: Gen2 disabled\n"); - } - - /* - * Start Directed Speed Change so the best possible speed both link - * partners support can be negotiated. - */ - tmp = dw_pcie_readl_dbi(pci, PCIE_LINK_WIDTH_SPEED_CONTROL); - tmp |= PORT_LOGIC_SPEED_CHANGE; - dw_pcie_writel_dbi(pci, PCIE_LINK_WIDTH_SPEED_CONTROL, tmp); - if (imx6_pcie->variant != IMX7D) { /* - * On i.MX7, DIRECT_SPEED_CHANGE behaves differently - * from i.MX6 family when no link speed transition - * occurs and we go Gen1 -> yep, Gen1. The difference - * is that, in such case, it will not be cleared by HW - * which will cause the following code to report false - * failure. + * Start Directed Speed Change so the best possible + * speed both link partners support can be negotiated. */ + tmp = dw_pcie_readl_dbi(pci, PCIE_LINK_WIDTH_SPEED_CONTROL); + tmp |= PORT_LOGIC_SPEED_CHANGE; + dw_pcie_writel_dbi(pci, PCIE_LINK_WIDTH_SPEED_CONTROL, tmp); + + if (imx6_pcie->variant != IMX7D) { + /* + * On i.MX7, DIRECT_SPEED_CHANGE behaves differently + * from i.MX6 family when no link speed transition + * occurs and we go Gen1 -> yep, Gen1. The difference + * is that, in such case, it will not be cleared by HW + * which will cause the following code to report false + * failure. + */ + + ret = imx6_pcie_wait_for_speed_change(imx6_pcie); + if (ret) { + dev_err(dev, "Failed to bring link up!\n"); + goto err_reset_phy; + } + } - ret = imx6_pcie_wait_for_speed_change(imx6_pcie); + /* Make sure link training is finished as well! */ + ret = imx6_pcie_wait_for_link(imx6_pcie); if (ret) { dev_err(dev, "Failed to bring link up!\n"); goto err_reset_phy; } - } - - /* Make sure link training is finished as well! */ - ret = imx6_pcie_wait_for_link(imx6_pcie); - if (ret) { - dev_err(dev, "Failed to bring link up!\n"); - goto err_reset_phy; + } else { + dev_info(dev, "Link: Gen2 disabled\n"); } tmp = dw_pcie_readl_dbi(pci, PCIE_RC_LCSR); -- cgit v1.2.3-59-g8ed1b From 5e8cb4033807e39849b753e5399ec130c0995f1f Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 10 Apr 2017 19:25:10 +0530 Subject: PCI: endpoint: Add EP core layer to enable EP controller and EP functions Introduce a new EP core layer in order to support endpoint functions in linux kernel. This comprises the EPC library (Endpoint Controller Library) and EPF library (Endpoint Function Library). EPC library implements functions specific to an endpoint controller and EPF library implements functions specific to an endpoint function. Signed-off-by: Kishon Vijay Abraham I Acked-by: Joao Pinto Signed-off-by: Bjorn Helgaas --- drivers/Makefile | 2 + drivers/pci/Kconfig | 1 + drivers/pci/endpoint/Kconfig | 20 ++ drivers/pci/endpoint/Makefile | 6 + drivers/pci/endpoint/pci-epc-core.c | 576 ++++++++++++++++++++++++++++++++++++ drivers/pci/endpoint/pci-epc-mem.c | 143 +++++++++ drivers/pci/endpoint/pci-epf-core.c | 355 ++++++++++++++++++++++ include/linux/mod_devicetable.h | 10 + include/linux/pci-epc.h | 142 +++++++++ include/linux/pci-epf.h | 160 ++++++++++ 10 files changed, 1415 insertions(+) create mode 100644 drivers/pci/endpoint/Kconfig create mode 100644 drivers/pci/endpoint/Makefile create mode 100644 drivers/pci/endpoint/pci-epc-core.c create mode 100644 drivers/pci/endpoint/pci-epc-mem.c create mode 100644 drivers/pci/endpoint/pci-epf-core.c create mode 100644 include/linux/pci-epc.h create mode 100644 include/linux/pci-epf.h diff --git a/drivers/Makefile b/drivers/Makefile index 2eced9afba53..a5f8e43b2c4d 100644 --- a/drivers/Makefile +++ b/drivers/Makefile @@ -14,7 +14,9 @@ obj-$(CONFIG_GENERIC_PHY) += phy/ obj-$(CONFIG_PINCTRL) += pinctrl/ obj-$(CONFIG_GPIOLIB) += gpio/ obj-y += pwm/ + obj-$(CONFIG_PCI) += pci/ +obj-$(CONFIG_PCI_ENDPOINT) += pci/endpoint/ # PCI dwc controller drivers obj-y += pci/dwc/ diff --git a/drivers/pci/Kconfig b/drivers/pci/Kconfig index df141420c902..9747c1ec8c74 100644 --- a/drivers/pci/Kconfig +++ b/drivers/pci/Kconfig @@ -134,3 +134,4 @@ config PCI_HYPERV source "drivers/pci/hotplug/Kconfig" source "drivers/pci/dwc/Kconfig" source "drivers/pci/host/Kconfig" +source "drivers/pci/endpoint/Kconfig" diff --git a/drivers/pci/endpoint/Kconfig b/drivers/pci/endpoint/Kconfig new file mode 100644 index 000000000000..a5442ace7077 --- /dev/null +++ b/drivers/pci/endpoint/Kconfig @@ -0,0 +1,20 @@ +# +# PCI Endpoint Support +# + +menu "PCI Endpoint" + +config PCI_ENDPOINT + bool "PCI Endpoint Support" + help + Enable this configuration option to support configurable PCI + endpoint. This should be enabled if the platform has a PCI + controller that can operate in endpoint mode. + + Enabling this option will build the endpoint library, which + includes endpoint controller library and endpoint function + library. + + If in doubt, say "N" to disable Endpoint support. + +endmenu diff --git a/drivers/pci/endpoint/Makefile b/drivers/pci/endpoint/Makefile new file mode 100644 index 000000000000..dc1bc16491e6 --- /dev/null +++ b/drivers/pci/endpoint/Makefile @@ -0,0 +1,6 @@ +# +# Makefile for PCI Endpoint Support +# + +obj-$(CONFIG_PCI_ENDPOINT) += pci-epc-core.o pci-epf-core.o\ + pci-epc-mem.o diff --git a/drivers/pci/endpoint/pci-epc-core.c b/drivers/pci/endpoint/pci-epc-core.c new file mode 100644 index 000000000000..7c71dd94721c --- /dev/null +++ b/drivers/pci/endpoint/pci-epc-core.c @@ -0,0 +1,576 @@ +/** + * PCI Endpoint *Controller* (EPC) library + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include + +#include +#include + +static struct class *pci_epc_class; + +static void devm_pci_epc_release(struct device *dev, void *res) +{ + struct pci_epc *epc = *(struct pci_epc **)res; + + pci_epc_destroy(epc); +} + +static int devm_pci_epc_match(struct device *dev, void *res, void *match_data) +{ + struct pci_epc **epc = res; + + return *epc == match_data; +} + +/** + * pci_epc_put() - release the PCI endpoint controller + * @epc: epc returned by pci_epc_get() + * + * release the refcount the caller obtained by invoking pci_epc_get() + */ +void pci_epc_put(struct pci_epc *epc) +{ + if (!epc || IS_ERR(epc)) + return; + + module_put(epc->ops->owner); + put_device(&epc->dev); +} +EXPORT_SYMBOL_GPL(pci_epc_put); + +/** + * pci_epc_get() - get the PCI endpoint controller + * @epc_name: device name of the endpoint controller + * + * Invoke to get struct pci_epc * corresponding to the device name of the + * endpoint controller + */ +struct pci_epc *pci_epc_get(const char *epc_name) +{ + int ret = -EINVAL; + struct pci_epc *epc; + struct device *dev; + struct class_dev_iter iter; + + class_dev_iter_init(&iter, pci_epc_class, NULL, NULL); + while ((dev = class_dev_iter_next(&iter))) { + if (strcmp(epc_name, dev_name(dev))) + continue; + + epc = to_pci_epc(dev); + if (!try_module_get(epc->ops->owner)) { + ret = -EINVAL; + goto err; + } + + class_dev_iter_exit(&iter); + get_device(&epc->dev); + return epc; + } + +err: + class_dev_iter_exit(&iter); + return ERR_PTR(ret); +} +EXPORT_SYMBOL_GPL(pci_epc_get); + +/** + * pci_epc_stop() - stop the PCI link + * @epc: the link of the EPC device that has to be stopped + * + * Invoke to stop the PCI link + */ +void pci_epc_stop(struct pci_epc *epc) +{ + unsigned long flags; + + if (IS_ERR(epc) || !epc->ops->stop) + return; + + spin_lock_irqsave(&epc->lock, flags); + epc->ops->stop(epc); + spin_unlock_irqrestore(&epc->lock, flags); +} +EXPORT_SYMBOL_GPL(pci_epc_stop); + +/** + * pci_epc_start() - start the PCI link + * @epc: the link of *this* EPC device has to be started + * + * Invoke to start the PCI link + */ +int pci_epc_start(struct pci_epc *epc) +{ + int ret; + unsigned long flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->start) + return 0; + + spin_lock_irqsave(&epc->lock, flags); + ret = epc->ops->start(epc); + spin_unlock_irqrestore(&epc->lock, flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_start); + +/** + * pci_epc_raise_irq() - interrupt the host system + * @epc: the EPC device which has to interrupt the host + * @type: specify the type of interrupt; legacy or MSI + * @interrupt_num: the MSI interrupt number + * + * Invoke to raise an MSI or legacy interrupt + */ +int pci_epc_raise_irq(struct pci_epc *epc, enum pci_epc_irq_type type, + u8 interrupt_num) +{ + int ret; + unsigned long flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->raise_irq) + return 0; + + spin_lock_irqsave(&epc->lock, flags); + ret = epc->ops->raise_irq(epc, type, interrupt_num); + spin_unlock_irqrestore(&epc->lock, flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_raise_irq); + +/** + * pci_epc_get_msi() - get the number of MSI interrupt numbers allocated + * @epc: the EPC device to which MSI interrupts was requested + * + * Invoke to get the number of MSI interrupts allocated by the RC + */ +int pci_epc_get_msi(struct pci_epc *epc) +{ + int interrupt; + unsigned long flags; + + if (IS_ERR(epc)) + return 0; + + if (!epc->ops->get_msi) + return 0; + + spin_lock_irqsave(&epc->lock, flags); + interrupt = epc->ops->get_msi(epc); + spin_unlock_irqrestore(&epc->lock, flags); + + if (interrupt < 0) + return 0; + + interrupt = 1 << interrupt; + + return interrupt; +} +EXPORT_SYMBOL_GPL(pci_epc_get_msi); + +/** + * pci_epc_set_msi() - set the number of MSI interrupt numbers required + * @epc: the EPC device on which MSI has to be configured + * @interrupts: number of MSI interrupts required by the EPF + * + * Invoke to set the required number of MSI interrupts. + */ +int pci_epc_set_msi(struct pci_epc *epc, u8 interrupts) +{ + int ret; + u8 encode_int; + unsigned long flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->set_msi) + return 0; + + encode_int = order_base_2(interrupts); + + spin_lock_irqsave(&epc->lock, flags); + ret = epc->ops->set_msi(epc, encode_int); + spin_unlock_irqrestore(&epc->lock, flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_set_msi); + +/** + * pci_epc_unmap_addr() - unmap CPU address from PCI address + * @epc: the EPC device on which address is allocated + * @phys_addr: physical address of the local system + * + * Invoke to unmap the CPU address from PCI address. + */ +void pci_epc_unmap_addr(struct pci_epc *epc, phys_addr_t phys_addr) +{ + unsigned long flags; + + if (IS_ERR(epc)) + return; + + if (!epc->ops->unmap_addr) + return; + + spin_lock_irqsave(&epc->lock, flags); + epc->ops->unmap_addr(epc, phys_addr); + spin_unlock_irqrestore(&epc->lock, flags); +} +EXPORT_SYMBOL_GPL(pci_epc_unmap_addr); + +/** + * pci_epc_map_addr() - map CPU address to PCI address + * @epc: the EPC device on which address is allocated + * @phys_addr: physical address of the local system + * @pci_addr: PCI address to which the physical address should be mapped + * @size: the size of the allocation + * + * Invoke to map CPU address with PCI address. + */ +int pci_epc_map_addr(struct pci_epc *epc, phys_addr_t phys_addr, + u64 pci_addr, size_t size) +{ + int ret; + unsigned long flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->map_addr) + return 0; + + spin_lock_irqsave(&epc->lock, flags); + ret = epc->ops->map_addr(epc, phys_addr, pci_addr, size); + spin_unlock_irqrestore(&epc->lock, flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_map_addr); + +/** + * pci_epc_clear_bar() - reset the BAR + * @epc: the EPC device for which the BAR has to be cleared + * @bar: the BAR number that has to be reset + * + * Invoke to reset the BAR of the endpoint device. + */ +void pci_epc_clear_bar(struct pci_epc *epc, int bar) +{ + unsigned long flags; + + if (IS_ERR(epc)) + return; + + if (!epc->ops->clear_bar) + return; + + spin_lock_irqsave(&epc->lock, flags); + epc->ops->clear_bar(epc, bar); + spin_unlock_irqrestore(&epc->lock, flags); +} +EXPORT_SYMBOL_GPL(pci_epc_clear_bar); + +/** + * pci_epc_set_bar() - configure BAR in order for host to assign PCI addr space + * @epc: the EPC device on which BAR has to be configured + * @bar: the BAR number that has to be configured + * @size: the size of the addr space + * @flags: specify memory allocation/io allocation/32bit address/64 bit address + * + * Invoke to configure the BAR of the endpoint device. + */ +int pci_epc_set_bar(struct pci_epc *epc, enum pci_barno bar, + dma_addr_t bar_phys, size_t size, int flags) +{ + int ret; + unsigned long irq_flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->set_bar) + return 0; + + spin_lock_irqsave(&epc->lock, irq_flags); + ret = epc->ops->set_bar(epc, bar, bar_phys, size, flags); + spin_unlock_irqrestore(&epc->lock, irq_flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_set_bar); + +/** + * pci_epc_write_header() - write standard configuration header + * @epc: the EPC device to which the configuration header should be written + * @header: standard configuration header fields + * + * Invoke to write the configuration header to the endpoint controller. Every + * endpoint controller will have a dedicated location to which the standard + * configuration header would be written. The callback function should write + * the header fields to this dedicated location. + */ +int pci_epc_write_header(struct pci_epc *epc, struct pci_epf_header *header) +{ + int ret; + unsigned long flags; + + if (IS_ERR(epc)) + return -EINVAL; + + if (!epc->ops->write_header) + return 0; + + spin_lock_irqsave(&epc->lock, flags); + ret = epc->ops->write_header(epc, header); + spin_unlock_irqrestore(&epc->lock, flags); + + return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_write_header); + +/** + * pci_epc_add_epf() - bind PCI endpoint function to an endpoint controller + * @epc: the EPC device to which the endpoint function should be added + * @epf: the endpoint function to be added + * + * A PCI endpoint device can have one or more functions. In the case of PCIe, + * the specification allows up to 8 PCIe endpoint functions. Invoke + * pci_epc_add_epf() to add a PCI endpoint function to an endpoint controller. + */ +int pci_epc_add_epf(struct pci_epc *epc, struct pci_epf *epf) +{ + unsigned long flags; + + if (epf->epc) + return -EBUSY; + + if (IS_ERR(epc)) + return -EINVAL; + + if (epf->func_no > epc->max_functions - 1) + return -EINVAL; + + epf->epc = epc; + dma_set_coherent_mask(&epf->dev, epc->dev.coherent_dma_mask); + epf->dev.dma_mask = epc->dev.dma_mask; + + spin_lock_irqsave(&epc->lock, flags); + list_add_tail(&epf->list, &epc->pci_epf); + spin_unlock_irqrestore(&epc->lock, flags); + + return 0; +} +EXPORT_SYMBOL_GPL(pci_epc_add_epf); + +/** + * pci_epc_remove_epf() - remove PCI endpoint function from endpoint controller + * @epc: the EPC device from which the endpoint function should be removed + * @epf: the endpoint function to be removed + * + * Invoke to remove PCI endpoint function from the endpoint controller. + */ +void pci_epc_remove_epf(struct pci_epc *epc, struct pci_epf *epf) +{ + unsigned long flags; + + if (!epc || IS_ERR(epc)) + return; + + spin_lock_irqsave(&epc->lock, flags); + list_del(&epf->list); + spin_unlock_irqrestore(&epc->lock, flags); +} +EXPORT_SYMBOL_GPL(pci_epc_remove_epf); + +/** + * pci_epc_linkup() - Notify the EPF device that EPC device has established a + * connection with the Root Complex. + * @epc: the EPC device which has established link with the host + * + * Invoke to Notify the EPF device that the EPC device has established a + * connection with the Root Complex. + */ +void pci_epc_linkup(struct pci_epc *epc) +{ + unsigned long flags; + struct pci_epf *epf; + + if (!epc || IS_ERR(epc)) + return; + + spin_lock_irqsave(&epc->lock, flags); + list_for_each_entry(epf, &epc->pci_epf, list) + pci_epf_linkup(epf); + spin_unlock_irqrestore(&epc->lock, flags); +} +EXPORT_SYMBOL_GPL(pci_epc_linkup); + +/** + * pci_epc_destroy() - destroy the EPC device + * @epc: the EPC device that has to be destroyed + * + * Invoke to destroy the PCI EPC device + */ +void pci_epc_destroy(struct pci_epc *epc) +{ + device_unregister(&epc->dev); + kfree(epc); +} +EXPORT_SYMBOL_GPL(pci_epc_destroy); + +/** + * devm_pci_epc_destroy() - destroy the EPC device + * @dev: device that wants to destroy the EPC + * @epc: the EPC device that has to be destroyed + * + * Invoke to destroy the devres associated with this + * pci_epc and destroy the EPC device. + */ +void devm_pci_epc_destroy(struct device *dev, struct pci_epc *epc) +{ + int r; + + r = devres_destroy(dev, devm_pci_epc_release, devm_pci_epc_match, + epc); + dev_WARN_ONCE(dev, r, "couldn't find PCI EPC resource\n"); +} +EXPORT_SYMBOL_GPL(devm_pci_epc_destroy); + +/** + * __pci_epc_create() - create a new endpoint controller (EPC) device + * @dev: device that is creating the new EPC + * @ops: function pointers for performing EPC operations + * @owner: the owner of the module that creates the EPC device + * + * Invoke to create a new EPC device and add it to pci_epc class. + */ +struct pci_epc * +__pci_epc_create(struct device *dev, const struct pci_epc_ops *ops, + struct module *owner) +{ + int ret; + struct pci_epc *epc; + + if (WARN_ON(!dev)) { + ret = -EINVAL; + goto err_ret; + } + + epc = kzalloc(sizeof(*epc), GFP_KERNEL); + if (!epc) { + ret = -ENOMEM; + goto err_ret; + } + + spin_lock_init(&epc->lock); + INIT_LIST_HEAD(&epc->pci_epf); + + device_initialize(&epc->dev); + dma_set_coherent_mask(&epc->dev, dev->coherent_dma_mask); + epc->dev.class = pci_epc_class; + epc->dev.dma_mask = dev->dma_mask; + epc->ops = ops; + + ret = dev_set_name(&epc->dev, "%s", dev_name(dev)); + if (ret) + goto put_dev; + + ret = device_add(&epc->dev); + if (ret) + goto put_dev; + + return epc; + +put_dev: + put_device(&epc->dev); + kfree(epc); + +err_ret: + return ERR_PTR(ret); +} +EXPORT_SYMBOL_GPL(__pci_epc_create); + +/** + * __devm_pci_epc_create() - create a new endpoint controller (EPC) device + * @dev: device that is creating the new EPC + * @ops: function pointers for performing EPC operations + * @owner: the owner of the module that creates the EPC device + * + * Invoke to create a new EPC device and add it to pci_epc class. + * While at that, it also associates the device with the pci_epc using devres. + * On driver detach, release function is invoked on the devres data, + * then, devres data is freed. + */ +struct pci_epc * +__devm_pci_epc_create(struct device *dev, const struct pci_epc_ops *ops, + struct module *owner) +{ + struct pci_epc **ptr, *epc; + + ptr = devres_alloc(devm_pci_epc_release, sizeof(*ptr), GFP_KERNEL); + if (!ptr) + return ERR_PTR(-ENOMEM); + + epc = __pci_epc_create(dev, ops, owner); + if (!IS_ERR(epc)) { + *ptr = epc; + devres_add(dev, ptr); + } else { + devres_free(ptr); + } + + return epc; +} +EXPORT_SYMBOL_GPL(__devm_pci_epc_create); + +static int __init pci_epc_init(void) +{ + pci_epc_class = class_create(THIS_MODULE, "pci_epc"); + if (IS_ERR(pci_epc_class)) { + pr_err("failed to create pci epc class --> %ld\n", + PTR_ERR(pci_epc_class)); + return PTR_ERR(pci_epc_class); + } + + return 0; +} +module_init(pci_epc_init); + +static void __exit pci_epc_exit(void) +{ + class_destroy(pci_epc_class); +} +module_exit(pci_epc_exit); + +MODULE_DESCRIPTION("PCI EPC Library"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/pci/endpoint/pci-epc-mem.c b/drivers/pci/endpoint/pci-epc-mem.c new file mode 100644 index 000000000000..3a94cc1caf22 --- /dev/null +++ b/drivers/pci/endpoint/pci-epc-mem.c @@ -0,0 +1,143 @@ +/** + * PCI Endpoint *Controller* Address Space Management + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include + +#include + +/** + * pci_epc_mem_init() - initialize the pci_epc_mem structure + * @epc: the EPC device that invoked pci_epc_mem_init + * @phys_base: the physical address of the base + * @size: the size of the address space + * + * Invoke to initialize the pci_epc_mem structure used by the + * endpoint functions to allocate mapped PCI address. + */ +int pci_epc_mem_init(struct pci_epc *epc, phys_addr_t phys_base, size_t size) +{ + int ret; + struct pci_epc_mem *mem; + unsigned long *bitmap; + int pages = size >> PAGE_SHIFT; + int bitmap_size = BITS_TO_LONGS(pages) * sizeof(long); + + mem = kzalloc(sizeof(*mem), GFP_KERNEL); + if (!mem) { + ret = -ENOMEM; + goto err; + } + + bitmap = kzalloc(bitmap_size, GFP_KERNEL); + if (!bitmap) { + ret = -ENOMEM; + goto err_mem; + } + + mem->bitmap = bitmap; + mem->phys_base = phys_base; + mem->pages = pages; + mem->size = size; + + epc->mem = mem; + + return 0; + +err_mem: + kfree(mem); + +err: +return ret; +} +EXPORT_SYMBOL_GPL(pci_epc_mem_init); + +/** + * pci_epc_mem_exit() - cleanup the pci_epc_mem structure + * @epc: the EPC device that invoked pci_epc_mem_exit + * + * Invoke to cleanup the pci_epc_mem structure allocated in + * pci_epc_mem_init(). + */ +void pci_epc_mem_exit(struct pci_epc *epc) +{ + struct pci_epc_mem *mem = epc->mem; + + epc->mem = NULL; + kfree(mem->bitmap); + kfree(mem); +} +EXPORT_SYMBOL_GPL(pci_epc_mem_exit); + +/** + * pci_epc_mem_alloc_addr() - allocate memory address from EPC addr space + * @epc: the EPC device on which memory has to be allocated + * @phys_addr: populate the allocated physical address here + * @size: the size of the address space that has to be allocated + * + * Invoke to allocate memory address from the EPC address space. This + * is usually done to map the remote RC address into the local system. + */ +void __iomem *pci_epc_mem_alloc_addr(struct pci_epc *epc, + phys_addr_t *phys_addr, size_t size) +{ + int pageno; + void __iomem *virt_addr; + struct pci_epc_mem *mem = epc->mem; + int order = get_order(size); + + pageno = bitmap_find_free_region(mem->bitmap, mem->pages, order); + if (pageno < 0) + return NULL; + + *phys_addr = mem->phys_base + (pageno << PAGE_SHIFT); + virt_addr = ioremap(*phys_addr, size); + if (!virt_addr) + bitmap_release_region(mem->bitmap, pageno, order); + + return virt_addr; +} +EXPORT_SYMBOL_GPL(pci_epc_mem_alloc_addr); + +/** + * pci_epc_mem_free_addr() - free the allocated memory address + * @epc: the EPC device on which memory was allocated + * @phys_addr: the allocated physical address + * @virt_addr: virtual address of the allocated mem space + * @size: the size of the allocated address space + * + * Invoke to free the memory allocated using pci_epc_mem_alloc_addr. + */ +void pci_epc_mem_free_addr(struct pci_epc *epc, phys_addr_t phys_addr, + void __iomem *virt_addr, size_t size) +{ + int pageno; + int order = get_order(size); + struct pci_epc_mem *mem = epc->mem; + + iounmap(virt_addr); + pageno = (phys_addr - mem->phys_base) >> PAGE_SHIFT; + bitmap_release_region(mem->bitmap, pageno, order); +} +EXPORT_SYMBOL_GPL(pci_epc_mem_free_addr); + +MODULE_DESCRIPTION("PCI EPC Address Space Management"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/pci/endpoint/pci-epf-core.c b/drivers/pci/endpoint/pci-epf-core.c new file mode 100644 index 000000000000..a281c599a504 --- /dev/null +++ b/drivers/pci/endpoint/pci-epf-core.c @@ -0,0 +1,355 @@ +/** + * PCI Endpoint *Function* (EPF) library + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include + +#include +#include + +static struct bus_type pci_epf_bus_type; +static struct device_type pci_epf_type; + +/** + * pci_epf_linkup() - Notify the function driver that EPC device has + * established a connection with the Root Complex. + * @epf: the EPF device bound to the EPC device which has established + * the connection with the host + * + * Invoke to notify the function driver that EPC device has established + * a connection with the Root Complex. + */ +void pci_epf_linkup(struct pci_epf *epf) +{ + if (!epf->driver) { + dev_WARN(&epf->dev, "epf device not bound to driver\n"); + return; + } + + epf->driver->ops->linkup(epf); +} +EXPORT_SYMBOL_GPL(pci_epf_linkup); + +/** + * pci_epf_unbind() - Notify the function driver that the binding between the + * EPF device and EPC device has been lost + * @epf: the EPF device which has lost the binding with the EPC device + * + * Invoke to notify the function driver that the binding between the EPF device + * and EPC device has been lost. + */ +void pci_epf_unbind(struct pci_epf *epf) +{ + if (!epf->driver) { + dev_WARN(&epf->dev, "epf device not bound to driver\n"); + return; + } + + epf->driver->ops->unbind(epf); + module_put(epf->driver->owner); +} +EXPORT_SYMBOL_GPL(pci_epf_unbind); + +/** + * pci_epf_bind() - Notify the function driver that the EPF device has been + * bound to a EPC device + * @epf: the EPF device which has been bound to the EPC device + * + * Invoke to notify the function driver that it has been bound to a EPC device + */ +int pci_epf_bind(struct pci_epf *epf) +{ + if (!epf->driver) { + dev_WARN(&epf->dev, "epf device not bound to driver\n"); + return -EINVAL; + } + + if (!try_module_get(epf->driver->owner)) + return -EAGAIN; + + return epf->driver->ops->bind(epf); +} +EXPORT_SYMBOL_GPL(pci_epf_bind); + +/** + * pci_epf_free_space() - free the allocated PCI EPF register space + * @addr: the virtual address of the PCI EPF register space + * @bar: the BAR number corresponding to the register space + * + * Invoke to free the allocated PCI EPF register space. + */ +void pci_epf_free_space(struct pci_epf *epf, void *addr, enum pci_barno bar) +{ + struct device *dev = &epf->dev; + + if (!addr) + return; + + dma_free_coherent(dev, epf->bar[bar].size, addr, + epf->bar[bar].phys_addr); + + epf->bar[bar].phys_addr = 0; + epf->bar[bar].size = 0; +} +EXPORT_SYMBOL_GPL(pci_epf_free_space); + +/** + * pci_epf_alloc_space() - allocate memory for the PCI EPF register space + * @size: the size of the memory that has to be allocated + * @bar: the BAR number corresponding to the allocated register space + * + * Invoke to allocate memory for the PCI EPF register space. + */ +void *pci_epf_alloc_space(struct pci_epf *epf, size_t size, enum pci_barno bar) +{ + void *space; + struct device *dev = &epf->dev; + dma_addr_t phys_addr; + + if (size < 128) + size = 128; + size = roundup_pow_of_two(size); + + space = dma_alloc_coherent(dev, size, &phys_addr, GFP_KERNEL); + if (!space) { + dev_err(dev, "failed to allocate mem space\n"); + return NULL; + } + + epf->bar[bar].phys_addr = phys_addr; + epf->bar[bar].size = size; + + return space; +} +EXPORT_SYMBOL_GPL(pci_epf_alloc_space); + +/** + * pci_epf_unregister_driver() - unregister the PCI EPF driver + * @driver: the PCI EPF driver that has to be unregistered + * + * Invoke to unregister the PCI EPF driver. + */ +void pci_epf_unregister_driver(struct pci_epf_driver *driver) +{ + driver_unregister(&driver->driver); +} +EXPORT_SYMBOL_GPL(pci_epf_unregister_driver); + +/** + * __pci_epf_register_driver() - register a new PCI EPF driver + * @driver: structure representing PCI EPF driver + * @owner: the owner of the module that registers the PCI EPF driver + * + * Invoke to register a new PCI EPF driver. + */ +int __pci_epf_register_driver(struct pci_epf_driver *driver, + struct module *owner) +{ + int ret; + + if (!driver->ops) + return -EINVAL; + + if (!driver->ops->bind || !driver->ops->unbind || !driver->ops->linkup) + return -EINVAL; + + driver->driver.bus = &pci_epf_bus_type; + driver->driver.owner = owner; + + ret = driver_register(&driver->driver); + if (ret) + return ret; + + return 0; +} +EXPORT_SYMBOL_GPL(__pci_epf_register_driver); + +/** + * pci_epf_destroy() - destroy the created PCI EPF device + * @epf: the PCI EPF device that has to be destroyed. + * + * Invoke to destroy the PCI EPF device created by invoking pci_epf_create(). + */ +void pci_epf_destroy(struct pci_epf *epf) +{ + device_unregister(&epf->dev); +} +EXPORT_SYMBOL_GPL(pci_epf_destroy); + +/** + * pci_epf_create() - create a new PCI EPF device + * @name: the name of the PCI EPF device. This name will be used to bind the + * the EPF device to a EPF driver + * + * Invoke to create a new PCI EPF device by providing the name of the function + * device. + */ +struct pci_epf *pci_epf_create(const char *name) +{ + int ret; + struct pci_epf *epf; + struct device *dev; + char *func_name; + char *buf; + + epf = kzalloc(sizeof(*epf), GFP_KERNEL); + if (!epf) { + ret = -ENOMEM; + goto err_ret; + } + + buf = kstrdup(name, GFP_KERNEL); + if (!buf) { + ret = -ENOMEM; + goto free_epf; + } + + func_name = buf; + buf = strchrnul(buf, '.'); + *buf = '\0'; + + epf->name = kstrdup(func_name, GFP_KERNEL); + if (!epf->name) { + ret = -ENOMEM; + goto free_func_name; + } + + dev = &epf->dev; + device_initialize(dev); + dev->bus = &pci_epf_bus_type; + dev->type = &pci_epf_type; + + ret = dev_set_name(dev, "%s", name); + if (ret) + goto put_dev; + + ret = device_add(dev); + if (ret) + goto put_dev; + + kfree(func_name); + return epf; + +put_dev: + put_device(dev); + kfree(epf->name); + +free_func_name: + kfree(func_name); + +free_epf: + kfree(epf); + +err_ret: + return ERR_PTR(ret); +} +EXPORT_SYMBOL_GPL(pci_epf_create); + +static void pci_epf_dev_release(struct device *dev) +{ + struct pci_epf *epf = to_pci_epf(dev); + + kfree(epf->name); + kfree(epf); +} + +static struct device_type pci_epf_type = { + .release = pci_epf_dev_release, +}; + +static int +pci_epf_match_id(const struct pci_epf_device_id *id, const struct pci_epf *epf) +{ + while (id->name[0]) { + if (strcmp(epf->name, id->name) == 0) + return true; + id++; + } + + return false; +} + +static int pci_epf_device_match(struct device *dev, struct device_driver *drv) +{ + struct pci_epf *epf = to_pci_epf(dev); + struct pci_epf_driver *driver = to_pci_epf_driver(drv); + + if (driver->id_table) + return pci_epf_match_id(driver->id_table, epf); + + return !strcmp(epf->name, drv->name); +} + +static int pci_epf_device_probe(struct device *dev) +{ + struct pci_epf *epf = to_pci_epf(dev); + struct pci_epf_driver *driver = to_pci_epf_driver(dev->driver); + + if (!driver->probe) + return -ENODEV; + + epf->driver = driver; + + return driver->probe(epf); +} + +static int pci_epf_device_remove(struct device *dev) +{ + int ret; + struct pci_epf *epf = to_pci_epf(dev); + struct pci_epf_driver *driver = to_pci_epf_driver(dev->driver); + + ret = driver->remove(epf); + epf->driver = NULL; + + return ret; +} + +static struct bus_type pci_epf_bus_type = { + .name = "pci-epf", + .match = pci_epf_device_match, + .probe = pci_epf_device_probe, + .remove = pci_epf_device_remove, +}; + +static int __init pci_epf_init(void) +{ + int ret; + + ret = bus_register(&pci_epf_bus_type); + if (ret) { + pr_err("failed to register pci epf bus --> %d\n", ret); + return ret; + } + + return 0; +} +module_init(pci_epf_init); + +static void __exit pci_epf_exit(void) +{ + bus_unregister(&pci_epf_bus_type); +} +module_exit(pci_epf_exit); + +MODULE_DESCRIPTION("PCI EPF Library"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/mod_devicetable.h b/include/linux/mod_devicetable.h index 8850fcaf50db..566fda587fcf 100644 --- a/include/linux/mod_devicetable.h +++ b/include/linux/mod_devicetable.h @@ -428,6 +428,16 @@ struct i2c_device_id { kernel_ulong_t driver_data; /* Data private to the driver */ }; +/* pci_epf */ + +#define PCI_EPF_NAME_SIZE 20 +#define PCI_EPF_MODULE_PREFIX "pci_epf:" + +struct pci_epf_device_id { + char name[PCI_EPF_NAME_SIZE]; + kernel_ulong_t driver_data; +}; + /* spi */ #define SPI_NAME_SIZE 32 diff --git a/include/linux/pci-epc.h b/include/linux/pci-epc.h new file mode 100644 index 000000000000..8c63d3c37f76 --- /dev/null +++ b/include/linux/pci-epc.h @@ -0,0 +1,142 @@ +/** + * PCI Endpoint *Controller* (EPC) header file + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + */ + +#ifndef __LINUX_PCI_EPC_H +#define __LINUX_PCI_EPC_H + +#include + +struct pci_epc; + +enum pci_epc_irq_type { + PCI_EPC_IRQ_UNKNOWN, + PCI_EPC_IRQ_LEGACY, + PCI_EPC_IRQ_MSI, +}; + +/** + * struct pci_epc_ops - set of function pointers for performing EPC operations + * @write_header: ops to populate configuration space header + * @set_bar: ops to configure the BAR + * @clear_bar: ops to reset the BAR + * @map_addr: ops to map CPU address to PCI address + * @unmap_addr: ops to unmap CPU address and PCI address + * @set_msi: ops to set the requested number of MSI interrupts in the MSI + * capability register + * @get_msi: ops to get the number of MSI interrupts allocated by the RC from + * the MSI capability register + * @raise_irq: ops to raise a legacy or MSI interrupt + * @start: ops to start the PCI link + * @stop: ops to stop the PCI link + * @owner: the module owner containing the ops + */ +struct pci_epc_ops { + int (*write_header)(struct pci_epc *pci_epc, + struct pci_epf_header *hdr); + int (*set_bar)(struct pci_epc *epc, enum pci_barno bar, + dma_addr_t bar_phys, size_t size, int flags); + void (*clear_bar)(struct pci_epc *epc, enum pci_barno bar); + int (*map_addr)(struct pci_epc *epc, phys_addr_t addr, + u64 pci_addr, size_t size); + void (*unmap_addr)(struct pci_epc *epc, phys_addr_t addr); + int (*set_msi)(struct pci_epc *epc, u8 interrupts); + int (*get_msi)(struct pci_epc *epc); + int (*raise_irq)(struct pci_epc *pci_epc, + enum pci_epc_irq_type type, u8 interrupt_num); + int (*start)(struct pci_epc *epc); + void (*stop)(struct pci_epc *epc); + struct module *owner; +}; + +/** + * struct pci_epc_mem - address space of the endpoint controller + * @phys_base: physical base address of the PCI address space + * @size: the size of the PCI address space + * @bitmap: bitmap to manage the PCI address space + * @pages: number of bits representing the address region + */ +struct pci_epc_mem { + phys_addr_t phys_base; + size_t size; + unsigned long *bitmap; + int pages; +}; + +/** + * struct pci_epc - represents the PCI EPC device + * @dev: PCI EPC device + * @pci_epf: list of endpoint functions present in this EPC device + * @ops: function pointers for performing endpoint operations + * @mem: address space of the endpoint controller + * @max_functions: max number of functions that can be configured in this EPC + * @lock: spinlock to protect pci_epc ops + */ +struct pci_epc { + struct device dev; + struct list_head pci_epf; + const struct pci_epc_ops *ops; + struct pci_epc_mem *mem; + u8 max_functions; + /* spinlock to protect against concurrent access of EP controller */ + spinlock_t lock; +}; + +#define to_pci_epc(device) container_of((device), struct pci_epc, dev) + +#define pci_epc_create(dev, ops) \ + __pci_epc_create((dev), (ops), THIS_MODULE) +#define devm_pci_epc_create(dev, ops) \ + __devm_pci_epc_create((dev), (ops), THIS_MODULE) + +static inline void epc_set_drvdata(struct pci_epc *epc, void *data) +{ + dev_set_drvdata(&epc->dev, data); +} + +static inline void *epc_get_drvdata(struct pci_epc *epc) +{ + return dev_get_drvdata(&epc->dev); +} + +struct pci_epc * +__devm_pci_epc_create(struct device *dev, const struct pci_epc_ops *ops, + struct module *owner); +struct pci_epc * +__pci_epc_create(struct device *dev, const struct pci_epc_ops *ops, + struct module *owner); +void devm_pci_epc_destroy(struct device *dev, struct pci_epc *epc); +void pci_epc_destroy(struct pci_epc *epc); +int pci_epc_add_epf(struct pci_epc *epc, struct pci_epf *epf); +void pci_epc_linkup(struct pci_epc *epc); +void pci_epc_remove_epf(struct pci_epc *epc, struct pci_epf *epf); +int pci_epc_write_header(struct pci_epc *epc, struct pci_epf_header *hdr); +int pci_epc_set_bar(struct pci_epc *epc, enum pci_barno bar, + dma_addr_t bar_phys, size_t size, int flags); +void pci_epc_clear_bar(struct pci_epc *epc, int bar); +int pci_epc_map_addr(struct pci_epc *epc, phys_addr_t phys_addr, + u64 pci_addr, size_t size); +void pci_epc_unmap_addr(struct pci_epc *epc, phys_addr_t phys_addr); +int pci_epc_set_msi(struct pci_epc *epc, u8 interrupts); +int pci_epc_get_msi(struct pci_epc *epc); +int pci_epc_raise_irq(struct pci_epc *epc, enum pci_epc_irq_type type, + u8 interrupt_num); +int pci_epc_start(struct pci_epc *epc); +void pci_epc_stop(struct pci_epc *epc); +struct pci_epc *pci_epc_get(const char *epc_name); +void pci_epc_put(struct pci_epc *epc); + +int pci_epc_mem_init(struct pci_epc *epc, phys_addr_t phys_addr, size_t size); +void pci_epc_mem_exit(struct pci_epc *epc); +void __iomem *pci_epc_mem_alloc_addr(struct pci_epc *epc, + phys_addr_t *phys_addr, size_t size); +void pci_epc_mem_free_addr(struct pci_epc *epc, phys_addr_t phys_addr, + void __iomem *virt_addr, size_t size); +#endif /* __LINUX_PCI_EPC_H */ diff --git a/include/linux/pci-epf.h b/include/linux/pci-epf.h new file mode 100644 index 000000000000..5628714f7bcf --- /dev/null +++ b/include/linux/pci-epf.h @@ -0,0 +1,160 @@ +/** + * PCI Endpoint *Function* (EPF) header file + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + */ + +#ifndef __LINUX_PCI_EPF_H +#define __LINUX_PCI_EPF_H + +#include +#include + +struct pci_epf; + +enum pci_interrupt_pin { + PCI_INTERRUPT_UNKNOWN, + PCI_INTERRUPT_INTA, + PCI_INTERRUPT_INTB, + PCI_INTERRUPT_INTC, + PCI_INTERRUPT_INTD, +}; + +enum pci_barno { + BAR_0, + BAR_1, + BAR_2, + BAR_3, + BAR_4, + BAR_5, +}; + +/** + * struct pci_epf_header - represents standard configuration header + * @vendorid: identifies device manufacturer + * @deviceid: identifies a particular device + * @revid: specifies a device-specific revision identifier + * @progif_code: identifies a specific register-level programming interface + * @subclass_code: identifies more specifically the function of the device + * @baseclass_code: broadly classifies the type of function the device performs + * @cache_line_size: specifies the system cacheline size in units of DWORDs + * @subsys_vendor_id: vendor of the add-in card or subsystem + * @subsys_id: id specific to vendor + * @interrupt_pin: interrupt pin the device (or device function) uses + */ +struct pci_epf_header { + u16 vendorid; + u16 deviceid; + u8 revid; + u8 progif_code; + u8 subclass_code; + u8 baseclass_code; + u8 cache_line_size; + u16 subsys_vendor_id; + u16 subsys_id; + enum pci_interrupt_pin interrupt_pin; +}; + +/** + * struct pci_epf_ops - set of function pointers for performing EPF operations + * @bind: ops to perform when a EPC device has been bound to EPF device + * @unbind: ops to perform when a binding has been lost between a EPC device + * and EPF device + * @linkup: ops to perform when the EPC device has established a connection with + * a host system + */ +struct pci_epf_ops { + int (*bind)(struct pci_epf *epf); + void (*unbind)(struct pci_epf *epf); + void (*linkup)(struct pci_epf *epf); +}; + +/** + * struct pci_epf_driver - represents the PCI EPF driver + * @probe: ops to perform when a new EPF device has been bound to the EPF driver + * @remove: ops to perform when the binding between the EPF device and EPF + * driver is broken + * @driver: PCI EPF driver + * @ops: set of function pointers for performing EPF operations + * @owner: the owner of the module that registers the PCI EPF driver + * @id_table: identifies EPF devices for probing + */ +struct pci_epf_driver { + int (*probe)(struct pci_epf *epf); + int (*remove)(struct pci_epf *epf); + + struct device_driver driver; + struct pci_epf_ops *ops; + struct module *owner; + const struct pci_epf_device_id *id_table; +}; + +#define to_pci_epf_driver(drv) (container_of((drv), struct pci_epf_driver, \ + driver)) + +/** + * struct pci_epf_bar - represents the BAR of EPF device + * @phys_addr: physical address that should be mapped to the BAR + * @size: the size of the address space present in BAR + */ +struct pci_epf_bar { + dma_addr_t phys_addr; + size_t size; +}; + +/** + * struct pci_epf - represents the PCI EPF device + * @dev: the PCI EPF device + * @name: the name of the PCI EPF device + * @header: represents standard configuration header + * @bar: represents the BAR of EPF device + * @msi_interrupts: number of MSI interrupts required by this function + * @func_no: unique function number within this endpoint device + * @epc: the EPC device to which this EPF device is bound + * @driver: the EPF driver to which this EPF device is bound + * @list: to add pci_epf as a list of PCI endpoint functions to pci_epc + */ +struct pci_epf { + struct device dev; + const char *name; + struct pci_epf_header *header; + struct pci_epf_bar bar[6]; + u8 msi_interrupts; + u8 func_no; + + struct pci_epc *epc; + struct pci_epf_driver *driver; + struct list_head list; +}; + +#define to_pci_epf(epf_dev) container_of((epf_dev), struct pci_epf, dev) + +#define pci_epf_register_driver(driver) \ + __pci_epf_register_driver((driver), THIS_MODULE) + +static inline void epf_set_drvdata(struct pci_epf *epf, void *data) +{ + dev_set_drvdata(&epf->dev, data); +} + +static inline void *epf_get_drvdata(struct pci_epf *epf) +{ + return dev_get_drvdata(&epf->dev); +} + +struct pci_epf *pci_epf_create(const char *name); +void pci_epf_destroy(struct pci_epf *epf); +int __pci_epf_register_driver(struct pci_epf_driver *driver, + struct module *owner); +void pci_epf_unregister_driver(struct pci_epf_driver *driver); +void *pci_epf_alloc_space(struct pci_epf *epf, size_t size, enum pci_barno bar); +void pci_epf_free_space(struct pci_epf *epf, void *addr, enum pci_barno bar); +int pci_epf_bind(struct pci_epf *epf); +void pci_epf_unbind(struct pci_epf *epf); +void pci_epf_linkup(struct pci_epf *epf); +#endif /* __LINUX_PCI_EPF_H */ -- cgit v1.2.3-59-g8ed1b From a8a5be098885e8394497a6b64fb3f96dd217846b Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:14:58 +0530 Subject: Documentation: PCI: Guide to use PCI Endpoint Core Layer Add Documentation to help users use endpoint library to enable endpoint mode in the PCI controller and add new PCI endpoint functions. Signed-off-by: Kishon Vijay Abraham I Acked-By: Joao Pinto Signed-off-by: Bjorn Helgaas --- Documentation/PCI/00-INDEX | 2 + Documentation/PCI/endpoint/pci-endpoint.txt | 215 ++++++++++++++++++++++++++++ 2 files changed, 217 insertions(+) create mode 100644 Documentation/PCI/endpoint/pci-endpoint.txt diff --git a/Documentation/PCI/00-INDEX b/Documentation/PCI/00-INDEX index 147231f1613e..ba950b296bd8 100644 --- a/Documentation/PCI/00-INDEX +++ b/Documentation/PCI/00-INDEX @@ -12,3 +12,5 @@ pci.txt - info on the PCI subsystem for device driver authors pcieaer-howto.txt - the PCI Express Advanced Error Reporting Driver Guide HOWTO +endpoint/pci-endpoint.txt + - guide to add endpoint controller driver and endpoint function driver. diff --git a/Documentation/PCI/endpoint/pci-endpoint.txt b/Documentation/PCI/endpoint/pci-endpoint.txt new file mode 100644 index 000000000000..9b1d66829290 --- /dev/null +++ b/Documentation/PCI/endpoint/pci-endpoint.txt @@ -0,0 +1,215 @@ + PCI ENDPOINT FRAMEWORK + Kishon Vijay Abraham I + +This document is a guide to use the PCI Endpoint Framework in order to create +endpoint controller driver, endpoint function driver, and using configfs +interface to bind the function driver to the controller driver. + +1. Introduction + +Linux has a comprehensive PCI subsystem to support PCI controllers that +operates in Root Complex mode. The subsystem has capability to scan PCI bus, +assign memory resources and IRQ resources, load PCI driver (based on +vendor ID, device ID), support other services like hot-plug, power management, +advanced error reporting and virtual channels. + +However the PCI controller IP integrated in some SoCs is capable of operating +either in Root Complex mode or Endpoint mode. PCI Endpoint Framework will +add endpoint mode support in Linux. This will help to run Linux in an +EP system which can have a wide variety of use cases from testing or +validation, co-processor accelerator, etc. + +2. PCI Endpoint Core + +The PCI Endpoint Core layer comprises 3 components: the Endpoint Controller +library, the Endpoint Function library, and the configfs layer to bind the +endpoint function with the endpoint controller. + +2.1 PCI Endpoint Controller(EPC) Library + +The EPC library provides APIs to be used by the controller that can operate +in endpoint mode. It also provides APIs to be used by function driver/library +in order to implement a particular endpoint function. + +2.1.1 APIs for the PCI controller Driver + +This section lists the APIs that the PCI Endpoint core provides to be used +by the PCI controller driver. + +*) devm_pci_epc_create()/pci_epc_create() + + The PCI controller driver should implement the following ops: + * write_header: ops to populate configuration space header + * set_bar: ops to configure the BAR + * clear_bar: ops to reset the BAR + * alloc_addr_space: ops to allocate in PCI controller address space + * free_addr_space: ops to free the allocated address space + * raise_irq: ops to raise a legacy or MSI interrupt + * start: ops to start the PCI link + * stop: ops to stop the PCI link + + The PCI controller driver can then create a new EPC device by invoking + devm_pci_epc_create()/pci_epc_create(). + +*) devm_pci_epc_destroy()/pci_epc_destroy() + + The PCI controller driver can destroy the EPC device created by either + devm_pci_epc_create() or pci_epc_create() using devm_pci_epc_destroy() or + pci_epc_destroy(). + +*) pci_epc_linkup() + + In order to notify all the function devices that the EPC device to which + they are linked has established a link with the host, the PCI controller + driver should invoke pci_epc_linkup(). + +*) pci_epc_mem_init() + + Initialize the pci_epc_mem structure used for allocating EPC addr space. + +*) pci_epc_mem_exit() + + Cleanup the pci_epc_mem structure allocated during pci_epc_mem_init(). + +2.1.2 APIs for the PCI Endpoint Function Driver + +This section lists the APIs that the PCI Endpoint core provides to be used +by the PCI endpoint function driver. + +*) pci_epc_write_header() + + The PCI endpoint function driver should use pci_epc_write_header() to + write the standard configuration header to the endpoint controller. + +*) pci_epc_set_bar() + + The PCI endpoint function driver should use pci_epc_set_bar() to configure + the Base Address Register in order for the host to assign PCI addr space. + Register space of the function driver is usually configured + using this API. + +*) pci_epc_clear_bar() + + The PCI endpoint function driver should use pci_epc_clear_bar() to reset + the BAR. + +*) pci_epc_raise_irq() + + The PCI endpoint function driver should use pci_epc_raise_irq() to raise + Legacy Interrupt or MSI Interrupt. + +*) pci_epc_mem_alloc_addr() + + The PCI endpoint function driver should use pci_epc_mem_alloc_addr(), to + allocate memory address from EPC addr space which is required to access + RC's buffer + +*) pci_epc_mem_free_addr() + + The PCI endpoint function driver should use pci_epc_mem_free_addr() to + free the memory space allocated using pci_epc_mem_alloc_addr(). + +2.1.3 Other APIs + +There are other APIs provided by the EPC library. These are used for binding +the EPF device with EPC device. pci-ep-cfs.c can be used as reference for +using these APIs. + +*) pci_epc_get() + + Get a reference to the PCI endpoint controller based on the device name of + the controller. + +*) pci_epc_put() + + Release the reference to the PCI endpoint controller obtained using + pci_epc_get() + +*) pci_epc_add_epf() + + Add a PCI endpoint function to a PCI endpoint controller. A PCIe device + can have up to 8 functions according to the specification. + +*) pci_epc_remove_epf() + + Remove the PCI endpoint function from PCI endpoint controller. + +*) pci_epc_start() + + The PCI endpoint function driver should invoke pci_epc_start() once it + has configured the endpoint function and wants to start the PCI link. + +*) pci_epc_stop() + + The PCI endpoint function driver should invoke pci_epc_stop() to stop + the PCI LINK. + +2.2 PCI Endpoint Function(EPF) Library + +The EPF library provides APIs to be used by the function driver and the EPC +library to provide endpoint mode functionality. + +2.2.1 APIs for the PCI Endpoint Function Driver + +This section lists the APIs that the PCI Endpoint core provides to be used +by the PCI endpoint function driver. + +*) pci_epf_register_driver() + + The PCI Endpoint Function driver should implement the following ops: + * bind: ops to perform when a EPC device has been bound to EPF device + * unbind: ops to perform when a binding has been lost between a EPC + device and EPF device + * linkup: ops to perform when the EPC device has established a + connection with a host system + + The PCI Function driver can then register the PCI EPF driver by using + pci_epf_register_driver(). + +*) pci_epf_unregister_driver() + + The PCI Function driver can unregister the PCI EPF driver by using + pci_epf_unregister_driver(). + +*) pci_epf_alloc_space() + + The PCI Function driver can allocate space for a particular BAR using + pci_epf_alloc_space(). + +*) pci_epf_free_space() + + The PCI Function driver can free the allocated space + (using pci_epf_alloc_space) by invoking pci_epf_free_space(). + +2.2.2 APIs for the PCI Endpoint Controller Library +This section lists the APIs that the PCI Endpoint core provides to be used +by the PCI endpoint controller library. + +*) pci_epf_linkup() + + The PCI endpoint controller library invokes pci_epf_linkup() when the + EPC device has established the connection to the host. + +2.2.2 Other APIs +There are other APIs provided by the EPF library. These are used to notify +the function driver when the EPF device is bound to the EPC device. +pci-ep-cfs.c can be used as reference for using these APIs. + +*) pci_epf_create() + + Create a new PCI EPF device by passing the name of the PCI EPF device. + This name will be used to bind the the EPF device to a EPF driver. + +*) pci_epf_destroy() + + Destroy the created PCI EPF device. + +*) pci_epf_bind() + + pci_epf_bind() should be invoked when the EPF device has been bound to + a EPC device. + +*) pci_epf_unbind() + + pci_epf_unbind() should be invoked when the binding between EPC device + and EPF device is lost. -- cgit v1.2.3-59-g8ed1b From d746799116103d857be203382b09035bbe225d03 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:14:59 +0530 Subject: PCI: endpoint: Introduce configfs entry for configuring EP functions Introduce a new configfs entry to configure the EP function (like configuring the standard configuration header entries) and to bind the EP function with EP controller. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/endpoint/Kconfig | 9 + drivers/pci/endpoint/Makefile | 1 + drivers/pci/endpoint/pci-ep-cfs.c | 509 ++++++++++++++++++++++++++++++++++++++ include/linux/pci-ep-cfs.h | 41 +++ 4 files changed, 560 insertions(+) create mode 100644 drivers/pci/endpoint/pci-ep-cfs.c create mode 100644 include/linux/pci-ep-cfs.h diff --git a/drivers/pci/endpoint/Kconfig b/drivers/pci/endpoint/Kconfig index a5442ace7077..c86bca9b7de3 100644 --- a/drivers/pci/endpoint/Kconfig +++ b/drivers/pci/endpoint/Kconfig @@ -17,4 +17,13 @@ config PCI_ENDPOINT If in doubt, say "N" to disable Endpoint support. +config PCI_ENDPOINT_CONFIGFS + bool "PCI Endpoint Configfs Support" + depends on PCI_ENDPOINT + select CONFIGFS_FS + help + This will enable the configfs entry that can be used to + configure the endpoint function and used to bind the + function with a endpoint controller. + endmenu diff --git a/drivers/pci/endpoint/Makefile b/drivers/pci/endpoint/Makefile index dc1bc16491e6..7219d51bb401 100644 --- a/drivers/pci/endpoint/Makefile +++ b/drivers/pci/endpoint/Makefile @@ -2,5 +2,6 @@ # Makefile for PCI Endpoint Support # +obj-$(CONFIG_PCI_ENDPOINT_CONFIGFS) += pci-ep-cfs.o obj-$(CONFIG_PCI_ENDPOINT) += pci-epc-core.o pci-epf-core.o\ pci-epc-mem.o diff --git a/drivers/pci/endpoint/pci-ep-cfs.c b/drivers/pci/endpoint/pci-ep-cfs.c new file mode 100644 index 000000000000..424fdd6ed1ca --- /dev/null +++ b/drivers/pci/endpoint/pci-ep-cfs.c @@ -0,0 +1,509 @@ +/** + * configfs to configure the PCI endpoint + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include + +#include +#include +#include + +static struct config_group *functions_group; +static struct config_group *controllers_group; + +struct pci_epf_group { + struct config_group group; + struct pci_epf *epf; +}; + +struct pci_epc_group { + struct config_group group; + struct pci_epc *epc; + bool start; + unsigned long function_num_map; +}; + +static inline struct pci_epf_group *to_pci_epf_group(struct config_item *item) +{ + return container_of(to_config_group(item), struct pci_epf_group, group); +} + +static inline struct pci_epc_group *to_pci_epc_group(struct config_item *item) +{ + return container_of(to_config_group(item), struct pci_epc_group, group); +} + +static ssize_t pci_epc_start_store(struct config_item *item, const char *page, + size_t len) +{ + int ret; + bool start; + struct pci_epc *epc; + struct pci_epc_group *epc_group = to_pci_epc_group(item); + + epc = epc_group->epc; + + ret = kstrtobool(page, &start); + if (ret) + return ret; + + if (!start) { + pci_epc_stop(epc); + return len; + } + + ret = pci_epc_start(epc); + if (ret) { + dev_err(&epc->dev, "failed to start endpoint controller\n"); + return -EINVAL; + } + + epc_group->start = start; + + return len; +} + +static ssize_t pci_epc_start_show(struct config_item *item, char *page) +{ + return sprintf(page, "%d\n", + to_pci_epc_group(item)->start); +} + +CONFIGFS_ATTR(pci_epc_, start); + +static struct configfs_attribute *pci_epc_attrs[] = { + &pci_epc_attr_start, + NULL, +}; + +static int pci_epc_epf_link(struct config_item *epc_item, + struct config_item *epf_item) +{ + int ret; + u32 func_no = 0; + struct pci_epc *epc; + struct pci_epf *epf; + struct pci_epf_group *epf_group = to_pci_epf_group(epf_item); + struct pci_epc_group *epc_group = to_pci_epc_group(epc_item); + + epc = epc_group->epc; + epf = epf_group->epf; + ret = pci_epc_add_epf(epc, epf); + if (ret) + goto err_add_epf; + + func_no = find_first_zero_bit(&epc_group->function_num_map, + sizeof(epc_group->function_num_map)); + set_bit(func_no, &epc_group->function_num_map); + epf->func_no = func_no; + + ret = pci_epf_bind(epf); + if (ret) + goto err_epf_bind; + + return 0; + +err_epf_bind: + pci_epc_remove_epf(epc, epf); + +err_add_epf: + clear_bit(func_no, &epc_group->function_num_map); + + return ret; +} + +static void pci_epc_epf_unlink(struct config_item *epc_item, + struct config_item *epf_item) +{ + struct pci_epc *epc; + struct pci_epf *epf; + struct pci_epf_group *epf_group = to_pci_epf_group(epf_item); + struct pci_epc_group *epc_group = to_pci_epc_group(epc_item); + + WARN_ON_ONCE(epc_group->start); + + epc = epc_group->epc; + epf = epf_group->epf; + clear_bit(epf->func_no, &epc_group->function_num_map); + pci_epf_unbind(epf); + pci_epc_remove_epf(epc, epf); +} + +static struct configfs_item_operations pci_epc_item_ops = { + .allow_link = pci_epc_epf_link, + .drop_link = pci_epc_epf_unlink, +}; + +static struct config_item_type pci_epc_type = { + .ct_item_ops = &pci_epc_item_ops, + .ct_attrs = pci_epc_attrs, + .ct_owner = THIS_MODULE, +}; + +struct config_group *pci_ep_cfs_add_epc_group(const char *name) +{ + int ret; + struct pci_epc *epc; + struct config_group *group; + struct pci_epc_group *epc_group; + + epc_group = kzalloc(sizeof(*epc_group), GFP_KERNEL); + if (!epc_group) { + ret = -ENOMEM; + goto err; + } + + group = &epc_group->group; + + config_group_init_type_name(group, name, &pci_epc_type); + ret = configfs_register_group(controllers_group, group); + if (ret) { + pr_err("failed to register configfs group for %s\n", name); + goto err_register_group; + } + + epc = pci_epc_get(name); + if (IS_ERR(epc)) { + ret = PTR_ERR(epc); + goto err_epc_get; + } + + epc_group->epc = epc; + + return group; + +err_epc_get: + configfs_unregister_group(group); + +err_register_group: + kfree(epc_group); + +err: + return ERR_PTR(ret); +} +EXPORT_SYMBOL(pci_ep_cfs_add_epc_group); + +void pci_ep_cfs_remove_epc_group(struct config_group *group) +{ + struct pci_epc_group *epc_group; + + if (!group) + return; + + epc_group = container_of(group, struct pci_epc_group, group); + pci_epc_put(epc_group->epc); + configfs_unregister_group(&epc_group->group); + kfree(epc_group); +} +EXPORT_SYMBOL(pci_ep_cfs_remove_epc_group); + +#define PCI_EPF_HEADER_R(_name) \ +static ssize_t pci_epf_##_name##_show(struct config_item *item, char *page) \ +{ \ + struct pci_epf *epf = to_pci_epf_group(item)->epf; \ + if (WARN_ON_ONCE(!epf->header)) \ + return -EINVAL; \ + return sprintf(page, "0x%04x\n", epf->header->_name); \ +} + +#define PCI_EPF_HEADER_W_u32(_name) \ +static ssize_t pci_epf_##_name##_store(struct config_item *item, \ + const char *page, size_t len) \ +{ \ + u32 val; \ + int ret; \ + struct pci_epf *epf = to_pci_epf_group(item)->epf; \ + if (WARN_ON_ONCE(!epf->header)) \ + return -EINVAL; \ + ret = kstrtou32(page, 0, &val); \ + if (ret) \ + return ret; \ + epf->header->_name = val; \ + return len; \ +} + +#define PCI_EPF_HEADER_W_u16(_name) \ +static ssize_t pci_epf_##_name##_store(struct config_item *item, \ + const char *page, size_t len) \ +{ \ + u16 val; \ + int ret; \ + struct pci_epf *epf = to_pci_epf_group(item)->epf; \ + if (WARN_ON_ONCE(!epf->header)) \ + return -EINVAL; \ + ret = kstrtou16(page, 0, &val); \ + if (ret) \ + return ret; \ + epf->header->_name = val; \ + return len; \ +} + +#define PCI_EPF_HEADER_W_u8(_name) \ +static ssize_t pci_epf_##_name##_store(struct config_item *item, \ + const char *page, size_t len) \ +{ \ + u8 val; \ + int ret; \ + struct pci_epf *epf = to_pci_epf_group(item)->epf; \ + if (WARN_ON_ONCE(!epf->header)) \ + return -EINVAL; \ + ret = kstrtou8(page, 0, &val); \ + if (ret) \ + return ret; \ + epf->header->_name = val; \ + return len; \ +} + +static ssize_t pci_epf_msi_interrupts_store(struct config_item *item, + const char *page, size_t len) +{ + u8 val; + int ret; + + ret = kstrtou8(page, 0, &val); + if (ret) + return ret; + + to_pci_epf_group(item)->epf->msi_interrupts = val; + + return len; +} + +static ssize_t pci_epf_msi_interrupts_show(struct config_item *item, + char *page) +{ + return sprintf(page, "%d\n", + to_pci_epf_group(item)->epf->msi_interrupts); +} + +PCI_EPF_HEADER_R(vendorid) +PCI_EPF_HEADER_W_u16(vendorid) + +PCI_EPF_HEADER_R(deviceid) +PCI_EPF_HEADER_W_u16(deviceid) + +PCI_EPF_HEADER_R(revid) +PCI_EPF_HEADER_W_u8(revid) + +PCI_EPF_HEADER_R(progif_code) +PCI_EPF_HEADER_W_u8(progif_code) + +PCI_EPF_HEADER_R(subclass_code) +PCI_EPF_HEADER_W_u8(subclass_code) + +PCI_EPF_HEADER_R(baseclass_code) +PCI_EPF_HEADER_W_u8(baseclass_code) + +PCI_EPF_HEADER_R(cache_line_size) +PCI_EPF_HEADER_W_u8(cache_line_size) + +PCI_EPF_HEADER_R(subsys_vendor_id) +PCI_EPF_HEADER_W_u16(subsys_vendor_id) + +PCI_EPF_HEADER_R(subsys_id) +PCI_EPF_HEADER_W_u16(subsys_id) + +PCI_EPF_HEADER_R(interrupt_pin) +PCI_EPF_HEADER_W_u8(interrupt_pin) + +CONFIGFS_ATTR(pci_epf_, vendorid); +CONFIGFS_ATTR(pci_epf_, deviceid); +CONFIGFS_ATTR(pci_epf_, revid); +CONFIGFS_ATTR(pci_epf_, progif_code); +CONFIGFS_ATTR(pci_epf_, subclass_code); +CONFIGFS_ATTR(pci_epf_, baseclass_code); +CONFIGFS_ATTR(pci_epf_, cache_line_size); +CONFIGFS_ATTR(pci_epf_, subsys_vendor_id); +CONFIGFS_ATTR(pci_epf_, subsys_id); +CONFIGFS_ATTR(pci_epf_, interrupt_pin); +CONFIGFS_ATTR(pci_epf_, msi_interrupts); + +static struct configfs_attribute *pci_epf_attrs[] = { + &pci_epf_attr_vendorid, + &pci_epf_attr_deviceid, + &pci_epf_attr_revid, + &pci_epf_attr_progif_code, + &pci_epf_attr_subclass_code, + &pci_epf_attr_baseclass_code, + &pci_epf_attr_cache_line_size, + &pci_epf_attr_subsys_vendor_id, + &pci_epf_attr_subsys_id, + &pci_epf_attr_interrupt_pin, + &pci_epf_attr_msi_interrupts, + NULL, +}; + +static void pci_epf_release(struct config_item *item) +{ + struct pci_epf_group *epf_group = to_pci_epf_group(item); + + pci_epf_destroy(epf_group->epf); + kfree(epf_group); +} + +static struct configfs_item_operations pci_epf_ops = { + .release = pci_epf_release, +}; + +static struct config_item_type pci_epf_type = { + .ct_item_ops = &pci_epf_ops, + .ct_attrs = pci_epf_attrs, + .ct_owner = THIS_MODULE, +}; + +static struct config_group *pci_epf_make(struct config_group *group, + const char *name) +{ + struct pci_epf_group *epf_group; + struct pci_epf *epf; + + epf_group = kzalloc(sizeof(*epf_group), GFP_KERNEL); + if (!epf_group) + return ERR_PTR(-ENOMEM); + + config_group_init_type_name(&epf_group->group, name, &pci_epf_type); + + epf = pci_epf_create(group->cg_item.ci_name); + if (IS_ERR(epf)) { + pr_err("failed to create endpoint function device\n"); + return ERR_PTR(-EINVAL); + } + + epf_group->epf = epf; + + return &epf_group->group; +} + +static void pci_epf_drop(struct config_group *group, struct config_item *item) +{ + config_item_put(item); +} + +static struct configfs_group_operations pci_epf_group_ops = { + .make_group = &pci_epf_make, + .drop_item = &pci_epf_drop, +}; + +static struct config_item_type pci_epf_group_type = { + .ct_group_ops = &pci_epf_group_ops, + .ct_owner = THIS_MODULE, +}; + +struct config_group *pci_ep_cfs_add_epf_group(const char *name) +{ + struct config_group *group; + + group = configfs_register_default_group(functions_group, name, + &pci_epf_group_type); + if (IS_ERR(group)) + pr_err("failed to register configfs group for %s function\n", + name); + + return group; +} +EXPORT_SYMBOL(pci_ep_cfs_add_epf_group); + +void pci_ep_cfs_remove_epf_group(struct config_group *group) +{ + if (IS_ERR_OR_NULL(group)) + return; + + configfs_unregister_default_group(group); +} +EXPORT_SYMBOL(pci_ep_cfs_remove_epf_group); + +static struct config_item_type pci_functions_type = { + .ct_owner = THIS_MODULE, +}; + +static struct config_item_type pci_controllers_type = { + .ct_owner = THIS_MODULE, +}; + +static struct config_item_type pci_ep_type = { + .ct_owner = THIS_MODULE, +}; + +static struct configfs_subsystem pci_ep_cfs_subsys = { + .su_group = { + .cg_item = { + .ci_namebuf = "pci_ep", + .ci_type = &pci_ep_type, + }, + }, + .su_mutex = __MUTEX_INITIALIZER(pci_ep_cfs_subsys.su_mutex), +}; + +static int __init pci_ep_cfs_init(void) +{ + int ret; + struct config_group *root = &pci_ep_cfs_subsys.su_group; + + config_group_init(root); + + ret = configfs_register_subsystem(&pci_ep_cfs_subsys); + if (ret) { + pr_err("Error %d while registering subsystem %s\n", + ret, root->cg_item.ci_namebuf); + goto err; + } + + functions_group = configfs_register_default_group(root, "functions", + &pci_functions_type); + if (IS_ERR(functions_group)) { + ret = PTR_ERR(functions_group); + pr_err("Error %d while registering functions group\n", + ret); + goto err_functions_group; + } + + controllers_group = + configfs_register_default_group(root, "controllers", + &pci_controllers_type); + if (IS_ERR(controllers_group)) { + ret = PTR_ERR(controllers_group); + pr_err("Error %d while registering controllers group\n", + ret); + goto err_controllers_group; + } + + return 0; + +err_controllers_group: + configfs_unregister_default_group(functions_group); + +err_functions_group: + configfs_unregister_subsystem(&pci_ep_cfs_subsys); + +err: + return ret; +} +module_init(pci_ep_cfs_init); + +static void __exit pci_ep_cfs_exit(void) +{ + configfs_unregister_default_group(controllers_group); + configfs_unregister_default_group(functions_group); + configfs_unregister_subsystem(&pci_ep_cfs_subsys); +} +module_exit(pci_ep_cfs_exit); + +MODULE_DESCRIPTION("PCI EP CONFIGFS"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); diff --git a/include/linux/pci-ep-cfs.h b/include/linux/pci-ep-cfs.h new file mode 100644 index 000000000000..263b89ea5705 --- /dev/null +++ b/include/linux/pci-ep-cfs.h @@ -0,0 +1,41 @@ +/** + * PCI Endpoint ConfigFS header file + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + */ + +#ifndef __LINUX_PCI_EP_CFS_H +#define __LINUX_PCI_EP_CFS_H + +#include + +#ifdef CONFIG_PCI_ENDPOINT_CONFIGFS +struct config_group *pci_ep_cfs_add_epc_group(const char *name); +void pci_ep_cfs_remove_epc_group(struct config_group *group); +struct config_group *pci_ep_cfs_add_epf_group(const char *name); +void pci_ep_cfs_remove_epf_group(struct config_group *group); +#else +static inline struct config_group *pci_ep_cfs_add_epc_group(const char *name) +{ + return 0; +} + +static inline void pci_ep_cfs_remove_epc_group(struct config_group *group) +{ +} + +static inline struct config_group *pci_ep_cfs_add_epf_group(const char *name) +{ + return 0; +} + +static inline void pci_ep_cfs_remove_epf_group(struct config_group *group) +{ +} +#endif +#endif /* __LINUX_PCI_EP_CFS_H */ -- cgit v1.2.3-59-g8ed1b From bea37d3d3121d051b12541f06327d7ac63996fc0 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:00 +0530 Subject: Documentation: PCI: Guide to use PCI endpoint configfs Add Documentation to help users use PCI endpoint to configure PCI endpoint function and to bind the endpoint function with endpoint controller. Signed-off-by: Kishon Vijay Abraham I Acked-By: Joao Pinto Signed-off-by: Bjorn Helgaas --- Documentation/PCI/00-INDEX | 2 + Documentation/PCI/endpoint/pci-endpoint-cfs.txt | 105 ++++++++++++++++++++++++ 2 files changed, 107 insertions(+) create mode 100644 Documentation/PCI/endpoint/pci-endpoint-cfs.txt diff --git a/Documentation/PCI/00-INDEX b/Documentation/PCI/00-INDEX index ba950b296bd8..bf8223a1bf08 100644 --- a/Documentation/PCI/00-INDEX +++ b/Documentation/PCI/00-INDEX @@ -14,3 +14,5 @@ pcieaer-howto.txt - the PCI Express Advanced Error Reporting Driver Guide HOWTO endpoint/pci-endpoint.txt - guide to add endpoint controller driver and endpoint function driver. +endpoint/pci-endpoint-cfs.txt + - guide to use configfs to configure the PCI endpoint function. diff --git a/Documentation/PCI/endpoint/pci-endpoint-cfs.txt b/Documentation/PCI/endpoint/pci-endpoint-cfs.txt new file mode 100644 index 000000000000..d740f29960a4 --- /dev/null +++ b/Documentation/PCI/endpoint/pci-endpoint-cfs.txt @@ -0,0 +1,105 @@ + CONFIGURING PCI ENDPOINT USING CONFIGFS + Kishon Vijay Abraham I + +The PCI Endpoint Core exposes configfs entry (pci_ep) to configure the +PCI endpoint function and to bind the endpoint function +with the endpoint controller. (For introducing other mechanisms to +configure the PCI Endpoint Function refer to [1]). + +*) Mounting configfs + +The PCI Endpoint Core layer creates pci_ep directory in the mounted configfs +directory. configfs can be mounted using the following command. + + mount -t configfs none /sys/kernel/config + +*) Directory Structure + +The pci_ep configfs has two directories at its root: controllers and +functions. Every EPC device present in the system will have an entry in +the *controllers* directory and and every EPF driver present in the system +will have an entry in the *functions* directory. + +/sys/kernel/config/pci_ep/ + .. controllers/ + .. functions/ + +*) Creating EPF Device + +Every registered EPF driver will be listed in controllers directory. The +entries corresponding to EPF driver will be created by the EPF core. + +/sys/kernel/config/pci_ep/functions/ + .. / + ... / + ... / + .. / + ... / + ... / + +In order to create a of the type probed by , the +user has to create a directory inside . + +Every directory consists of the following entries that can be +used to configure the standard configuration header of the endpoint function. +(These entries are created by the framework when any new is +created) + + .. / + ... / + ... vendorid + ... deviceid + ... revid + ... progif_code + ... subclass_code + ... baseclass_code + ... cache_line_size + ... subsys_vendor_id + ... subsys_id + ... interrupt_pin + +*) EPC Device + +Every registered EPC device will be listed in controllers directory. The +entries corresponding to EPC device will be created by the EPC core. + +/sys/kernel/config/pci_ep/controllers/ + .. / + ... / + ... / + ... start + .. / + ... / + ... / + ... start + +The directory will have a list of symbolic links to +. These symbolic links should be created by the user to +represent the functions present in the endpoint device. + +The directory will also have a *start* field. Once +"1" is written to this field, the endpoint device will be ready to +establish the link with the host. This is usually done after +all the EPF devices are created and linked with the EPC device. + + + | controllers/ + | / + | + | start + | functions/ + | / + | / + | vendorid + | deviceid + | revid + | progif_code + | subclass_code + | baseclass_code + | cache_line_size + | subsys_vendor_id + | subsys_id + | interrupt_pin + | function + +[1] -> Documentation/PCI/endpoint/pci-endpoint.txt -- cgit v1.2.3-59-g8ed1b From 3a401a2ce1cb6f6e52b78f21aa82e5d90e35c430 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:01 +0530 Subject: PCI: endpoint: Create configfs entry for EPC device and EPF driver Invoke APIs provided by pci-ep-cfs to create configfs entry for every EPC device and EPF driver to help users in creating EPF device and binding the EPF device to the EPC device. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/endpoint/pci-epc-core.c | 4 ++++ drivers/pci/endpoint/pci-epf-core.c | 4 ++++ include/linux/pci-epc.h | 2 ++ include/linux/pci-epf.h | 2 ++ 4 files changed, 12 insertions(+) diff --git a/drivers/pci/endpoint/pci-epc-core.c b/drivers/pci/endpoint/pci-epc-core.c index 7c71dd94721c..caa7be10e473 100644 --- a/drivers/pci/endpoint/pci-epc-core.c +++ b/drivers/pci/endpoint/pci-epc-core.c @@ -24,6 +24,7 @@ #include #include +#include static struct class *pci_epc_class; @@ -442,6 +443,7 @@ EXPORT_SYMBOL_GPL(pci_epc_linkup); */ void pci_epc_destroy(struct pci_epc *epc) { + pci_ep_cfs_remove_epc_group(epc->group); device_unregister(&epc->dev); kfree(epc); } @@ -508,6 +510,8 @@ __pci_epc_create(struct device *dev, const struct pci_epc_ops *ops, if (ret) goto put_dev; + epc->group = pci_ep_cfs_add_epc_group(dev_name(dev)); + return epc; put_dev: diff --git a/drivers/pci/endpoint/pci-epf-core.c b/drivers/pci/endpoint/pci-epf-core.c index a281c599a504..6877d6a5bcc9 100644 --- a/drivers/pci/endpoint/pci-epf-core.c +++ b/drivers/pci/endpoint/pci-epf-core.c @@ -24,6 +24,7 @@ #include #include +#include static struct bus_type pci_epf_bus_type; static struct device_type pci_epf_type; @@ -149,6 +150,7 @@ EXPORT_SYMBOL_GPL(pci_epf_alloc_space); */ void pci_epf_unregister_driver(struct pci_epf_driver *driver) { + pci_ep_cfs_remove_epf_group(driver->group); driver_unregister(&driver->driver); } EXPORT_SYMBOL_GPL(pci_epf_unregister_driver); @@ -178,6 +180,8 @@ int __pci_epf_register_driver(struct pci_epf_driver *driver, if (ret) return ret; + driver->group = pci_ep_cfs_add_epf_group(driver->driver.name); + return 0; } EXPORT_SYMBOL_GPL(__pci_epf_register_driver); diff --git a/include/linux/pci-epc.h b/include/linux/pci-epc.h index 8c63d3c37f76..af5edbf3eea3 100644 --- a/include/linux/pci-epc.h +++ b/include/linux/pci-epc.h @@ -77,6 +77,7 @@ struct pci_epc_mem { * @ops: function pointers for performing endpoint operations * @mem: address space of the endpoint controller * @max_functions: max number of functions that can be configured in this EPC + * @group: configfs group representing the PCI EPC device * @lock: spinlock to protect pci_epc ops */ struct pci_epc { @@ -85,6 +86,7 @@ struct pci_epc { const struct pci_epc_ops *ops; struct pci_epc_mem *mem; u8 max_functions; + struct config_group *group; /* spinlock to protect against concurrent access of EP controller */ spinlock_t lock; }; diff --git a/include/linux/pci-epf.h b/include/linux/pci-epf.h index 5628714f7bcf..0d529cb90143 100644 --- a/include/linux/pci-epf.h +++ b/include/linux/pci-epf.h @@ -82,6 +82,7 @@ struct pci_epf_ops { * @driver: PCI EPF driver * @ops: set of function pointers for performing EPF operations * @owner: the owner of the module that registers the PCI EPF driver + * @group: configfs group corresponding to the PCI EPF driver * @id_table: identifies EPF devices for probing */ struct pci_epf_driver { @@ -91,6 +92,7 @@ struct pci_epf_driver { struct device_driver driver; struct pci_epf_ops *ops; struct module *owner; + struct config_group *group; const struct pci_epf_device_id *id_table; }; -- cgit v1.2.3-59-g8ed1b From 42fc2ac6e809a9f383e8ffa78b7f3fcd2fab1649 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:02 +0530 Subject: Documentation: PCI: Add specification for the *PCI test* function device Add specification for the *PCI test* virtual function device. The endpoint function driver and the host PCI driver should be created based on this specification. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- Documentation/PCI/00-INDEX | 2 + Documentation/PCI/endpoint/pci-test-function.txt | 66 ++++++++++++++++++++++++ 2 files changed, 68 insertions(+) create mode 100644 Documentation/PCI/endpoint/pci-test-function.txt diff --git a/Documentation/PCI/00-INDEX b/Documentation/PCI/00-INDEX index bf8223a1bf08..ab35e4bbdb1c 100644 --- a/Documentation/PCI/00-INDEX +++ b/Documentation/PCI/00-INDEX @@ -16,3 +16,5 @@ endpoint/pci-endpoint.txt - guide to add endpoint controller driver and endpoint function driver. endpoint/pci-endpoint-cfs.txt - guide to use configfs to configure the PCI endpoint function. +endpoint/pci-test-function.txt + - specification of *PCI test* function device. diff --git a/Documentation/PCI/endpoint/pci-test-function.txt b/Documentation/PCI/endpoint/pci-test-function.txt new file mode 100644 index 000000000000..0c519c9bf94a --- /dev/null +++ b/Documentation/PCI/endpoint/pci-test-function.txt @@ -0,0 +1,66 @@ + PCI TEST + Kishon Vijay Abraham I + +Traditionally PCI RC has always been validated by using standard +PCI cards like ethernet PCI cards or USB PCI cards or SATA PCI cards. +However with the addition of EP-core in linux kernel, it is possible +to configure a PCI controller that can operate in EP mode to work as +a test device. + +The PCI endpoint test device is a virtual device (defined in software) +used to test the endpoint functionality and serve as a sample driver +for other PCI endpoint devices (to use the EP framework). + +The PCI endpoint test device has the following registers: + + 1) PCI_ENDPOINT_TEST_MAGIC + 2) PCI_ENDPOINT_TEST_COMMAND + 3) PCI_ENDPOINT_TEST_STATUS + 4) PCI_ENDPOINT_TEST_SRC_ADDR + 5) PCI_ENDPOINT_TEST_DST_ADDR + 6) PCI_ENDPOINT_TEST_SIZE + 7) PCI_ENDPOINT_TEST_CHECKSUM + +*) PCI_ENDPOINT_TEST_MAGIC + +This register will be used to test BAR0. A known pattern will be written +and read back from MAGIC register to verify BAR0. + +*) PCI_ENDPOINT_TEST_COMMAND: + +This register will be used by the host driver to indicate the function +that the endpoint device must perform. + +Bitfield Description: + Bit 0 : raise legacy IRQ + Bit 1 : raise MSI IRQ + Bit 2 - 7 : MSI interrupt number + Bit 8 : read command (read data from RC buffer) + Bit 9 : write command (write data to RC buffer) + Bit 10 : copy command (copy data from one RC buffer to another + RC buffer) + +*) PCI_ENDPOINT_TEST_STATUS + +This register reflects the status of the PCI endpoint device. + +Bitfield Description: + Bit 0 : read success + Bit 1 : read fail + Bit 2 : write success + Bit 3 : write fail + Bit 4 : copy success + Bit 5 : copy fail + Bit 6 : IRQ raised + Bit 7 : source address is invalid + Bit 8 : destination address is invalid + +*) PCI_ENDPOINT_TEST_SRC_ADDR + +This register contains the source address (RC buffer address) for the +COPY/READ command. + +*) PCI_ENDPOINT_TEST_DST_ADDR + +This register contains the destination address (RC buffer address) for +the COPY/WRITE command. -- cgit v1.2.3-59-g8ed1b From 349e7a85b25fa6ee82902d9be2cc5f3bec815120 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:03 +0530 Subject: PCI: endpoint: functions: Add an EP function to test PCI Adds a new endpoint function driver (to program the virtual test device) making use of the EP-core library. [bhelgaas: fold in pci_epf_test_probe() -ENOMEM test from Wei Yongjun ] Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/endpoint/Kconfig | 2 + drivers/pci/endpoint/Makefile | 2 +- drivers/pci/endpoint/functions/Kconfig | 12 + drivers/pci/endpoint/functions/Makefile | 5 + drivers/pci/endpoint/functions/pci-epf-test.c | 510 ++++++++++++++++++++++++++ 5 files changed, 530 insertions(+), 1 deletion(-) create mode 100644 drivers/pci/endpoint/functions/Kconfig create mode 100644 drivers/pci/endpoint/functions/Makefile create mode 100644 drivers/pci/endpoint/functions/pci-epf-test.c diff --git a/drivers/pci/endpoint/Kconfig b/drivers/pci/endpoint/Kconfig index c86bca9b7de3..c23f146fb5a6 100644 --- a/drivers/pci/endpoint/Kconfig +++ b/drivers/pci/endpoint/Kconfig @@ -26,4 +26,6 @@ config PCI_ENDPOINT_CONFIGFS configure the endpoint function and used to bind the function with a endpoint controller. +source "drivers/pci/endpoint/functions/Kconfig" + endmenu diff --git a/drivers/pci/endpoint/Makefile b/drivers/pci/endpoint/Makefile index 7219d51bb401..1041f80a4645 100644 --- a/drivers/pci/endpoint/Makefile +++ b/drivers/pci/endpoint/Makefile @@ -4,4 +4,4 @@ obj-$(CONFIG_PCI_ENDPOINT_CONFIGFS) += pci-ep-cfs.o obj-$(CONFIG_PCI_ENDPOINT) += pci-epc-core.o pci-epf-core.o\ - pci-epc-mem.o + pci-epc-mem.o functions/ diff --git a/drivers/pci/endpoint/functions/Kconfig b/drivers/pci/endpoint/functions/Kconfig new file mode 100644 index 000000000000..175edad42d2f --- /dev/null +++ b/drivers/pci/endpoint/functions/Kconfig @@ -0,0 +1,12 @@ +# +# PCI Endpoint Functions +# + +config PCI_EPF_TEST + tristate "PCI Endpoint Test driver" + depends on PCI_ENDPOINT + help + Enable this configuration option to enable the test driver + for PCI Endpoint. + + If in doubt, say "N" to disable Endpoint test driver. diff --git a/drivers/pci/endpoint/functions/Makefile b/drivers/pci/endpoint/functions/Makefile new file mode 100644 index 000000000000..6d94a4801838 --- /dev/null +++ b/drivers/pci/endpoint/functions/Makefile @@ -0,0 +1,5 @@ +# +# Makefile for PCI Endpoint Functions +# + +obj-$(CONFIG_PCI_EPF_TEST) += pci-epf-test.o diff --git a/drivers/pci/endpoint/functions/pci-epf-test.c b/drivers/pci/endpoint/functions/pci-epf-test.c new file mode 100644 index 000000000000..53fff8030337 --- /dev/null +++ b/drivers/pci/endpoint/functions/pci-epf-test.c @@ -0,0 +1,510 @@ +/** + * Test driver to test endpoint functionality + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#define COMMAND_RAISE_LEGACY_IRQ BIT(0) +#define COMMAND_RAISE_MSI_IRQ BIT(1) +#define MSI_NUMBER_SHIFT 2 +#define MSI_NUMBER_MASK (0x3f << MSI_NUMBER_SHIFT) +#define COMMAND_READ BIT(8) +#define COMMAND_WRITE BIT(9) +#define COMMAND_COPY BIT(10) + +#define STATUS_READ_SUCCESS BIT(0) +#define STATUS_READ_FAIL BIT(1) +#define STATUS_WRITE_SUCCESS BIT(2) +#define STATUS_WRITE_FAIL BIT(3) +#define STATUS_COPY_SUCCESS BIT(4) +#define STATUS_COPY_FAIL BIT(5) +#define STATUS_IRQ_RAISED BIT(6) +#define STATUS_SRC_ADDR_INVALID BIT(7) +#define STATUS_DST_ADDR_INVALID BIT(8) + +#define TIMER_RESOLUTION 1 + +static struct workqueue_struct *kpcitest_workqueue; + +struct pci_epf_test { + void *reg[6]; + struct pci_epf *epf; + struct delayed_work cmd_handler; +}; + +struct pci_epf_test_reg { + u32 magic; + u32 command; + u32 status; + u64 src_addr; + u64 dst_addr; + u32 size; + u32 checksum; +} __packed; + +static struct pci_epf_header test_header = { + .vendorid = PCI_ANY_ID, + .deviceid = PCI_ANY_ID, + .baseclass_code = PCI_CLASS_OTHERS, + .interrupt_pin = PCI_INTERRUPT_INTA, +}; + +static int bar_size[] = { 512, 1024, 16384, 131072, 1048576 }; + +static int pci_epf_test_copy(struct pci_epf_test *epf_test) +{ + int ret; + void __iomem *src_addr; + void __iomem *dst_addr; + phys_addr_t src_phys_addr; + phys_addr_t dst_phys_addr; + struct pci_epf *epf = epf_test->epf; + struct device *dev = &epf->dev; + struct pci_epc *epc = epf->epc; + struct pci_epf_test_reg *reg = epf_test->reg[0]; + + src_addr = pci_epc_mem_alloc_addr(epc, &src_phys_addr, reg->size); + if (!src_addr) { + dev_err(dev, "failed to allocate source address\n"); + reg->status = STATUS_SRC_ADDR_INVALID; + ret = -ENOMEM; + goto err; + } + + ret = pci_epc_map_addr(epc, src_phys_addr, reg->src_addr, reg->size); + if (ret) { + dev_err(dev, "failed to map source address\n"); + reg->status = STATUS_SRC_ADDR_INVALID; + goto err_src_addr; + } + + dst_addr = pci_epc_mem_alloc_addr(epc, &dst_phys_addr, reg->size); + if (!dst_addr) { + dev_err(dev, "failed to allocate destination address\n"); + reg->status = STATUS_DST_ADDR_INVALID; + ret = -ENOMEM; + goto err_src_map_addr; + } + + ret = pci_epc_map_addr(epc, dst_phys_addr, reg->dst_addr, reg->size); + if (ret) { + dev_err(dev, "failed to map destination address\n"); + reg->status = STATUS_DST_ADDR_INVALID; + goto err_dst_addr; + } + + memcpy(dst_addr, src_addr, reg->size); + + pci_epc_unmap_addr(epc, dst_phys_addr); + +err_dst_addr: + pci_epc_mem_free_addr(epc, dst_phys_addr, dst_addr, reg->size); + +err_src_map_addr: + pci_epc_unmap_addr(epc, src_phys_addr); + +err_src_addr: + pci_epc_mem_free_addr(epc, src_phys_addr, src_addr, reg->size); + +err: + return ret; +} + +static int pci_epf_test_read(struct pci_epf_test *epf_test) +{ + int ret; + void __iomem *src_addr; + void *buf; + u32 crc32; + phys_addr_t phys_addr; + struct pci_epf *epf = epf_test->epf; + struct device *dev = &epf->dev; + struct pci_epc *epc = epf->epc; + struct pci_epf_test_reg *reg = epf_test->reg[0]; + + src_addr = pci_epc_mem_alloc_addr(epc, &phys_addr, reg->size); + if (!src_addr) { + dev_err(dev, "failed to allocate address\n"); + reg->status = STATUS_SRC_ADDR_INVALID; + ret = -ENOMEM; + goto err; + } + + ret = pci_epc_map_addr(epc, phys_addr, reg->src_addr, reg->size); + if (ret) { + dev_err(dev, "failed to map address\n"); + reg->status = STATUS_SRC_ADDR_INVALID; + goto err_addr; + } + + buf = kzalloc(reg->size, GFP_KERNEL); + if (!buf) { + ret = -ENOMEM; + goto err_map_addr; + } + + memcpy(buf, src_addr, reg->size); + + crc32 = crc32_le(~0, buf, reg->size); + if (crc32 != reg->checksum) + ret = -EIO; + + kfree(buf); + +err_map_addr: + pci_epc_unmap_addr(epc, phys_addr); + +err_addr: + pci_epc_mem_free_addr(epc, phys_addr, src_addr, reg->size); + +err: + return ret; +} + +static int pci_epf_test_write(struct pci_epf_test *epf_test) +{ + int ret; + void __iomem *dst_addr; + void *buf; + phys_addr_t phys_addr; + struct pci_epf *epf = epf_test->epf; + struct device *dev = &epf->dev; + struct pci_epc *epc = epf->epc; + struct pci_epf_test_reg *reg = epf_test->reg[0]; + + dst_addr = pci_epc_mem_alloc_addr(epc, &phys_addr, reg->size); + if (!dst_addr) { + dev_err(dev, "failed to allocate address\n"); + reg->status = STATUS_DST_ADDR_INVALID; + ret = -ENOMEM; + goto err; + } + + ret = pci_epc_map_addr(epc, phys_addr, reg->dst_addr, reg->size); + if (ret) { + dev_err(dev, "failed to map address\n"); + reg->status = STATUS_DST_ADDR_INVALID; + goto err_addr; + } + + buf = kzalloc(reg->size, GFP_KERNEL); + if (!buf) { + ret = -ENOMEM; + goto err_map_addr; + } + + get_random_bytes(buf, reg->size); + reg->checksum = crc32_le(~0, buf, reg->size); + + memcpy(dst_addr, buf, reg->size); + + /* + * wait 1ms inorder for the write to complete. Without this delay L3 + * error in observed in the host system. + */ + mdelay(1); + + kfree(buf); + +err_map_addr: + pci_epc_unmap_addr(epc, phys_addr); + +err_addr: + pci_epc_mem_free_addr(epc, phys_addr, dst_addr, reg->size); + +err: + return ret; +} + +static void pci_epf_test_raise_irq(struct pci_epf_test *epf_test) +{ + u8 irq; + u8 msi_count; + struct pci_epf *epf = epf_test->epf; + struct pci_epc *epc = epf->epc; + struct pci_epf_test_reg *reg = epf_test->reg[0]; + + reg->status |= STATUS_IRQ_RAISED; + msi_count = pci_epc_get_msi(epc); + irq = (reg->command & MSI_NUMBER_MASK) >> MSI_NUMBER_SHIFT; + if (irq > msi_count || msi_count <= 0) + pci_epc_raise_irq(epc, PCI_EPC_IRQ_LEGACY, 0); + else + pci_epc_raise_irq(epc, PCI_EPC_IRQ_MSI, irq); +} + +static void pci_epf_test_cmd_handler(struct work_struct *work) +{ + int ret; + u8 irq; + u8 msi_count; + struct pci_epf_test *epf_test = container_of(work, struct pci_epf_test, + cmd_handler.work); + struct pci_epf *epf = epf_test->epf; + struct pci_epc *epc = epf->epc; + struct pci_epf_test_reg *reg = epf_test->reg[0]; + + if (!reg->command) + goto reset_handler; + + if (reg->command & COMMAND_RAISE_LEGACY_IRQ) { + reg->status = STATUS_IRQ_RAISED; + pci_epc_raise_irq(epc, PCI_EPC_IRQ_LEGACY, 0); + goto reset_handler; + } + + if (reg->command & COMMAND_WRITE) { + ret = pci_epf_test_write(epf_test); + if (ret) + reg->status |= STATUS_WRITE_FAIL; + else + reg->status |= STATUS_WRITE_SUCCESS; + pci_epf_test_raise_irq(epf_test); + goto reset_handler; + } + + if (reg->command & COMMAND_READ) { + ret = pci_epf_test_read(epf_test); + if (!ret) + reg->status |= STATUS_READ_SUCCESS; + else + reg->status |= STATUS_READ_FAIL; + pci_epf_test_raise_irq(epf_test); + goto reset_handler; + } + + if (reg->command & COMMAND_COPY) { + ret = pci_epf_test_copy(epf_test); + if (!ret) + reg->status |= STATUS_COPY_SUCCESS; + else + reg->status |= STATUS_COPY_FAIL; + pci_epf_test_raise_irq(epf_test); + goto reset_handler; + } + + if (reg->command & COMMAND_RAISE_MSI_IRQ) { + msi_count = pci_epc_get_msi(epc); + irq = (reg->command & MSI_NUMBER_MASK) >> MSI_NUMBER_SHIFT; + if (irq > msi_count || msi_count <= 0) + goto reset_handler; + reg->status = STATUS_IRQ_RAISED; + pci_epc_raise_irq(epc, PCI_EPC_IRQ_MSI, irq); + goto reset_handler; + } + +reset_handler: + reg->command = 0; + + queue_delayed_work(kpcitest_workqueue, &epf_test->cmd_handler, + msecs_to_jiffies(1)); +} + +static void pci_epf_test_linkup(struct pci_epf *epf) +{ + struct pci_epf_test *epf_test = epf_get_drvdata(epf); + + queue_delayed_work(kpcitest_workqueue, &epf_test->cmd_handler, + msecs_to_jiffies(1)); +} + +static void pci_epf_test_unbind(struct pci_epf *epf) +{ + struct pci_epf_test *epf_test = epf_get_drvdata(epf); + struct pci_epc *epc = epf->epc; + int bar; + + cancel_delayed_work(&epf_test->cmd_handler); + pci_epc_stop(epc); + for (bar = BAR_0; bar <= BAR_5; bar++) { + if (epf_test->reg[bar]) { + pci_epf_free_space(epf, epf_test->reg[bar], bar); + pci_epc_clear_bar(epc, bar); + } + } +} + +static int pci_epf_test_set_bar(struct pci_epf *epf) +{ + int flags; + int bar; + int ret; + struct pci_epf_bar *epf_bar; + struct pci_epc *epc = epf->epc; + struct device *dev = &epf->dev; + struct pci_epf_test *epf_test = epf_get_drvdata(epf); + + flags = PCI_BASE_ADDRESS_SPACE_MEMORY | PCI_BASE_ADDRESS_MEM_TYPE_32; + if (sizeof(dma_addr_t) == 0x8) + flags |= PCI_BASE_ADDRESS_MEM_TYPE_64; + + for (bar = BAR_0; bar <= BAR_5; bar++) { + epf_bar = &epf->bar[bar]; + ret = pci_epc_set_bar(epc, bar, epf_bar->phys_addr, + epf_bar->size, flags); + if (ret) { + pci_epf_free_space(epf, epf_test->reg[bar], bar); + dev_err(dev, "failed to set BAR%d\n", bar); + if (bar == BAR_0) + return ret; + } + } + + return 0; +} + +static int pci_epf_test_alloc_space(struct pci_epf *epf) +{ + struct pci_epf_test *epf_test = epf_get_drvdata(epf); + struct device *dev = &epf->dev; + void *base; + int bar; + + base = pci_epf_alloc_space(epf, sizeof(struct pci_epf_test_reg), + BAR_0); + if (!base) { + dev_err(dev, "failed to allocated register space\n"); + return -ENOMEM; + } + epf_test->reg[0] = base; + + for (bar = BAR_1; bar <= BAR_5; bar++) { + base = pci_epf_alloc_space(epf, bar_size[bar - 1], bar); + if (!base) + dev_err(dev, "failed to allocate space for BAR%d\n", + bar); + epf_test->reg[bar] = base; + } + + return 0; +} + +static int pci_epf_test_bind(struct pci_epf *epf) +{ + int ret; + struct pci_epf_header *header = epf->header; + struct pci_epc *epc = epf->epc; + struct device *dev = &epf->dev; + + if (WARN_ON_ONCE(!epc)) + return -EINVAL; + + ret = pci_epc_write_header(epc, header); + if (ret) { + dev_err(dev, "configuration header write failed\n"); + return ret; + } + + ret = pci_epf_test_alloc_space(epf); + if (ret) + return ret; + + ret = pci_epf_test_set_bar(epf); + if (ret) + return ret; + + ret = pci_epc_set_msi(epc, epf->msi_interrupts); + if (ret) + return ret; + + return 0; +} + +static int pci_epf_test_probe(struct pci_epf *epf) +{ + struct pci_epf_test *epf_test; + struct device *dev = &epf->dev; + + epf_test = devm_kzalloc(dev, sizeof(*epf_test), GFP_KERNEL); + if (!epf_test) + return -ENOMEM; + + epf->header = &test_header; + epf_test->epf = epf; + + INIT_DELAYED_WORK(&epf_test->cmd_handler, pci_epf_test_cmd_handler); + + epf_set_drvdata(epf, epf_test); + return 0; +} + +static int pci_epf_test_remove(struct pci_epf *epf) +{ + struct pci_epf_test *epf_test = epf_get_drvdata(epf); + + kfree(epf_test); + return 0; +} + +static struct pci_epf_ops ops = { + .unbind = pci_epf_test_unbind, + .bind = pci_epf_test_bind, + .linkup = pci_epf_test_linkup, +}; + +static const struct pci_epf_device_id pci_epf_test_ids[] = { + { + .name = "pci_epf_test", + }, + {}, +}; + +static struct pci_epf_driver test_driver = { + .driver.name = "pci_epf_test", + .probe = pci_epf_test_probe, + .remove = pci_epf_test_remove, + .id_table = pci_epf_test_ids, + .ops = &ops, + .owner = THIS_MODULE, +}; + +static int __init pci_epf_test_init(void) +{ + int ret; + + kpcitest_workqueue = alloc_workqueue("kpcitest", + WQ_MEM_RECLAIM | WQ_HIGHPRI, 0); + ret = pci_epf_register_driver(&test_driver); + if (ret) { + pr_err("failed to register pci epf test driver --> %d\n", ret); + return ret; + } + + return 0; +} +module_init(pci_epf_test_init); + +static void __exit pci_epf_test_exit(void) +{ + pci_epf_unregister_driver(&test_driver); +} +module_exit(pci_epf_test_exit); + +MODULE_DESCRIPTION("PCI EPF TEST DRIVER"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); -- cgit v1.2.3-59-g8ed1b From 64d6ea602ce619633a6e0af979e2c73738f6aeba Mon Sep 17 00:00:00 2001 From: Shawn Lin Date: Tue, 11 Apr 2017 16:27:02 -0500 Subject: PCI: rockchip: Set PCI_EXP_LNKSTA_SLC in the Root Port All platforms using Rockchip use a common clock for the Root Port and the slot connected to it. Indicate this by setting the Slot Clock Configuration (PCI_EXP_LNKSTA_SLC) bit in the Root Port's Link Status. Per the Implementation Note in the spec (PCIe r3.1, sec 7.8.7), if the downstream component also sets PCI_EXP_LNKSTA_SLC, software may set the Common Clock Configuration (PCI_EXP_LNKCTL_CCC) bits on both ends of the Link. This is done by pcie_aspm_configure_common_clock(). Signed-off-by: Shawn Lin Cc: Brian Norris Cc: jeffy.chen --- drivers/pci/host/pcie-rockchip.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index a7467212ea18..94feb7dfd8f4 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -598,7 +598,7 @@ static int rockchip_pcie_init_port(struct rockchip_pcie *rockchip) /* Set RC's clock architecture as common clock */ status = rockchip_pcie_read(rockchip, PCIE_RC_CONFIG_LCS); - status |= PCI_EXP_LNKCTL_CCC; + status |= PCI_EXP_LNKSTA_SLC << 16; rockchip_pcie_write(rockchip, status, PCIE_RC_CONFIG_LCS); /* Set RC's RCB to 128 */ -- cgit v1.2.3-59-g8ed1b From 44fc691b4fac3aff3e99b42cad3a994f5a6621ac Mon Sep 17 00:00:00 2001 From: Logan Gunthorpe Date: Thu, 2 Mar 2017 16:24:32 -0700 Subject: switchtec: Add user interface documentation Add standard documentation for the sysfs switchtec attributes and a RST formatted text file which documents the char device interface. Jonathan Corbet has indicated he will move this to a new user-space developer documentation book once it's created. Tested-by: Krishna Dhulipala Signed-off-by: Logan Gunthorpe Signed-off-by: Stephen Bates Signed-off-by: Bjorn Helgaas Reviewed-by: Wei Zhang Reviewed-by: Jens Axboe --- Documentation/switchtec.txt | 53 +++++++++++++++++++++++++++++++++++++++++++++ MAINTAINERS | 1 + 2 files changed, 54 insertions(+) create mode 100644 Documentation/switchtec.txt diff --git a/Documentation/switchtec.txt b/Documentation/switchtec.txt new file mode 100644 index 000000000000..4bced4c78446 --- /dev/null +++ b/Documentation/switchtec.txt @@ -0,0 +1,53 @@ +======================== +Linux Switchtec Support +======================== + +Microsemi's "Switchtec" line of PCI switch devices is already +supported by the kernel with standard PCI switch drivers. However, the +Switchtec device advertises a special management endpoint which +enables some additional functionality. This includes: + +* Packet and Byte Counters +* Firmware Upgrades +* Event and Error logs +* Querying port link status +* Custom user firmware commands + +The switchtec kernel module implements this functionality. + + +Interface +========= + +The primary means of communicating with the Switchtec management firmware is +through the Memory-mapped Remote Procedure Call (MRPC) interface. +Commands are submitted to the interface with a 4-byte command +identifier and up to 1KB of command specific data. The firmware will +respond with a 4 bytes return code and up to 1KB of command specific +data. The interface only processes a single command at a time. + + +Userspace Interface +=================== + +The MRPC interface will be exposed to userspace through a simple char +device: /dev/switchtec#, one for each management endpoint in the system. + +The char device has the following semantics: + +* A write must consist of at least 4 bytes and no more than 1028 bytes. + The first four bytes will be interpreted as the command to run and + the remainder will be used as the input data. A write will send the + command to the firmware to begin processing. + +* Each write must be followed by exactly one read. Any double write will + produce an error and any read that doesn't follow a write will + produce an error. + +* A read will block until the firmware completes the command and return + the four bytes of status plus up to 1024 bytes of output data. (The + length will be specified by the size parameter of the read call -- + reading less than 4 bytes will produce an error. + +* The poll call will also be supported for userspace applications that + need to do other things while waiting for the command to complete. diff --git a/MAINTAINERS b/MAINTAINERS index 16bd75d30ac1..5bc3b7c7e0d9 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9661,6 +9661,7 @@ M: Stephen Bates M: Logan Gunthorpe L: linux-pci@vger.kernel.org S: Maintained +F: Documentation/switchtec.txt F: drivers/pci/switch/switchtec* PCI DRIVER FOR NVIDIA TEGRA -- cgit v1.2.3-59-g8ed1b From 5d8e1881f431cf470170813eb048e6a384340904 Mon Sep 17 00:00:00 2001 From: Logan Gunthorpe Date: Thu, 2 Mar 2017 16:24:33 -0700 Subject: switchtec: Add sysfs attributes to the Switchtec driver Add a few read-only sysfs attributes which provide some device information that is exposed from the devices, primarily component and device names and versions. These are documented in Documentation/ABI/testing/sysfs-class-switchtec. Tested-by: Krishna Dhulipala Signed-off-by: Logan Gunthorpe Signed-off-by: Stephen Bates Signed-off-by: Bjorn Helgaas Reviewed-by: Wei Zhang Reviewed-by: Jens Axboe Reviewed-by: Greg Kroah-Hartman --- Documentation/ABI/testing/sysfs-class-switchtec | 96 ++++++++++++++++++++ MAINTAINERS | 1 + drivers/pci/switch/switchtec.c | 113 ++++++++++++++++++++++++ 3 files changed, 210 insertions(+) create mode 100644 Documentation/ABI/testing/sysfs-class-switchtec diff --git a/Documentation/ABI/testing/sysfs-class-switchtec b/Documentation/ABI/testing/sysfs-class-switchtec new file mode 100644 index 000000000000..48cb4c15e430 --- /dev/null +++ b/Documentation/ABI/testing/sysfs-class-switchtec @@ -0,0 +1,96 @@ +switchtec - Microsemi Switchtec PCI Switch Management Endpoint + +For details on this subsystem look at Documentation/switchtec.txt. + +What: /sys/class/switchtec +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: The switchtec class subsystem folder. + Each registered switchtec driver is represented by a switchtecX + subfolder (X being an integer >= 0). + + +What: /sys/class/switchtec/switchtec[0-9]+/component_id +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Component identifier as stored in the hardware (eg. PM8543) + (read only) +Values: arbitrary string. + + +What: /sys/class/switchtec/switchtec[0-9]+/component_revision +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Component revision stored in the hardware (read only) +Values: integer. + + +What: /sys/class/switchtec/switchtec[0-9]+/component_vendor +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Component vendor as stored in the hardware (eg. MICROSEM) + (read only) +Values: arbitrary string. + + +What: /sys/class/switchtec/switchtec[0-9]+/device_version +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Device version as stored in the hardware (read only) +Values: integer. + + +What: /sys/class/switchtec/switchtec[0-9]+/fw_version +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Currently running firmware version (read only) +Values: integer (in hexadecimal). + + +What: /sys/class/switchtec/switchtec[0-9]+/partition +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Partition number for this device in the switch (read only) +Values: integer. + + +What: /sys/class/switchtec/switchtec[0-9]+/partition_count +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Total number of partitions in the switch (read only) +Values: integer. + + +What: /sys/class/switchtec/switchtec[0-9]+/product_id +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Product identifier as stored in the hardware (eg. PSX 48XG3) + (read only) +Values: arbitrary string. + + +What: /sys/class/switchtec/switchtec[0-9]+/product_revision +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Product revision stored in the hardware (eg. RevB) + (read only) +Values: arbitrary string. + + +What: /sys/class/switchtec/switchtec[0-9]+/product_vendor +Date: 05-Jan-2017 +KernelVersion: v4.11 +Contact: Logan Gunthorpe +Description: Product vendor as stored in the hardware (eg. MICROSEM) + (read only) +Values: arbitrary string. diff --git a/MAINTAINERS b/MAINTAINERS index 5bc3b7c7e0d9..76ccc5a805dd 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9662,6 +9662,7 @@ M: Logan Gunthorpe L: linux-pci@vger.kernel.org S: Maintained F: Documentation/switchtec.txt +F: Documentation/ABI/testing/sysfs-class-switchtec F: drivers/pci/switch/switchtec* PCI DRIVER FOR NVIDIA TEGRA diff --git a/drivers/pci/switch/switchtec.c b/drivers/pci/switch/switchtec.c index 0bbe90df968d..15ff61c7718b 100644 --- a/drivers/pci/switch/switchtec.c +++ b/drivers/pci/switch/switchtec.c @@ -486,6 +486,118 @@ out: mutex_unlock(&stdev->mrpc_mutex); } +static ssize_t device_version_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + u32 ver; + + ver = ioread32(&stdev->mmio_sys_info->device_version); + + return sprintf(buf, "%x\n", ver); +} +static DEVICE_ATTR_RO(device_version); + +static ssize_t fw_version_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + u32 ver; + + ver = ioread32(&stdev->mmio_sys_info->firmware_version); + + return sprintf(buf, "%08x\n", ver); +} +static DEVICE_ATTR_RO(fw_version); + +static ssize_t io_string_show(char *buf, void __iomem *attr, size_t len) +{ + int i; + + memcpy_fromio(buf, attr, len); + buf[len] = '\n'; + buf[len + 1] = 0; + + for (i = len - 1; i > 0; i--) { + if (buf[i] != ' ') + break; + buf[i] = '\n'; + buf[i + 1] = 0; + } + + return strlen(buf); +} + +#define DEVICE_ATTR_SYS_INFO_STR(field) \ +static ssize_t field ## _show(struct device *dev, \ + struct device_attribute *attr, char *buf) \ +{ \ + struct switchtec_dev *stdev = to_stdev(dev); \ + return io_string_show(buf, &stdev->mmio_sys_info->field, \ + sizeof(stdev->mmio_sys_info->field)); \ +} \ +\ +static DEVICE_ATTR_RO(field) + +DEVICE_ATTR_SYS_INFO_STR(vendor_id); +DEVICE_ATTR_SYS_INFO_STR(product_id); +DEVICE_ATTR_SYS_INFO_STR(product_revision); +DEVICE_ATTR_SYS_INFO_STR(component_vendor); + +static ssize_t component_id_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + int id = ioread16(&stdev->mmio_sys_info->component_id); + + return sprintf(buf, "PM%04X\n", id); +} +static DEVICE_ATTR_RO(component_id); + +static ssize_t component_revision_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + int rev = ioread8(&stdev->mmio_sys_info->component_revision); + + return sprintf(buf, "%d\n", rev); +} +static DEVICE_ATTR_RO(component_revision); + +static ssize_t partition_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + + return sprintf(buf, "%d\n", stdev->partition); +} +static DEVICE_ATTR_RO(partition); + +static ssize_t partition_count_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct switchtec_dev *stdev = to_stdev(dev); + + return sprintf(buf, "%d\n", stdev->partition_count); +} +static DEVICE_ATTR_RO(partition_count); + +static struct attribute *switchtec_device_attrs[] = { + &dev_attr_device_version.attr, + &dev_attr_fw_version.attr, + &dev_attr_vendor_id.attr, + &dev_attr_product_id.attr, + &dev_attr_product_revision.attr, + &dev_attr_component_vendor.attr, + &dev_attr_component_id.attr, + &dev_attr_component_revision.attr, + &dev_attr_partition.attr, + &dev_attr_partition_count.attr, + NULL, +}; + +ATTRIBUTE_GROUPS(switchtec_device); + static int switchtec_dev_open(struct inode *inode, struct file *filp) { struct switchtec_dev *stdev; @@ -734,6 +846,7 @@ static struct switchtec_dev *stdev_create(struct pci_dev *pdev) device_initialize(dev); dev->class = switchtec_class; dev->parent = &pdev->dev; + dev->groups = switchtec_device_groups; dev->release = stdev_release; minor = ida_simple_get(&switchtec_minor_ida, 0, 0, -- cgit v1.2.3-59-g8ed1b From 52eabba5bcdb2853dec6ef007ba427b092034738 Mon Sep 17 00:00:00 2001 From: Logan Gunthorpe Date: Thu, 2 Mar 2017 16:24:34 -0700 Subject: switchtec: Add IOCTLs to the Switchtec driver Add a couple of special IOCTLs to: * Inform userspace of firmware partition locations * Pass event counts and allow userspace to wait on events * Translate PFF numbers used by the switch to port numbers [Dan Carpenter : fix off-by-one in ioctl_event_ctl()] Tested-by: Krishna Dhulipala Signed-off-by: Logan Gunthorpe Signed-off-by: Stephen Bates Signed-off-by: Bjorn Helgaas Reviewed-by: Wei Zhang Reviewed-by: Jens Axboe --- Documentation/ioctl/ioctl-number.txt | 1 + Documentation/switchtec.txt | 27 ++ MAINTAINERS | 1 + drivers/pci/switch/switchtec.c | 481 +++++++++++++++++++++++++++++++++++ include/uapi/linux/switchtec_ioctl.h | 132 ++++++++++ 5 files changed, 642 insertions(+) create mode 100644 include/uapi/linux/switchtec_ioctl.h diff --git a/Documentation/ioctl/ioctl-number.txt b/Documentation/ioctl/ioctl-number.txt index 08244bea5048..0682bd3eaa8a 100644 --- a/Documentation/ioctl/ioctl-number.txt +++ b/Documentation/ioctl/ioctl-number.txt @@ -191,6 +191,7 @@ Code Seq#(hex) Include File Comments 'W' 00-1F linux/watchdog.h conflict! 'W' 00-1F linux/wanrouter.h conflict! (pre 3.9) 'W' 00-3F sound/asound.h conflict! +'W' 40-5F drivers/pci/switch/switchtec.c 'X' all fs/xfs/xfs_fs.h conflict! and fs/xfs/linux-2.6/xfs_ioctl32.h and include/linux/falloc.h diff --git a/Documentation/switchtec.txt b/Documentation/switchtec.txt index 4bced4c78446..a0a9c7b3d4d5 100644 --- a/Documentation/switchtec.txt +++ b/Documentation/switchtec.txt @@ -51,3 +51,30 @@ The char device has the following semantics: * The poll call will also be supported for userspace applications that need to do other things while waiting for the command to complete. + +The following IOCTLs are also supported by the device: + +* SWITCHTEC_IOCTL_FLASH_INFO - Retrieve firmware length and number + of partitions in the device. + +* SWITCHTEC_IOCTL_FLASH_PART_INFO - Retrieve address and lengeth for + any specified partition in flash. + +* SWITCHTEC_IOCTL_EVENT_SUMMARY - Read a structure of bitmaps + indicating all uncleared events. + +* SWITCHTEC_IOCTL_EVENT_CTL - Get the current count, clear and set flags + for any event. This ioctl takes in a switchtec_ioctl_event_ctl struct + with the event_id, index and flags set (index being the partition or PFF + number for non-global events). It returns whether the event has + occurred, the number of times and any event specific data. The flags + can be used to clear the count or enable and disable actions to + happen when the event occurs. + By using the SWITCHTEC_IOCTL_EVENT_FLAG_EN_POLL flag, + you can set an event to trigger a poll command to return with + POLLPRI. In this way, userspace can wait for events to occur. + +* SWITCHTEC_IOCTL_PFF_TO_PORT and SWITCHTEC_IOCTL_PORT_TO_PFF convert + between PCI Function Framework number (used by the event system) + and Switchtec Logic Port ID and Partition number (which is more + user friendly). diff --git a/MAINTAINERS b/MAINTAINERS index 76ccc5a805dd..3744019a8853 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9664,6 +9664,7 @@ S: Maintained F: Documentation/switchtec.txt F: Documentation/ABI/testing/sysfs-class-switchtec F: drivers/pci/switch/switchtec* +F: include/uapi/linux/switchtec_ioctl.h PCI DRIVER FOR NVIDIA TEGRA M: Thierry Reding diff --git a/drivers/pci/switch/switchtec.c b/drivers/pci/switch/switchtec.c index 15ff61c7718b..cc6e085008fb 100644 --- a/drivers/pci/switch/switchtec.c +++ b/drivers/pci/switch/switchtec.c @@ -13,6 +13,8 @@ * */ +#include + #include #include #include @@ -778,6 +780,431 @@ static unsigned int switchtec_dev_poll(struct file *filp, poll_table *wait) return ret; } +static int ioctl_flash_info(struct switchtec_dev *stdev, + struct switchtec_ioctl_flash_info __user *uinfo) +{ + struct switchtec_ioctl_flash_info info = {0}; + struct flash_info_regs __iomem *fi = stdev->mmio_flash_info; + + info.flash_length = ioread32(&fi->flash_length); + info.num_partitions = SWITCHTEC_IOCTL_NUM_PARTITIONS; + + if (copy_to_user(uinfo, &info, sizeof(info))) + return -EFAULT; + + return 0; +} + +static void set_fw_info_part(struct switchtec_ioctl_flash_part_info *info, + struct partition_info __iomem *pi) +{ + info->address = ioread32(&pi->address); + info->length = ioread32(&pi->length); +} + +static int ioctl_flash_part_info(struct switchtec_dev *stdev, + struct switchtec_ioctl_flash_part_info __user *uinfo) +{ + struct switchtec_ioctl_flash_part_info info = {0}; + struct flash_info_regs __iomem *fi = stdev->mmio_flash_info; + u32 active_addr = -1; + + if (copy_from_user(&info, uinfo, sizeof(info))) + return -EFAULT; + + switch (info.flash_partition) { + case SWITCHTEC_IOCTL_PART_CFG0: + active_addr = ioread32(&fi->active_cfg); + set_fw_info_part(&info, &fi->cfg0); + break; + case SWITCHTEC_IOCTL_PART_CFG1: + active_addr = ioread32(&fi->active_cfg); + set_fw_info_part(&info, &fi->cfg1); + break; + case SWITCHTEC_IOCTL_PART_IMG0: + active_addr = ioread32(&fi->active_img); + set_fw_info_part(&info, &fi->img0); + break; + case SWITCHTEC_IOCTL_PART_IMG1: + active_addr = ioread32(&fi->active_img); + set_fw_info_part(&info, &fi->img1); + break; + case SWITCHTEC_IOCTL_PART_NVLOG: + set_fw_info_part(&info, &fi->nvlog); + break; + case SWITCHTEC_IOCTL_PART_VENDOR0: + set_fw_info_part(&info, &fi->vendor[0]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR1: + set_fw_info_part(&info, &fi->vendor[1]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR2: + set_fw_info_part(&info, &fi->vendor[2]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR3: + set_fw_info_part(&info, &fi->vendor[3]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR4: + set_fw_info_part(&info, &fi->vendor[4]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR5: + set_fw_info_part(&info, &fi->vendor[5]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR6: + set_fw_info_part(&info, &fi->vendor[6]); + break; + case SWITCHTEC_IOCTL_PART_VENDOR7: + set_fw_info_part(&info, &fi->vendor[7]); + break; + default: + return -EINVAL; + } + + if (info.address == active_addr) + info.active = 1; + + if (copy_to_user(uinfo, &info, sizeof(info))) + return -EFAULT; + + return 0; +} + +static int ioctl_event_summary(struct switchtec_dev *stdev, + struct switchtec_user *stuser, + struct switchtec_ioctl_event_summary __user *usum) +{ + struct switchtec_ioctl_event_summary s = {0}; + int i; + u32 reg; + + s.global = ioread32(&stdev->mmio_sw_event->global_summary); + s.part_bitmap = ioread32(&stdev->mmio_sw_event->part_event_bitmap); + s.local_part = ioread32(&stdev->mmio_part_cfg->part_event_summary); + + for (i = 0; i < stdev->partition_count; i++) { + reg = ioread32(&stdev->mmio_part_cfg_all[i].part_event_summary); + s.part[i] = reg; + } + + for (i = 0; i < SWITCHTEC_MAX_PFF_CSR; i++) { + reg = ioread16(&stdev->mmio_pff_csr[i].vendor_id); + if (reg != MICROSEMI_VENDOR_ID) + break; + + reg = ioread32(&stdev->mmio_pff_csr[i].pff_event_summary); + s.pff[i] = reg; + } + + if (copy_to_user(usum, &s, sizeof(s))) + return -EFAULT; + + stuser->event_cnt = atomic_read(&stdev->event_cnt); + + return 0; +} + +static u32 __iomem *global_ev_reg(struct switchtec_dev *stdev, + size_t offset, int index) +{ + return (void __iomem *)stdev->mmio_sw_event + offset; +} + +static u32 __iomem *part_ev_reg(struct switchtec_dev *stdev, + size_t offset, int index) +{ + return (void __iomem *)&stdev->mmio_part_cfg_all[index] + offset; +} + +static u32 __iomem *pff_ev_reg(struct switchtec_dev *stdev, + size_t offset, int index) +{ + return (void __iomem *)&stdev->mmio_pff_csr[index] + offset; +} + +#define EV_GLB(i, r)[i] = {offsetof(struct sw_event_regs, r), global_ev_reg} +#define EV_PAR(i, r)[i] = {offsetof(struct part_cfg_regs, r), part_ev_reg} +#define EV_PFF(i, r)[i] = {offsetof(struct pff_csr_regs, r), pff_ev_reg} + +const struct event_reg { + size_t offset; + u32 __iomem *(*map_reg)(struct switchtec_dev *stdev, + size_t offset, int index); +} event_regs[] = { + EV_GLB(SWITCHTEC_IOCTL_EVENT_STACK_ERROR, stack_error_event_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_PPU_ERROR, ppu_error_event_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_ISP_ERROR, isp_error_event_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_SYS_RESET, sys_reset_event_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_FW_EXC, fw_exception_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_FW_NMI, fw_nmi_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_FW_NON_FATAL, fw_non_fatal_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_FW_FATAL, fw_fatal_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_TWI_MRPC_COMP, twi_mrpc_comp_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_TWI_MRPC_COMP_ASYNC, + twi_mrpc_comp_async_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_CLI_MRPC_COMP, cli_mrpc_comp_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_CLI_MRPC_COMP_ASYNC, + cli_mrpc_comp_async_hdr), + EV_GLB(SWITCHTEC_IOCTL_EVENT_GPIO_INT, gpio_interrupt_hdr), + EV_PAR(SWITCHTEC_IOCTL_EVENT_PART_RESET, part_reset_hdr), + EV_PAR(SWITCHTEC_IOCTL_EVENT_MRPC_COMP, mrpc_comp_hdr), + EV_PAR(SWITCHTEC_IOCTL_EVENT_MRPC_COMP_ASYNC, mrpc_comp_async_hdr), + EV_PAR(SWITCHTEC_IOCTL_EVENT_DYN_PART_BIND_COMP, dyn_binding_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_AER_IN_P2P, aer_in_p2p_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_AER_IN_VEP, aer_in_vep_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_DPC, dpc_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_CTS, cts_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_HOTPLUG, hotplug_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_IER, ier_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_THRESH, threshold_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_POWER_MGMT, power_mgmt_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_TLP_THROTTLING, tlp_throttling_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_FORCE_SPEED, force_speed_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_CREDIT_TIMEOUT, credit_timeout_hdr), + EV_PFF(SWITCHTEC_IOCTL_EVENT_LINK_STATE, link_state_hdr), +}; + +static u32 __iomem *event_hdr_addr(struct switchtec_dev *stdev, + int event_id, int index) +{ + size_t off; + + if (event_id < 0 || event_id >= SWITCHTEC_IOCTL_MAX_EVENTS) + return ERR_PTR(-EINVAL); + + off = event_regs[event_id].offset; + + if (event_regs[event_id].map_reg == part_ev_reg) { + if (index == SWITCHTEC_IOCTL_EVENT_LOCAL_PART_IDX) + index = stdev->partition; + else if (index < 0 || index >= stdev->partition_count) + return ERR_PTR(-EINVAL); + } else if (event_regs[event_id].map_reg == pff_ev_reg) { + if (index < 0 || index >= stdev->pff_csr_count) + return ERR_PTR(-EINVAL); + } + + return event_regs[event_id].map_reg(stdev, off, index); +} + +static int event_ctl(struct switchtec_dev *stdev, + struct switchtec_ioctl_event_ctl *ctl) +{ + int i; + u32 __iomem *reg; + u32 hdr; + + reg = event_hdr_addr(stdev, ctl->event_id, ctl->index); + if (IS_ERR(reg)) + return PTR_ERR(reg); + + hdr = ioread32(reg); + for (i = 0; i < ARRAY_SIZE(ctl->data); i++) + ctl->data[i] = ioread32(®[i + 1]); + + ctl->occurred = hdr & SWITCHTEC_EVENT_OCCURRED; + ctl->count = (hdr >> 5) & 0xFF; + + if (!(ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_CLEAR)) + hdr &= ~SWITCHTEC_EVENT_CLEAR; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_EN_POLL) + hdr |= SWITCHTEC_EVENT_EN_IRQ; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_DIS_POLL) + hdr &= ~SWITCHTEC_EVENT_EN_IRQ; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_EN_LOG) + hdr |= SWITCHTEC_EVENT_EN_LOG; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_DIS_LOG) + hdr &= ~SWITCHTEC_EVENT_EN_LOG; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_EN_CLI) + hdr |= SWITCHTEC_EVENT_EN_CLI; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_DIS_CLI) + hdr &= ~SWITCHTEC_EVENT_EN_CLI; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_EN_FATAL) + hdr |= SWITCHTEC_EVENT_FATAL; + if (ctl->flags & SWITCHTEC_IOCTL_EVENT_FLAG_DIS_FATAL) + hdr &= ~SWITCHTEC_EVENT_FATAL; + + if (ctl->flags) + iowrite32(hdr, reg); + + ctl->flags = 0; + if (hdr & SWITCHTEC_EVENT_EN_IRQ) + ctl->flags |= SWITCHTEC_IOCTL_EVENT_FLAG_EN_POLL; + if (hdr & SWITCHTEC_EVENT_EN_LOG) + ctl->flags |= SWITCHTEC_IOCTL_EVENT_FLAG_EN_LOG; + if (hdr & SWITCHTEC_EVENT_EN_CLI) + ctl->flags |= SWITCHTEC_IOCTL_EVENT_FLAG_EN_CLI; + if (hdr & SWITCHTEC_EVENT_FATAL) + ctl->flags |= SWITCHTEC_IOCTL_EVENT_FLAG_EN_FATAL; + + return 0; +} + +static int ioctl_event_ctl(struct switchtec_dev *stdev, + struct switchtec_ioctl_event_ctl __user *uctl) +{ + int ret; + int nr_idxs; + struct switchtec_ioctl_event_ctl ctl; + + if (copy_from_user(&ctl, uctl, sizeof(ctl))) + return -EFAULT; + + if (ctl.event_id >= SWITCHTEC_IOCTL_MAX_EVENTS) + return -EINVAL; + + if (ctl.flags & SWITCHTEC_IOCTL_EVENT_FLAG_UNUSED) + return -EINVAL; + + if (ctl.index == SWITCHTEC_IOCTL_EVENT_IDX_ALL) { + if (event_regs[ctl.event_id].map_reg == global_ev_reg) + nr_idxs = 1; + else if (event_regs[ctl.event_id].map_reg == part_ev_reg) + nr_idxs = stdev->partition_count; + else if (event_regs[ctl.event_id].map_reg == pff_ev_reg) + nr_idxs = stdev->pff_csr_count; + else + return -EINVAL; + + for (ctl.index = 0; ctl.index < nr_idxs; ctl.index++) { + ret = event_ctl(stdev, &ctl); + if (ret < 0) + return ret; + } + } else { + ret = event_ctl(stdev, &ctl); + if (ret < 0) + return ret; + } + + if (copy_to_user(uctl, &ctl, sizeof(ctl))) + return -EFAULT; + + return 0; +} + +static int ioctl_pff_to_port(struct switchtec_dev *stdev, + struct switchtec_ioctl_pff_port *up) +{ + int i, part; + u32 reg; + struct part_cfg_regs *pcfg; + struct switchtec_ioctl_pff_port p; + + if (copy_from_user(&p, up, sizeof(p))) + return -EFAULT; + + p.port = -1; + for (part = 0; part < stdev->partition_count; part++) { + pcfg = &stdev->mmio_part_cfg_all[part]; + p.partition = part; + + reg = ioread32(&pcfg->usp_pff_inst_id); + if (reg == p.pff) { + p.port = 0; + break; + } + + reg = ioread32(&pcfg->vep_pff_inst_id); + if (reg == p.pff) { + p.port = SWITCHTEC_IOCTL_PFF_VEP; + break; + } + + for (i = 0; i < ARRAY_SIZE(pcfg->dsp_pff_inst_id); i++) { + reg = ioread32(&pcfg->dsp_pff_inst_id[i]); + if (reg != p.pff) + continue; + + p.port = i + 1; + break; + } + + if (p.port != -1) + break; + } + + if (copy_to_user(up, &p, sizeof(p))) + return -EFAULT; + + return 0; +} + +static int ioctl_port_to_pff(struct switchtec_dev *stdev, + struct switchtec_ioctl_pff_port *up) +{ + struct switchtec_ioctl_pff_port p; + struct part_cfg_regs *pcfg; + + if (copy_from_user(&p, up, sizeof(p))) + return -EFAULT; + + if (p.partition == SWITCHTEC_IOCTL_EVENT_LOCAL_PART_IDX) + pcfg = stdev->mmio_part_cfg; + else if (p.partition < stdev->partition_count) + pcfg = &stdev->mmio_part_cfg_all[p.partition]; + else + return -EINVAL; + + switch (p.port) { + case 0: + p.pff = ioread32(&pcfg->usp_pff_inst_id); + break; + case SWITCHTEC_IOCTL_PFF_VEP: + p.pff = ioread32(&pcfg->vep_pff_inst_id); + break; + default: + if (p.port > ARRAY_SIZE(pcfg->dsp_pff_inst_id)) + return -EINVAL; + p.pff = ioread32(&pcfg->dsp_pff_inst_id[p.port - 1]); + break; + } + + if (copy_to_user(up, &p, sizeof(p))) + return -EFAULT; + + return 0; +} + +static long switchtec_dev_ioctl(struct file *filp, unsigned int cmd, + unsigned long arg) +{ + struct switchtec_user *stuser = filp->private_data; + struct switchtec_dev *stdev = stuser->stdev; + int rc; + void __user *argp = (void __user *)arg; + + rc = lock_mutex_and_test_alive(stdev); + if (rc) + return rc; + + switch (cmd) { + case SWITCHTEC_IOCTL_FLASH_INFO: + rc = ioctl_flash_info(stdev, argp); + break; + case SWITCHTEC_IOCTL_FLASH_PART_INFO: + rc = ioctl_flash_part_info(stdev, argp); + break; + case SWITCHTEC_IOCTL_EVENT_SUMMARY: + rc = ioctl_event_summary(stdev, stuser, argp); + break; + case SWITCHTEC_IOCTL_EVENT_CTL: + rc = ioctl_event_ctl(stdev, argp); + break; + case SWITCHTEC_IOCTL_PFF_TO_PORT: + rc = ioctl_pff_to_port(stdev, argp); + break; + case SWITCHTEC_IOCTL_PORT_TO_PFF: + rc = ioctl_port_to_pff(stdev, argp); + break; + default: + rc = -ENOTTY; + break; + } + + mutex_unlock(&stdev->mrpc_mutex); + return rc; +} + static const struct file_operations switchtec_fops = { .owner = THIS_MODULE, .open = switchtec_dev_open, @@ -785,6 +1212,8 @@ static const struct file_operations switchtec_fops = { .write = switchtec_dev_write, .read = switchtec_dev_read, .poll = switchtec_dev_poll, + .unlocked_ioctl = switchtec_dev_ioctl, + .compat_ioctl = switchtec_dev_ioctl, }; static void stdev_release(struct device *dev) @@ -871,11 +1300,52 @@ err_put: return ERR_PTR(rc); } +static int mask_event(struct switchtec_dev *stdev, int eid, int idx) +{ + size_t off = event_regs[eid].offset; + u32 __iomem *hdr_reg; + u32 hdr; + + hdr_reg = event_regs[eid].map_reg(stdev, off, idx); + hdr = ioread32(hdr_reg); + + if (!(hdr & SWITCHTEC_EVENT_OCCURRED && hdr & SWITCHTEC_EVENT_EN_IRQ)) + return 0; + + dev_dbg(&stdev->dev, "%s: %d %d %x\n", __func__, eid, idx, hdr); + hdr &= ~(SWITCHTEC_EVENT_EN_IRQ | SWITCHTEC_EVENT_OCCURRED); + iowrite32(hdr, hdr_reg); + + return 1; +} + +static int mask_all_events(struct switchtec_dev *stdev, int eid) +{ + int idx; + int count = 0; + + if (event_regs[eid].map_reg == part_ev_reg) { + for (idx = 0; idx < stdev->partition_count; idx++) + count += mask_event(stdev, eid, idx); + } else if (event_regs[eid].map_reg == pff_ev_reg) { + for (idx = 0; idx < stdev->pff_csr_count; idx++) { + if (!stdev->pff_local[idx]) + continue; + count += mask_event(stdev, eid, idx); + } + } else { + count += mask_event(stdev, eid, 0); + } + + return count; +} + static irqreturn_t switchtec_event_isr(int irq, void *dev) { struct switchtec_dev *stdev = dev; u32 reg; irqreturn_t ret = IRQ_NONE; + int eid, event_count = 0; reg = ioread32(&stdev->mmio_part_cfg->mrpc_comp_hdr); if (reg & SWITCHTEC_EVENT_OCCURRED) { @@ -885,6 +1355,17 @@ static irqreturn_t switchtec_event_isr(int irq, void *dev) iowrite32(reg, &stdev->mmio_part_cfg->mrpc_comp_hdr); } + for (eid = 0; eid < SWITCHTEC_IOCTL_MAX_EVENTS; eid++) + event_count += mask_all_events(stdev, eid); + + if (event_count) { + atomic_inc(&stdev->event_cnt); + wake_up_interruptible(&stdev->event_wq); + dev_dbg(&stdev->dev, "%s: %d events\n", __func__, + event_count); + return IRQ_HANDLED; + } + return ret; } diff --git a/include/uapi/linux/switchtec_ioctl.h b/include/uapi/linux/switchtec_ioctl.h new file mode 100644 index 000000000000..3e824e1a6495 --- /dev/null +++ b/include/uapi/linux/switchtec_ioctl.h @@ -0,0 +1,132 @@ +/* + * Microsemi Switchtec PCIe Driver + * Copyright (c) 2017, Microsemi Corporation + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + */ + +#ifndef _UAPI_LINUX_SWITCHTEC_IOCTL_H +#define _UAPI_LINUX_SWITCHTEC_IOCTL_H + +#include + +#define SWITCHTEC_IOCTL_PART_CFG0 0 +#define SWITCHTEC_IOCTL_PART_CFG1 1 +#define SWITCHTEC_IOCTL_PART_IMG0 2 +#define SWITCHTEC_IOCTL_PART_IMG1 3 +#define SWITCHTEC_IOCTL_PART_NVLOG 4 +#define SWITCHTEC_IOCTL_PART_VENDOR0 5 +#define SWITCHTEC_IOCTL_PART_VENDOR1 6 +#define SWITCHTEC_IOCTL_PART_VENDOR2 7 +#define SWITCHTEC_IOCTL_PART_VENDOR3 8 +#define SWITCHTEC_IOCTL_PART_VENDOR4 9 +#define SWITCHTEC_IOCTL_PART_VENDOR5 10 +#define SWITCHTEC_IOCTL_PART_VENDOR6 11 +#define SWITCHTEC_IOCTL_PART_VENDOR7 12 +#define SWITCHTEC_IOCTL_NUM_PARTITIONS 13 + +struct switchtec_ioctl_flash_info { + __u64 flash_length; + __u32 num_partitions; + __u32 padding; +}; + +struct switchtec_ioctl_flash_part_info { + __u32 flash_partition; + __u32 address; + __u32 length; + __u32 active; +}; + +struct switchtec_ioctl_event_summary { + __u64 global; + __u64 part_bitmap; + __u32 local_part; + __u32 padding; + __u32 part[48]; + __u32 pff[48]; +}; + +#define SWITCHTEC_IOCTL_EVENT_STACK_ERROR 0 +#define SWITCHTEC_IOCTL_EVENT_PPU_ERROR 1 +#define SWITCHTEC_IOCTL_EVENT_ISP_ERROR 2 +#define SWITCHTEC_IOCTL_EVENT_SYS_RESET 3 +#define SWITCHTEC_IOCTL_EVENT_FW_EXC 4 +#define SWITCHTEC_IOCTL_EVENT_FW_NMI 5 +#define SWITCHTEC_IOCTL_EVENT_FW_NON_FATAL 6 +#define SWITCHTEC_IOCTL_EVENT_FW_FATAL 7 +#define SWITCHTEC_IOCTL_EVENT_TWI_MRPC_COMP 8 +#define SWITCHTEC_IOCTL_EVENT_TWI_MRPC_COMP_ASYNC 9 +#define SWITCHTEC_IOCTL_EVENT_CLI_MRPC_COMP 10 +#define SWITCHTEC_IOCTL_EVENT_CLI_MRPC_COMP_ASYNC 11 +#define SWITCHTEC_IOCTL_EVENT_GPIO_INT 12 +#define SWITCHTEC_IOCTL_EVENT_PART_RESET 13 +#define SWITCHTEC_IOCTL_EVENT_MRPC_COMP 14 +#define SWITCHTEC_IOCTL_EVENT_MRPC_COMP_ASYNC 15 +#define SWITCHTEC_IOCTL_EVENT_DYN_PART_BIND_COMP 16 +#define SWITCHTEC_IOCTL_EVENT_AER_IN_P2P 17 +#define SWITCHTEC_IOCTL_EVENT_AER_IN_VEP 18 +#define SWITCHTEC_IOCTL_EVENT_DPC 19 +#define SWITCHTEC_IOCTL_EVENT_CTS 20 +#define SWITCHTEC_IOCTL_EVENT_HOTPLUG 21 +#define SWITCHTEC_IOCTL_EVENT_IER 22 +#define SWITCHTEC_IOCTL_EVENT_THRESH 23 +#define SWITCHTEC_IOCTL_EVENT_POWER_MGMT 24 +#define SWITCHTEC_IOCTL_EVENT_TLP_THROTTLING 25 +#define SWITCHTEC_IOCTL_EVENT_FORCE_SPEED 26 +#define SWITCHTEC_IOCTL_EVENT_CREDIT_TIMEOUT 27 +#define SWITCHTEC_IOCTL_EVENT_LINK_STATE 28 +#define SWITCHTEC_IOCTL_MAX_EVENTS 29 + +#define SWITCHTEC_IOCTL_EVENT_LOCAL_PART_IDX -1 +#define SWITCHTEC_IOCTL_EVENT_IDX_ALL -2 + +#define SWITCHTEC_IOCTL_EVENT_FLAG_CLEAR (1 << 0) +#define SWITCHTEC_IOCTL_EVENT_FLAG_EN_POLL (1 << 1) +#define SWITCHTEC_IOCTL_EVENT_FLAG_EN_LOG (1 << 2) +#define SWITCHTEC_IOCTL_EVENT_FLAG_EN_CLI (1 << 3) +#define SWITCHTEC_IOCTL_EVENT_FLAG_EN_FATAL (1 << 4) +#define SWITCHTEC_IOCTL_EVENT_FLAG_DIS_POLL (1 << 5) +#define SWITCHTEC_IOCTL_EVENT_FLAG_DIS_LOG (1 << 6) +#define SWITCHTEC_IOCTL_EVENT_FLAG_DIS_CLI (1 << 7) +#define SWITCHTEC_IOCTL_EVENT_FLAG_DIS_FATAL (1 << 8) +#define SWITCHTEC_IOCTL_EVENT_FLAG_UNUSED (~0x1ff) + +struct switchtec_ioctl_event_ctl { + __u32 event_id; + __s32 index; + __u32 flags; + __u32 occurred; + __u32 count; + __u32 data[5]; +}; + +#define SWITCHTEC_IOCTL_PFF_VEP 100 +struct switchtec_ioctl_pff_port { + __u32 pff; + __u32 partition; + __u32 port; +}; + +#define SWITCHTEC_IOCTL_FLASH_INFO \ + _IOR('W', 0x40, struct switchtec_ioctl_flash_info) +#define SWITCHTEC_IOCTL_FLASH_PART_INFO \ + _IOWR('W', 0x41, struct switchtec_ioctl_flash_part_info) +#define SWITCHTEC_IOCTL_EVENT_SUMMARY \ + _IOR('W', 0x42, struct switchtec_ioctl_event_summary) +#define SWITCHTEC_IOCTL_EVENT_CTL \ + _IOWR('W', 0x43, struct switchtec_ioctl_event_ctl) +#define SWITCHTEC_IOCTL_PFF_TO_PORT \ + _IOWR('W', 0x44, struct switchtec_ioctl_pff_port) +#define SWITCHTEC_IOCTL_PORT_TO_PFF \ + _IOWR('W', 0x45, struct switchtec_ioctl_pff_port) + +#endif -- cgit v1.2.3-59-g8ed1b From fce5d57e5754fa35245f7c41e3942040393e2fb1 Mon Sep 17 00:00:00 2001 From: Jarod Wilson Date: Wed, 12 Apr 2017 12:33:04 -0500 Subject: PCI: Add bridge DMA alias quirk for ITE 8893 bridge The ITE 8893 bridge has the same problems as the ITE 8892, which were resulting in crippling an older PCI 1Gbps NIC down to 45Mbps throughput with IOMMU and VT-d enabled. With the patch, this old e1000 goes back up to ~900Mbps. Suggested-by: Alex Williamson Signed-off-by: Jarod Wilson Signed-off-by: Bjorn Helgaas Reviewed-by: Alex Williamson --- drivers/pci/quirks.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f754453fe754..e2e2f7ab1d75 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -3938,6 +3938,8 @@ DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_ASMEDIA, 0x1080, DECLARE_PCI_FIXUP_HEADER(0x10e3, 0x8113, quirk_use_pcie_bridge_dma_alias); /* ITE 8892, https://bugzilla.kernel.org/show_bug.cgi?id=73551 */ DECLARE_PCI_FIXUP_HEADER(0x1283, 0x8892, quirk_use_pcie_bridge_dma_alias); +/* ITE 8893 has the same problem as the 8892 */ +DECLARE_PCI_FIXUP_HEADER(0x1283, 0x8893, quirk_use_pcie_bridge_dma_alias); /* Intel 82801, https://bugzilla.kernel.org/show_bug.cgi?id=44881#c49 */ DECLARE_PCI_FIXUP_HEADER(0x8086, 0x244e, quirk_use_pcie_bridge_dma_alias); -- cgit v1.2.3-59-g8ed1b From 6bccc7f426abd640f08d8c75fb22f99483f201b4 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:50 +0100 Subject: PCI: Fix pci_mmap_fits() for HAVE_PCI_RESOURCE_TO_USER platforms In the PCI_MMAP_PROCFS case when the address being passed by the user is a 'user visible' resource address based on the bus window, and not the actual contents of the resource, that's what we need to be checking it against. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org --- drivers/pci/pci-sysfs.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 25d010d449a3..7ac258fd3c5c 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -985,15 +985,19 @@ void pci_remove_legacy_files(struct pci_bus *b) int pci_mmap_fits(struct pci_dev *pdev, int resno, struct vm_area_struct *vma, enum pci_mmap_api mmap_api) { - unsigned long nr, start, size, pci_start; + unsigned long nr, start, size; + resource_size_t pci_start = 0, pci_end; if (pci_resource_len(pdev, resno) == 0) return 0; nr = vma_pages(vma); start = vma->vm_pgoff; size = ((pci_resource_len(pdev, resno) - 1) >> PAGE_SHIFT) + 1; - pci_start = (mmap_api == PCI_MMAP_PROCFS) ? - pci_resource_start(pdev, resno) >> PAGE_SHIFT : 0; + if (mmap_api == PCI_MMAP_PROCFS) { + pci_resource_to_user(pdev, resno, &pdev->resource[resno], + &pci_start, &pci_end); + pci_start >>= PAGE_SHIFT; + } if (start >= pci_start && start < pci_start + size && start + nr <= pci_start + size) return 1; -- cgit v1.2.3-59-g8ed1b From ffff885832101543c002cef7abcab0fd27a9aee1 Mon Sep 17 00:00:00 2001 From: Jayachandran C Date: Thu, 13 Apr 2017 20:30:44 +0000 Subject: PCI: Add device flag PCI_DEV_FLAGS_BRIDGE_XLATE_ROOT Add a new quirk flag PCI_DEV_FLAGS_BRIDGE_XLATE_ROOT to limit the DMA alias search to go no further than the bridge where the IOMMU unit is attached. The flag will be used to indicate a bridge device which forwards the address translation requests to the IOMMU, i.e., where the interrupt and DMA requests leave the PCIe hierarchy and go into the system blocks. Usually this happens at the PCI RC, so this flag is not needed. But on systems where there are bridges that introduce aliases above the IOMMU, this flag prevents pci_for_each_dma_alias() from generating aliases that the IOMMU will never see. The function pci_for_each_dma_alias() is updated to stop when it see a bridge with this flag set. Link: https://bugzilla.kernel.org/show_bug.cgi?id=195447 Signed-off-by: Jayachandran C Signed-off-by: Bjorn Helgaas Reviewed-by: Robin Murphy Acked-by: David Daney --- drivers/pci/search.c | 4 ++++ include/linux/pci.h | 2 ++ 2 files changed, 6 insertions(+) diff --git a/drivers/pci/search.c b/drivers/pci/search.c index 33e0f033a48e..4c6044ad7368 100644 --- a/drivers/pci/search.c +++ b/drivers/pci/search.c @@ -60,6 +60,10 @@ int pci_for_each_dma_alias(struct pci_dev *pdev, tmp = bus->self; + /* stop at bridge where translation unit is associated */ + if (tmp->dev_flags & PCI_DEV_FLAGS_BRIDGE_XLATE_ROOT) + return ret; + /* * PCIe-to-PCI/X bridges alias transactions from downstream * devices using the subordinate bus number (PCI Express to diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..3f596acc05be 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -178,6 +178,8 @@ enum pci_dev_flags { PCI_DEV_FLAGS_NO_PM_RESET = (__force pci_dev_flags_t) (1 << 7), /* Get VPD from function 0 VPD */ PCI_DEV_FLAGS_VPD_REF_F0 = (__force pci_dev_flags_t) (1 << 8), + /* a non-root bridge where translation occurs, stop alias search here */ + PCI_DEV_FLAGS_BRIDGE_XLATE_ROOT = (__force pci_dev_flags_t) (1 << 9), }; enum pci_irq_reroute_variant { -- cgit v1.2.3-59-g8ed1b From 45a2329367386342d41ea9414c88b023f5a79055 Mon Sep 17 00:00:00 2001 From: Jayachandran C Date: Thu, 13 Apr 2017 20:30:45 +0000 Subject: PCI: Avoid generating invalid ThunderX2 DMA aliases On Cavium ThunderX2 arm64 SoCs (formerly known as Broadcom Vulcan), the PCI topology is slightly unusual. For a multi-node system, it looks like: 00:00.0 PCI bridge to [bus 01-1e] 01:0a.0 PCI-to-PCIe bridge to [bus 02-04] 02:00.0 PCIe Root Port bridge to [bus 03-04] (XLATE_ROOT) 03:00.0 PCIe Endpoint pci_for_each_dma_alias() assumes IOMMU translation is done at the root of the PCI hierarchy. It generates 03:00.0, 01:0a.0, and 00:00.0 as DMA aliases for 03:00.0 because buses 01 and 00 are non-PCIe buses that don't carry the Requester ID. Because the ThunderX2 IOMMU is at 02:00.0, the Requester IDs 01:0a.0 and 00:00.0 are never valid for the endpoint. This quirk stops alias generation at the XLATE_ROOT bridge so we won't generate 01:0a.0 or 00:00.0. The current IOMMU code only maps the last alias (this is a separate bug in itself). Prior to this quirk, we only created IOMMU mappings for the invalid Requester ID 00:00:0, which never matched any DMA transactions. With this quirk, we create IOMMU mappings for a valid Requester ID, which fixes devices with no aliases but leaves devices with aliases still broken. The last alias for the endpoint is also used by the ARM GICv3 MSI-X code. Without this quirk, the GIC Interrupt Translation Tables are setup with the invalid Requester ID, and the MSI-X generated by the device fails to be translated and routed. Link: https://bugzilla.kernel.org/show_bug.cgi?id=195447 Signed-off-by: Jayachandran C Signed-off-by: Bjorn Helgaas Reviewed-by: Robin Murphy Acked-by: David Daney --- drivers/pci/quirks.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index 178242d0d9df..74344db4ed4a 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -3956,6 +3956,20 @@ static void quirk_mic_x200_dma_alias(struct pci_dev *pdev) DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2260, quirk_mic_x200_dma_alias); DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_INTEL, 0x2264, quirk_mic_x200_dma_alias); +/* + * The IOMMU and interrupt controller on Broadcom Vulcan/Cavium ThunderX2 are + * associated not at the root bus, but at a bridge below. This quirk avoids + * generating invalid DMA aliases. + */ +static void quirk_bridge_cavm_thrx2_pcie_root(struct pci_dev *pdev) +{ + pdev->dev_flags |= PCI_DEV_FLAGS_BRIDGE_XLATE_ROOT; +} +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_BROADCOM, 0x9000, + quirk_bridge_cavm_thrx2_pcie_root); +DECLARE_PCI_FIXUP_HEADER(PCI_VENDOR_ID_BROADCOM, 0x9084, + quirk_bridge_cavm_thrx2_pcie_root); + /* * Intersil/Techwell TW686[4589]-based video capture cards have an empty (zero) * class code. Fix it. -- cgit v1.2.3-59-g8ed1b From 24196f0c7d4bba093dfa8074507f31509970319f Mon Sep 17 00:00:00 2001 From: Elena Reshetova Date: Tue, 18 Apr 2017 09:02:48 -0500 Subject: PCI: hv: Convert hv_pci_dev.refs from atomic_t to refcount_t refcount_t type and corresponding API should be used instead of atomic_t when the variable is used as a reference counter. This allows to avoid accidental refcounter overflows that might lead to use-after-free situations. Signed-off-by: Elena Reshetova Signed-off-by: Hans Liljestrand Signed-off-by: Kees Cook Signed-off-by: David Windsor Reviewed-by: Stephen Hemminger --- drivers/pci/host/pci-hyperv.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/pci/host/pci-hyperv.c b/drivers/pci/host/pci-hyperv.c index e73880c5d979..84936383e269 100644 --- a/drivers/pci/host/pci-hyperv.c +++ b/drivers/pci/host/pci-hyperv.c @@ -56,6 +56,7 @@ #include #include #include +#include #include /* @@ -423,7 +424,7 @@ enum hv_pcidev_ref_reason { struct hv_pci_dev { /* List protected by pci_rescan_remove_lock */ struct list_head list_entry; - atomic_t refs; + refcount_t refs; enum hv_pcichild_state state; struct pci_function_description desc; bool reported_missing; @@ -1262,13 +1263,13 @@ static void q_resource_requirements(void *context, struct pci_response *resp, static void get_pcichild(struct hv_pci_dev *hpdev, enum hv_pcidev_ref_reason reason) { - atomic_inc(&hpdev->refs); + refcount_inc(&hpdev->refs); } static void put_pcichild(struct hv_pci_dev *hpdev, enum hv_pcidev_ref_reason reason) { - if (atomic_dec_and_test(&hpdev->refs)) + if (refcount_dec_and_test(&hpdev->refs)) kfree(hpdev); } @@ -1322,7 +1323,7 @@ static struct hv_pci_dev *new_pcichild_device(struct hv_pcibus_device *hbus, wait_for_completion(&comp_pkt.host_event); hpdev->desc = *desc; - get_pcichild(hpdev, hv_pcidev_ref_initial); + refcount_set(&hpdev->refs, 1); get_pcichild(hpdev, hv_pcidev_ref_childlist); spin_lock_irqsave(&hbus->device_list_lock, flags); -- cgit v1.2.3-59-g8ed1b From 17caf56731311c9596e7d38a70c88fcb6afa6a1b Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:51 +0100 Subject: PCI: Fix another sanity check bug in /proc/pci mmap Don't match MMIO maps with I/O BARs and vice versa. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org --- drivers/pci/proc.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index f82710a8694d..62a0c3e788d3 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -231,14 +231,20 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma) { struct pci_dev *dev = PDE_DATA(file_inode(file)); struct pci_filp_private *fpriv = file->private_data; - int i, ret, write_combine; + int i, ret, write_combine, res_bit; if (!capable(CAP_SYS_RAWIO)) return -EPERM; + if (fpriv->mmap_state == pci_mmap_io) + res_bit = IORESOURCE_IO; + else + res_bit = IORESOURCE_MEM; + /* Make sure the caller is mapping a real resource for this device */ for (i = 0; i < PCI_ROM_RESOURCE; i++) { - if (pci_mmap_fits(dev, i, vma, PCI_MMAP_PROCFS)) + if (dev->resource[i].flags & res_bit && + pci_mmap_fits(dev, i, vma, PCI_MMAP_PROCFS)) break; } -- cgit v1.2.3-59-g8ed1b From cef4d02305a06be581bb7f4353446717a1b319ec Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:52 +0100 Subject: PCI: Only allow WC mmap on prefetchable resources The /proc/bus/pci mmap interface allows the user to specify whether they want WC or not. Don't let them do so on non-prefetchable BARs. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas Cc: stable@vger.kernel.org --- drivers/pci/proc.c | 13 ++++++++----- 1 file changed, 8 insertions(+), 5 deletions(-) diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index 62a0c3e788d3..dc8912e2d4a1 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -231,7 +231,7 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma) { struct pci_dev *dev = PDE_DATA(file_inode(file)); struct pci_filp_private *fpriv = file->private_data; - int i, ret, write_combine, res_bit; + int i, ret, write_combine = 0, res_bit; if (!capable(CAP_SYS_RAWIO)) return -EPERM; @@ -251,10 +251,13 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma) if (i >= PCI_ROM_RESOURCE) return -ENODEV; - if (fpriv->mmap_state == pci_mmap_mem) - write_combine = fpriv->write_combine; - else - write_combine = 0; + if (fpriv->mmap_state == pci_mmap_mem && + fpriv->write_combine) { + if (dev->resource[i].flags & IORESOURCE_PREFETCH) + write_combine = 1; + else + return -EINVAL; + } ret = pci_mmap_page_range(dev, vma, fpriv->mmap_state, write_combine); if (ret < 0) -- cgit v1.2.3-59-g8ed1b From 03a064b431eb5cb0a91012699ac1e4d6302b327d Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:53 +0100 Subject: xtensa/PCI: Do not mmap PCI BARs to userspace as write-through These should be uncached, not write-through. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/xtensa/kernel/pci.c | 21 +-------------------- 1 file changed, 1 insertion(+), 20 deletions(-) diff --git a/arch/xtensa/kernel/pci.c b/arch/xtensa/kernel/pci.c index b848cc3dc913..5b73fc2f076c 100644 --- a/arch/xtensa/kernel/pci.c +++ b/arch/xtensa/kernel/pci.c @@ -333,25 +333,6 @@ __pci_mmap_make_offset(struct pci_dev *dev, struct vm_area_struct *vma, return -EINVAL; } -/* - * Set vm_page_prot of VMA, as appropriate for this architecture, for a pci - * device mapping. - */ -static __inline__ void -__pci_mmap_set_pgprot(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - int prot = pgprot_val(vma->vm_page_prot); - - /* Set to write-through */ - prot = (prot & _PAGE_CA_MASK) | _PAGE_CA_WT; -#if 0 - if (!write_combine) - prot |= _PAGE_WRITETHRU; -#endif - vma->vm_page_prot = __pgprot(prot); -} - /* * Perform the actual remap of the pages for a PCI device mapping, as * appropriate for this architecture. The region in the process to map @@ -372,7 +353,7 @@ int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, if (ret < 0) return ret; - __pci_mmap_set_pgprot(dev, vma, mmap_state, write_combine); + vma->vm_page_prot = pgprot_device(vma->vm_page_prot); ret = io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, vma->vm_end - vma->vm_start,vma->vm_page_prot); -- cgit v1.2.3-59-g8ed1b From ae749c7ab475de2c9c427058db19921c91846e89 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:54 +0100 Subject: PCI: Add arch_can_pci_mmap_wc() macro Most of the almost-identical versions of pci_mmap_page_range() silently ignore the 'write_combine' argument and give uncached mappings. Yet we allow the PCIIOC_WRITE_COMBINE ioctl in /proc/bus/pci, expose the 'resourceX_wc' file in sysfs, and allow an attempted mapping to apparently succeed. To fix this, introduce a macro arch_can_pci_mmap_wc() which indicates whether the platform can do a write-combining mapping. On x86 this ends up being pat_enabled(), while the few other platforms that support it can just set it to a literal '1'. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- Documentation/filesystems/sysfs-pci.txt | 4 ++++ arch/ia64/include/asm/pci.h | 2 ++ arch/powerpc/include/asm/pci.h | 5 +++-- arch/x86/include/asm/pci.h | 2 ++ drivers/pci/pci-sysfs.c | 4 ++-- drivers/pci/proc.c | 15 ++++++++------- include/linux/pci.h | 4 ++++ 7 files changed, 25 insertions(+), 11 deletions(-) diff --git a/Documentation/filesystems/sysfs-pci.txt b/Documentation/filesystems/sysfs-pci.txt index 6ea1ceda6f52..25b7f1c1b868 100644 --- a/Documentation/filesystems/sysfs-pci.txt +++ b/Documentation/filesystems/sysfs-pci.txt @@ -117,6 +117,10 @@ code must define HAVE_PCI_MMAP and provide a pci_mmap_page_range function. Platforms are free to only support subsets of the mmap functionality, but useful return codes should be provided. +Platforms which support write-combining maps of PCI resources must define +arch_can_pci_mmap_wc() which shall evaluate to non-zero at runtime when +write-combining is permitted. + Legacy resources are protected by the HAVE_PCI_LEGACY define. Platforms wishing to support legacy functionality should define it and provide pci_legacy_read, pci_legacy_write and pci_mmap_legacy_page_range functions. diff --git a/arch/ia64/include/asm/pci.h b/arch/ia64/include/asm/pci.h index c0835b0dc722..6283758474ad 100644 --- a/arch/ia64/include/asm/pci.h +++ b/arch/ia64/include/asm/pci.h @@ -51,6 +51,8 @@ extern unsigned long ia64_max_iommu_merge_mask; #define PCI_DMA_BUS_IS_PHYS (ia64_max_iommu_merge_mask == ~0UL) #define HAVE_PCI_MMAP +#define arch_can_pci_mmap_wc() 1 + extern int pci_mmap_page_range (struct pci_dev *dev, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine); #define HAVE_PCI_LEGACY diff --git a/arch/powerpc/include/asm/pci.h b/arch/powerpc/include/asm/pci.h index 93eded8d3843..b5b68c6a10b1 100644 --- a/arch/powerpc/include/asm/pci.h +++ b/arch/powerpc/include/asm/pci.h @@ -81,8 +81,9 @@ struct vm_area_struct; int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine); -/* Tell drivers/pci/proc.c that we have pci_mmap_page_range() */ -#define HAVE_PCI_MMAP 1 +/* Tell drivers/pci/proc.c that we have pci_mmap_page_range() and it does WC */ +#define HAVE_PCI_MMAP 1 +#define arch_can_pci_mmap_wc() 1 extern int pci_legacy_read(struct pci_bus *bus, loff_t port, u32 *val, size_t count); diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h index 1411dbed5e5e..f6e22c271efa 100644 --- a/arch/x86/include/asm/pci.h +++ b/arch/x86/include/asm/pci.h @@ -7,6 +7,7 @@ #include #include #include +#include #include #ifdef __KERNEL__ @@ -102,6 +103,7 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); #define HAVE_PCI_MMAP +#define arch_can_pci_mmap_wc() pat_enabled() extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine); diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 7ac258fd3c5c..7d494bd66a97 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -1211,9 +1211,9 @@ static int pci_create_resource_files(struct pci_dev *pdev) retval = pci_create_attr(pdev, i, 0); /* for prefetchable resources, create a WC mappable file */ - if (!retval && pdev->resource[i].flags & IORESOURCE_PREFETCH) + if (!retval && arch_can_pci_mmap_wc() && + pdev->resource[i].flags & IORESOURCE_PREFETCH) retval = pci_create_attr(pdev, i, 1); - if (retval) { pci_remove_resource_files(pdev); return retval; diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index dc8912e2d4a1..a2aa58a8fb96 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -210,14 +210,15 @@ static long proc_bus_pci_ioctl(struct file *file, unsigned int cmd, break; case PCIIOC_WRITE_COMBINE: - if (arg) - fpriv->write_combine = 1; - else - fpriv->write_combine = 0; - break; - + if (arch_can_pci_mmap_wc()) { + if (arg) + fpriv->write_combine = 1; + else + fpriv->write_combine = 0; + break; + } + /* If arch decided it can't, fall through... */ #endif /* HAVE_PCI_MMAP */ - default: ret = -EINVAL; break; diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..e614fb42d8bb 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1626,6 +1626,10 @@ static inline int pci_get_new_domain_nr(void) { return -ENOSYS; } #include +#ifndef arch_can_pci_mmap_wc +#define arch_can_pci_mmap_wc() 0 +#endif + #ifndef pci_root_bus_fwnode #define pci_root_bus_fwnode(bus) NULL #endif -- cgit v1.2.3-59-g8ed1b From 11df19546fe4a6135cdae62e96a1e25b3fabf6ea Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:55 +0100 Subject: PCI: Move multiple declarations of pci_mmap_page_range() to We can declare it even on platforms where it isn't going to be defined. There's no need to have it littered through the various files. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/arm/include/asm/pci.h | 2 -- arch/cris/include/asm/pci.h | 3 --- arch/ia64/include/asm/pci.h | 2 -- arch/microblaze/include/asm/pci.h | 3 --- arch/mips/include/asm/pci.h | 3 --- arch/mn10300/include/asm/pci.h | 3 --- arch/parisc/include/asm/pci.h | 3 --- arch/powerpc/include/asm/pci.h | 3 --- arch/sh/include/asm/pci.h | 3 +-- arch/sparc/include/asm/pci_64.h | 4 ---- arch/unicore32/include/asm/pci.h | 2 -- arch/x86/include/asm/pci.h | 4 ---- arch/xtensa/include/asm/pci.h | 4 ---- include/linux/pci.h | 7 +++++++ 14 files changed, 8 insertions(+), 38 deletions(-) diff --git a/arch/arm/include/asm/pci.h b/arch/arm/include/asm/pci.h index 057d381f4e57..51118a0cabbc 100644 --- a/arch/arm/include/asm/pci.h +++ b/arch/arm/include/asm/pci.h @@ -29,8 +29,6 @@ static inline int pci_proc_domain(struct pci_bus *bus) #define PCI_DMA_BUS_IS_PHYS (1) #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel) { diff --git a/arch/cris/include/asm/pci.h b/arch/cris/include/asm/pci.h index b1b289df04c7..65198cb46c85 100644 --- a/arch/cris/include/asm/pci.h +++ b/arch/cris/include/asm/pci.h @@ -42,9 +42,6 @@ struct pci_dev; #define PCI_DMA_BUS_IS_PHYS (1) #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); - #endif /* __KERNEL__ */ diff --git a/arch/ia64/include/asm/pci.h b/arch/ia64/include/asm/pci.h index 6283758474ad..fc60b3d139f4 100644 --- a/arch/ia64/include/asm/pci.h +++ b/arch/ia64/include/asm/pci.h @@ -53,8 +53,6 @@ extern unsigned long ia64_max_iommu_merge_mask; #define HAVE_PCI_MMAP #define arch_can_pci_mmap_wc() 1 -extern int pci_mmap_page_range (struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); #define HAVE_PCI_LEGACY extern int pci_mmap_legacy_page_range(struct pci_bus *bus, struct vm_area_struct *vma, diff --git a/arch/microblaze/include/asm/pci.h b/arch/microblaze/include/asm/pci.h index 2a120bb70e54..ab381d283fb9 100644 --- a/arch/microblaze/include/asm/pci.h +++ b/arch/microblaze/include/asm/pci.h @@ -46,9 +46,6 @@ extern int pci_domain_nr(struct pci_bus *bus); extern int pci_proc_domain(struct pci_bus *bus); struct vm_area_struct; -/* Map a range of PCI memory or I/O space for a device into user space */ -int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() */ #define HAVE_PCI_MMAP 1 diff --git a/arch/mips/include/asm/pci.h b/arch/mips/include/asm/pci.h index 30d1129d8624..3141e2a8113a 100644 --- a/arch/mips/include/asm/pci.h +++ b/arch/mips/include/asm/pci.h @@ -111,9 +111,6 @@ extern void pcibios_set_master(struct pci_dev *dev); #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); - #define HAVE_ARCH_PCI_RESOURCE_TO_USER /* diff --git a/arch/mn10300/include/asm/pci.h b/arch/mn10300/include/asm/pci.h index 51159fff025a..082b6de90936 100644 --- a/arch/mn10300/include/asm/pci.h +++ b/arch/mn10300/include/asm/pci.h @@ -74,9 +74,6 @@ static inline int pci_controller_num(struct pci_dev *dev) } #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, - int write_combine); #endif /* __KERNEL__ */ diff --git a/arch/parisc/include/asm/pci.h b/arch/parisc/include/asm/pci.h index defebd956585..bb9ea9003e08 100644 --- a/arch/parisc/include/asm/pci.h +++ b/arch/parisc/include/asm/pci.h @@ -201,7 +201,4 @@ static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel) #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); - #endif /* __ASM_PARISC_PCI_H */ diff --git a/arch/powerpc/include/asm/pci.h b/arch/powerpc/include/asm/pci.h index b5b68c6a10b1..55887d171205 100644 --- a/arch/powerpc/include/asm/pci.h +++ b/arch/powerpc/include/asm/pci.h @@ -77,9 +77,6 @@ extern int pci_domain_nr(struct pci_bus *bus); extern int pci_proc_domain(struct pci_bus *bus); struct vm_area_struct; -/* Map a range of PCI memory or I/O space for a device into user space */ -int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() and it does WC */ #define HAVE_PCI_MMAP 1 diff --git a/arch/sh/include/asm/pci.h b/arch/sh/include/asm/pci.h index 644314f2b1ef..46abbc9983c9 100644 --- a/arch/sh/include/asm/pci.h +++ b/arch/sh/include/asm/pci.h @@ -66,8 +66,7 @@ extern unsigned long PCIBIOS_MIN_IO, PCIBIOS_MIN_MEM; struct pci_dev; #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); + extern void pcibios_set_master(struct pci_dev *dev); /* Dynamic DMA mapping stuff. diff --git a/arch/sparc/include/asm/pci_64.h b/arch/sparc/include/asm/pci_64.h index 2303635158f5..516fda70a1bb 100644 --- a/arch/sparc/include/asm/pci_64.h +++ b/arch/sparc/include/asm/pci_64.h @@ -45,10 +45,6 @@ static inline int pci_proc_domain(struct pci_bus *bus) #define HAVE_ARCH_PCI_GET_UNMAPPED_AREA #define get_pci_unmapped_area get_fb_unmapped_area -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, - int write_combine); - static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel) { return PCI_IRQ_NONE; diff --git a/arch/unicore32/include/asm/pci.h b/arch/unicore32/include/asm/pci.h index 37e55d018de5..a51290862acd 100644 --- a/arch/unicore32/include/asm/pci.h +++ b/arch/unicore32/include/asm/pci.h @@ -17,8 +17,6 @@ #include /* for PCIBIOS_MIN_* */ #define HAVE_PCI_MMAP -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); #endif /* __KERNEL__ */ #endif diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h index f6e22c271efa..734cc9411b3a 100644 --- a/arch/x86/include/asm/pci.h +++ b/arch/x86/include/asm/pci.h @@ -104,10 +104,6 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); #define HAVE_PCI_MMAP #define arch_can_pci_mmap_wc() pat_enabled() -extern int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, - int write_combine); - #ifdef CONFIG_PCI extern void early_quirks(void); diff --git a/arch/xtensa/include/asm/pci.h b/arch/xtensa/include/asm/pci.h index 5d6bd932ba4e..bb5510b28754 100644 --- a/arch/xtensa/include/asm/pci.h +++ b/arch/xtensa/include/asm/pci.h @@ -46,10 +46,6 @@ struct pci_dev; #define PCI_DMA_BUS_IS_PHYS (1) -/* Map a range of PCI memory or I/O space for a device into user space */ -int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine); - /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() */ #define HAVE_PCI_MMAP 1 diff --git a/include/linux/pci.h b/include/linux/pci.h index e614fb42d8bb..e7bb4b62cc97 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1626,6 +1626,13 @@ static inline int pci_get_new_domain_nr(void) { return -ENOSYS; } #include +/* Map a range of PCI memory or I/O space for a device into user space. + * Architectures provide this function if they set HAVE_PCI_MMAP, and + * it accepts the 'write_combine' argument when arch_can_pci_mmap_wc() + * evaluates to nonzero. */ +int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine); + #ifndef arch_can_pci_mmap_wc #define arch_can_pci_mmap_wc() 0 #endif -- cgit v1.2.3-59-g8ed1b From e854d8b2a82ef76521ad2bed68211fde0511d417 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:56 +0100 Subject: PCI: Add arch_can_pci_mmap_io() on architectures which can mmap() I/O space This is relatively esoteric, and knowing that we don't have it makes life easier in some cases rather than just an eventual -EINVAL from pci_mmap_page_range(). Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- Documentation/filesystems/sysfs-pci.txt | 3 ++- arch/microblaze/include/asm/pci.h | 3 ++- arch/powerpc/include/asm/pci.h | 1 + arch/sparc/include/asm/pci_64.h | 1 + arch/xtensa/include/asm/pci.h | 3 ++- drivers/pci/pci-sysfs.c | 13 ++++++++----- drivers/pci/proc.c | 11 +++++++---- include/linux/pci.h | 3 +++ 8 files changed, 26 insertions(+), 12 deletions(-) diff --git a/Documentation/filesystems/sysfs-pci.txt b/Documentation/filesystems/sysfs-pci.txt index 25b7f1c1b868..46b95d82c4fd 100644 --- a/Documentation/filesystems/sysfs-pci.txt +++ b/Documentation/filesystems/sysfs-pci.txt @@ -119,7 +119,8 @@ useful return codes should be provided. Platforms which support write-combining maps of PCI resources must define arch_can_pci_mmap_wc() which shall evaluate to non-zero at runtime when -write-combining is permitted. +write-combining is permitted. Platforms which support maps of I/O resources +define arch_can_pci_mmap_io() similarly. Legacy resources are protected by the HAVE_PCI_LEGACY define. Platforms wishing to support legacy functionality should define it and provide diff --git a/arch/microblaze/include/asm/pci.h b/arch/microblaze/include/asm/pci.h index ab381d283fb9..efd4983cb697 100644 --- a/arch/microblaze/include/asm/pci.h +++ b/arch/microblaze/include/asm/pci.h @@ -48,7 +48,8 @@ extern int pci_proc_domain(struct pci_bus *bus); struct vm_area_struct; /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() */ -#define HAVE_PCI_MMAP 1 +#define HAVE_PCI_MMAP 1 +#define arch_can_pci_mmap_io() 1 extern int pci_legacy_read(struct pci_bus *bus, loff_t port, u32 *val, size_t count); diff --git a/arch/powerpc/include/asm/pci.h b/arch/powerpc/include/asm/pci.h index 55887d171205..c8975dac535f 100644 --- a/arch/powerpc/include/asm/pci.h +++ b/arch/powerpc/include/asm/pci.h @@ -80,6 +80,7 @@ struct vm_area_struct; /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() and it does WC */ #define HAVE_PCI_MMAP 1 +#define arch_can_pci_mmap_io() 1 #define arch_can_pci_mmap_wc() 1 extern int pci_legacy_read(struct pci_bus *bus, loff_t port, u32 *val, diff --git a/arch/sparc/include/asm/pci_64.h b/arch/sparc/include/asm/pci_64.h index 516fda70a1bb..b957ca5527a3 100644 --- a/arch/sparc/include/asm/pci_64.h +++ b/arch/sparc/include/asm/pci_64.h @@ -42,6 +42,7 @@ static inline int pci_proc_domain(struct pci_bus *bus) /* Platform support for /proc/bus/pci/X/Y mmap()s. */ #define HAVE_PCI_MMAP +#define arch_can_pci_mmap_io() 1 #define HAVE_ARCH_PCI_GET_UNMAPPED_AREA #define get_pci_unmapped_area get_fb_unmapped_area diff --git a/arch/xtensa/include/asm/pci.h b/arch/xtensa/include/asm/pci.h index bb5510b28754..e4f366a488d3 100644 --- a/arch/xtensa/include/asm/pci.h +++ b/arch/xtensa/include/asm/pci.h @@ -47,7 +47,8 @@ struct pci_dev; #define PCI_DMA_BUS_IS_PHYS (1) /* Tell drivers/pci/proc.c that we have pci_mmap_page_range() */ -#define HAVE_PCI_MMAP 1 +#define HAVE_PCI_MMAP 1 +#define arch_can_pci_mmap_io() 1 #endif /* __KERNEL__ */ diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 7d494bd66a97..cb04bc29943f 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -1174,11 +1174,14 @@ static int pci_create_attr(struct pci_dev *pdev, int num, int write_combine) } else { pdev->res_attr[num] = res_attr; sprintf(res_attr_name, "resource%d", num); - res_attr->mmap = pci_mmap_resource_uc; - } - if (pci_resource_flags(pdev, num) & IORESOURCE_IO) { - res_attr->read = pci_read_resource_io; - res_attr->write = pci_write_resource_io; + if (pci_resource_flags(pdev, num) & IORESOURCE_IO) { + res_attr->read = pci_read_resource_io; + res_attr->write = pci_write_resource_io; + if (arch_can_pci_mmap_io()) + res_attr->mmap = pci_mmap_resource_uc; + } else { + res_attr->mmap = pci_mmap_resource_uc; + } } res_attr->attr.name = res_attr_name; res_attr->attr.mode = S_IRUSR | S_IWUSR; diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index a2aa58a8fb96..45e5cf7e9193 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -202,6 +202,8 @@ static long proc_bus_pci_ioctl(struct file *file, unsigned int cmd, #ifdef HAVE_PCI_MMAP case PCIIOC_MMAP_IS_IO: + if (!arch_can_pci_mmap_io()) + return -EINVAL; fpriv->mmap_state = pci_mmap_io; break; @@ -232,15 +234,16 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma) { struct pci_dev *dev = PDE_DATA(file_inode(file)); struct pci_filp_private *fpriv = file->private_data; - int i, ret, write_combine = 0, res_bit; + int i, ret, write_combine = 0, res_bit = IORESOURCE_MEM; if (!capable(CAP_SYS_RAWIO)) return -EPERM; - if (fpriv->mmap_state == pci_mmap_io) + if (fpriv->mmap_state == pci_mmap_io) { + if (!arch_can_pci_mmap_io()) + return -EINVAL; res_bit = IORESOURCE_IO; - else - res_bit = IORESOURCE_MEM; + } /* Make sure the caller is mapping a real resource for this device */ for (i = 0; i < PCI_ROM_RESOURCE; i++) { diff --git a/include/linux/pci.h b/include/linux/pci.h index e7bb4b62cc97..590cfcf6acf5 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1636,6 +1636,9 @@ int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, #ifndef arch_can_pci_mmap_wc #define arch_can_pci_mmap_wc() 0 #endif +#ifndef arch_can_pci_mmap_io +#define arch_can_pci_mmap_io() 0 +#endif #ifndef pci_root_bus_fwnode #define pci_root_bus_fwnode(bus) NULL -- cgit v1.2.3-59-g8ed1b From a7e60e55d73c39df7bcfedb6ccf9b6b1100d960d Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 13 Apr 2017 09:06:40 +0200 Subject: genirq: Fix indentation in remove_irq() Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Reviewed-by: Thomas Gleixner --- kernel/irq/manage.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index a4afe5cc5af1..391cb738b2db 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -1557,7 +1557,7 @@ void remove_irq(unsigned int irq, struct irqaction *act) struct irq_desc *desc = irq_to_desc(irq); if (desc && !WARN_ON(irq_settings_is_per_cpu_devid(desc))) - __free_irq(irq, act->dev_id); + __free_irq(irq, act->dev_id); } EXPORT_SYMBOL_GPL(remove_irq); -- cgit v1.2.3-59-g8ed1b From 25ce4be72411867e0471908ee9319599035cc624 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 13 Apr 2017 09:06:41 +0200 Subject: genirq: Return the IRQ name from free_irq() This allows callers to get back at them instead of having to store it in another variable. Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Reviewed-by: Thomas Gleixner --- include/linux/interrupt.h | 2 +- kernel/irq/manage.c | 13 ++++++++++--- 2 files changed, 11 insertions(+), 4 deletions(-) diff --git a/include/linux/interrupt.h b/include/linux/interrupt.h index 53144e78a369..a6fba4804672 100644 --- a/include/linux/interrupt.h +++ b/include/linux/interrupt.h @@ -155,7 +155,7 @@ extern int __must_check request_percpu_irq(unsigned int irq, irq_handler_t handler, const char *devname, void __percpu *percpu_dev_id); -extern void free_irq(unsigned int, void *); +extern const void *free_irq(unsigned int, void *); extern void free_percpu_irq(unsigned int, void __percpu *); struct device; diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index 391cb738b2db..e688e7e06772 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -1574,20 +1574,27 @@ EXPORT_SYMBOL_GPL(remove_irq); * have completed. * * This function must not be called from interrupt context. + * + * Returns the devname argument passed to request_irq. */ -void free_irq(unsigned int irq, void *dev_id) +const void *free_irq(unsigned int irq, void *dev_id) { struct irq_desc *desc = irq_to_desc(irq); + struct irqaction *action; + const char *devname; if (!desc || WARN_ON(irq_settings_is_per_cpu_devid(desc))) - return; + return NULL; #ifdef CONFIG_SMP if (WARN_ON(desc->affinity_notify)) desc->affinity_notify = NULL; #endif - kfree(__free_irq(irq, dev_id)); + action = __free_irq(irq, dev_id); + devname = action->name; + kfree(action); + return devname; } EXPORT_SYMBOL(free_irq); -- cgit v1.2.3-59-g8ed1b From 704e8953d3e9db29d5d93c0bf6973d86fe15e679 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 13 Apr 2017 09:06:42 +0200 Subject: PCI/irq: Add pci_request_irq() and pci_free_irq() helpers These are small wrappers around request_threaded_irq() and free_irq(), which dynamically allocate space for the device name so that drivers don't need to keep static buffers for these around. Additionally it works with device-relative vector numbers to make the usage easier, and force the IRQF_SHARED flag on given that it has no runtime overhead and should be supported by all PCI devices. Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Reviewed-by: Thomas Gleixner --- drivers/pci/irq.c | 61 ++++++++++++++++++++++++++++++++++++++++++++++++++++- include/linux/pci.h | 6 ++++++ 2 files changed, 66 insertions(+), 1 deletion(-) diff --git a/drivers/pci/irq.c b/drivers/pci/irq.c index 6684f153ab57..f1e46d255c5f 100644 --- a/drivers/pci/irq.c +++ b/drivers/pci/irq.c @@ -1,7 +1,8 @@ /* - * PCI IRQ failure handing code + * PCI IRQ handling code * * Copyright (c) 2008 James Bottomley + * Copyright (C) 2017 Christoph Hellwig. */ #include @@ -59,3 +60,61 @@ enum pci_lost_interrupt_reason pci_lost_interrupt(struct pci_dev *pdev) return PCI_LOST_IRQ_NO_INFORMATION; } EXPORT_SYMBOL(pci_lost_interrupt); + +/** + * pci_request_irq - allocate an interrupt line for a PCI device + * @dev: PCI device to operate on + * @nr: device-relative interrupt vector index (0-based). + * @handler: Function to be called when the IRQ occurs. + * Primary handler for threaded interrupts. + * If NULL and thread_fn != NULL the default primary handler is + * installed. + * @thread_fn: Function called from the IRQ handler thread + * If NULL, no IRQ thread is created + * @dev_id: Cookie passed back to the handler function + * @fmt: Printf-like format string naming the handler + * + * This call allocates interrupt resources and enables the interrupt line and + * IRQ handling. From the point this call is made @handler and @thread_fn may + * be invoked. All interrupts requested using this function might be shared. + * + * @dev_id must not be NULL and must be globally unique. + */ +int pci_request_irq(struct pci_dev *dev, unsigned int nr, irq_handler_t handler, + irq_handler_t thread_fn, void *dev_id, const char *fmt, ...) +{ + va_list ap; + int ret; + char *devname; + + va_start(ap, fmt); + devname = kvasprintf(GFP_KERNEL, fmt, ap); + va_end(ap); + + ret = request_threaded_irq(pci_irq_vector(dev, nr), handler, thread_fn, + IRQF_SHARED, devname, dev_id); + if (ret) + kfree(devname); + return ret; +} +EXPORT_SYMBOL(pci_request_irq); + +/** + * pci_free_irq - free an interrupt allocated with pci_request_irq + * @dev: PCI device to operate on + * @nr: device-relative interrupt vector index (0-based). + * @dev_id: Device identity to free + * + * Remove an interrupt handler. The handler is removed and if the interrupt + * line is no longer in use by any driver it is disabled. The caller must + * ensure the interrupt is disabled on the device before calling this function. + * The function does not return until any executing interrupts for this IRQ + * have completed. + * + * This function must not be called from interrupt context. + */ +void pci_free_irq(struct pci_dev *dev, unsigned int nr, void *dev_id) +{ + kfree(free_irq(pci_irq_vector(dev, nr), dev_id)); +} +EXPORT_SYMBOL(pci_free_irq); diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..b23f81b583ab 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -28,6 +28,7 @@ #include #include #include +#include #include #include #include @@ -1072,6 +1073,11 @@ int pci_select_bars(struct pci_dev *dev, unsigned long flags); bool pci_device_is_present(struct pci_dev *pdev); void pci_ignore_hotplug(struct pci_dev *dev); +int __printf(6, 7) pci_request_irq(struct pci_dev *dev, unsigned int nr, + irq_handler_t handler, irq_handler_t thread_fn, void *dev_id, + const char *fmt, ...); +void pci_free_irq(struct pci_dev *dev, unsigned int nr, void *dev_id); + /* ROM control related routines */ int pci_enable_rom(struct pci_dev *pdev); void pci_disable_rom(struct pci_dev *pdev); -- cgit v1.2.3-59-g8ed1b From 0ff199cb48b4af6f29a1bf15d92d93f44a22eeb4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 13 Apr 2017 09:06:43 +0200 Subject: nvme/pci: Switch to pci_request_irq() Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Reviewed-by: Thomas Gleixner Reviewed-by: Keith Busch --- drivers/nvme/host/pci.c | 30 +++++++++++++----------------- 1 file changed, 13 insertions(+), 17 deletions(-) diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c index 26a5fd05fe88..925997127a6b 100644 --- a/drivers/nvme/host/pci.c +++ b/drivers/nvme/host/pci.c @@ -117,7 +117,6 @@ static inline struct nvme_dev *to_nvme_dev(struct nvme_ctrl *ctrl) struct nvme_queue { struct device *q_dmadev; struct nvme_dev *dev; - char irqname[24]; /* nvme4294967295-65535\0 */ spinlock_t q_lock; struct nvme_command *sq_cmds; struct nvme_command __iomem *sq_cmds_io; @@ -204,11 +203,6 @@ static unsigned int nvme_cmd_size(struct nvme_dev *dev) nvme_iod_alloc_size(dev, NVME_INT_BYTES(dev), NVME_INT_PAGES); } -static int nvmeq_irq(struct nvme_queue *nvmeq) -{ - return pci_irq_vector(to_pci_dev(nvmeq->dev->dev), nvmeq->cq_vector); -} - static int nvme_admin_init_hctx(struct blk_mq_hw_ctx *hctx, void *data, unsigned int hctx_idx) { @@ -962,7 +956,7 @@ static int nvme_suspend_queue(struct nvme_queue *nvmeq) spin_unlock_irq(&nvmeq->q_lock); return 1; } - vector = nvmeq_irq(nvmeq); + vector = nvmeq->cq_vector; nvmeq->dev->online_queues--; nvmeq->cq_vector = -1; spin_unlock_irq(&nvmeq->q_lock); @@ -970,7 +964,7 @@ static int nvme_suspend_queue(struct nvme_queue *nvmeq) if (!nvmeq->qid && nvmeq->dev->ctrl.admin_q) blk_mq_stop_hw_queues(nvmeq->dev->ctrl.admin_q); - free_irq(vector, nvmeq); + pci_free_irq(to_pci_dev(nvmeq->dev->dev), vector, nvmeq); return 0; } @@ -1055,8 +1049,6 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, nvmeq->q_dmadev = dev->dev; nvmeq->dev = dev; - snprintf(nvmeq->irqname, sizeof(nvmeq->irqname), "nvme%dq%d", - dev->ctrl.instance, qid); spin_lock_init(&nvmeq->q_lock); nvmeq->cq_head = 0; nvmeq->cq_phase = 1; @@ -1079,12 +1071,16 @@ static struct nvme_queue *nvme_alloc_queue(struct nvme_dev *dev, int qid, static int queue_request_irq(struct nvme_queue *nvmeq) { - if (use_threaded_interrupts) - return request_threaded_irq(nvmeq_irq(nvmeq), nvme_irq_check, - nvme_irq, IRQF_SHARED, nvmeq->irqname, nvmeq); - else - return request_irq(nvmeq_irq(nvmeq), nvme_irq, IRQF_SHARED, - nvmeq->irqname, nvmeq); + struct pci_dev *pdev = to_pci_dev(nvmeq->dev->dev); + int nr = nvmeq->dev->ctrl.instance; + + if (use_threaded_interrupts) { + return pci_request_irq(pdev, nvmeq->cq_vector, nvme_irq_check, + nvme_irq, nvmeq, "nvme%dq%d", nr, nvmeq->qid); + } else { + return pci_request_irq(pdev, nvmeq->cq_vector, nvme_irq, + NULL, nvmeq, "nvme%dq%d", nr, nvmeq->qid); + } } static void nvme_init_queue(struct nvme_queue *nvmeq, u16 qid) @@ -1440,7 +1436,7 @@ static int nvme_setup_io_queues(struct nvme_dev *dev) } /* Deregister the admin queue's interrupt */ - free_irq(pci_irq_vector(pdev, 0), adminq); + pci_free_irq(pdev, 0, adminq); /* * If we enable msix early due to not intx, disable it again before -- cgit v1.2.3-59-g8ed1b From dc5205ef46231e848562c9275933d7b359f515ba Mon Sep 17 00:00:00 2001 From: Marc Gonzalez Date: Mon, 10 Apr 2017 19:46:54 +0200 Subject: PCI: Improve __pci_read_base() robustness Local variables 'l' and 'sz' are uninitialized. Normally, they would be initialized by pci_read_config_dword() but when an error occurs, some drivers immediately return an error code, which leaves the argument uninitialized. Provide a safe initial value to make the code more robust. Signed-off-by: Marc Gonzalez Signed-off-by: Bjorn Helgaas --- drivers/pci/probe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index dfc9a2794141..f2bd0024cb88 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -175,7 +175,7 @@ static inline unsigned long decode_bar(struct pci_dev *dev, u32 bar) int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type, struct resource *res, unsigned int pos) { - u32 l, sz, mask; + u32 l = 0, sz = 0, mask; u64 l64, sz64, mask64; u16 orig_cmd; struct pci_bus_region region, inverted_region; -- cgit v1.2.3-59-g8ed1b From de5bbdd01cf9ee3cd4586b5a970d3ea015c6d7e3 Mon Sep 17 00:00:00 2001 From: Marc Gonzalez Date: Tue, 18 Apr 2017 14:21:04 -0500 Subject: PCI: Change pci_host_common_probe() visibility pci_host_common_probe() is defined when CONFIG_PCI_HOST_COMMON=y; therefore the function declaration should match that. drivers/pci/host/pcie-tango.c:300:9: error: implicit declaration of function 'pci_host_common_probe' Signed-off-by: Marc Gonzalez Signed-off-by: Bjorn Helgaas --- include/linux/pci-ecam.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/pci-ecam.h b/include/linux/pci-ecam.h index b8f11d783a11..809c2f1873ac 100644 --- a/include/linux/pci-ecam.h +++ b/include/linux/pci-ecam.h @@ -69,7 +69,7 @@ extern struct pci_ecam_ops xgene_v1_pcie_ecam_ops; /* APM X-Gene PCIe v1 */ extern struct pci_ecam_ops xgene_v2_pcie_ecam_ops; /* APM X-Gene PCIe v2.x */ #endif -#ifdef CONFIG_PCI_HOST_GENERIC +#ifdef CONFIG_PCI_HOST_COMMON /* for DT-based PCI controllers that support ECAM */ int pci_host_common_probe(struct platform_device *pdev, struct pci_ecam_ops *ops); -- cgit v1.2.3-59-g8ed1b From 76dc52684d0f72971d9f6cc7d5ae198061b715bd Mon Sep 17 00:00:00 2001 From: Matthias Kaehlcke Date: Fri, 14 Apr 2017 13:38:02 -0700 Subject: PCI: Make PCI_ROM_ADDRESS_MASK a 32-bit constant A 64-bit value is not needed since a PCI ROM address consists in 32 bits. This fixes a clang warning about "implicit conversion from 'unsigned long' to 'u32'". Also remove now unnecessary casts to u32 from __pci_read_base() and pci_std_update_resource(). Signed-off-by: Matthias Kaehlcke Signed-off-by: Bjorn Helgaas --- drivers/pci/probe.c | 2 +- drivers/pci/setup-res.c | 2 +- include/uapi/linux/pci_regs.h | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c index dfc9a2794141..7d5d4a56a186 100644 --- a/drivers/pci/probe.c +++ b/drivers/pci/probe.c @@ -231,7 +231,7 @@ int __pci_read_base(struct pci_dev *dev, enum pci_bar_type type, res->flags |= IORESOURCE_ROM_ENABLE; l64 = l & PCI_ROM_ADDRESS_MASK; sz64 = sz & PCI_ROM_ADDRESS_MASK; - mask64 = (u32)PCI_ROM_ADDRESS_MASK; + mask64 = PCI_ROM_ADDRESS_MASK; } if (res->flags & IORESOURCE_MEM_64) { diff --git a/drivers/pci/setup-res.c b/drivers/pci/setup-res.c index 4bc589ee78d0..85774b7a316a 100644 --- a/drivers/pci/setup-res.c +++ b/drivers/pci/setup-res.c @@ -63,7 +63,7 @@ static void pci_std_update_resource(struct pci_dev *dev, int resno) mask = (u32)PCI_BASE_ADDRESS_IO_MASK; new |= res->flags & ~PCI_BASE_ADDRESS_IO_MASK; } else if (resno == PCI_ROM_RESOURCE) { - mask = (u32)PCI_ROM_ADDRESS_MASK; + mask = PCI_ROM_ADDRESS_MASK; } else { mask = (u32)PCI_BASE_ADDRESS_MEM_MASK; new |= res->flags & ~PCI_BASE_ADDRESS_MEM_MASK; diff --git a/include/uapi/linux/pci_regs.h b/include/uapi/linux/pci_regs.h index 634c9c44ed6c..fff521c9458c 100644 --- a/include/uapi/linux/pci_regs.h +++ b/include/uapi/linux/pci_regs.h @@ -114,7 +114,7 @@ #define PCI_SUBSYSTEM_ID 0x2e #define PCI_ROM_ADDRESS 0x30 /* Bits 31..11 are address, 10..1 reserved */ #define PCI_ROM_ADDRESS_ENABLE 0x01 -#define PCI_ROM_ADDRESS_MASK (~0x7ffUL) +#define PCI_ROM_ADDRESS_MASK (~0x7ffU) #define PCI_CAPABILITY_LIST 0x34 /* Offset of first capability list entry */ -- cgit v1.2.3-59-g8ed1b From ea629d873f3e140fb2e3181c30413e485ee9002b Mon Sep 17 00:00:00 2001 From: Yongji Xie Date: Wed, 15 Feb 2017 14:50:22 +0800 Subject: PCI: Ignore requested alignment for IOV BARs We would call pci_reassigndev_resource_alignment() before pci_init_capabilities(). So the requested alignment would never work for IOV BARs. Furthermore, it's meaningless to request additional alignment for IOV BARs, the IOV BAR alignment is only determined by the VF BAR size. Signed-off-by: Yongji Xie Signed-off-by: Bjorn Helgaas Reviewed-by: Gavin Shan --- drivers/pci/pci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..679af2a253ad 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -5084,7 +5084,7 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) command &= ~PCI_COMMAND_MEMORY; pci_write_config_word(dev, PCI_COMMAND, command); - for (i = 0; i < PCI_BRIDGE_RESOURCES; i++) { + for (i = 0; i <= PCI_ROM_RESOURCE; i++) { r = &dev->resource[i]; if (!(r->flags & IORESOURCE_MEM)) continue; -- cgit v1.2.3-59-g8ed1b From c9c75143a5962c4c26d2f2c99b7a6e06f421f5e1 Mon Sep 17 00:00:00 2001 From: Yongji Xie Date: Mon, 10 Apr 2017 19:58:11 +0800 Subject: PCI: Fix calculation of bridge window's size and alignment In case that one device's alignment is greater than its size, we may get an incorrect size and alignment for its bus's memory window in pbus_size_mem(). Fix this case. Signed-off-by: Yongji Xie Signed-off-by: Bjorn Helgaas --- drivers/pci/setup-bus.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/setup-bus.c b/drivers/pci/setup-bus.c index cb389277df41..958da7db9033 100644 --- a/drivers/pci/setup-bus.c +++ b/drivers/pci/setup-bus.c @@ -1066,10 +1066,10 @@ static int pbus_size_mem(struct pci_bus *bus, unsigned long mask, r->flags = 0; continue; } - size += r_size; + size += max(r_size, align); /* Exclude ranges with size > align from calculation of the alignment. */ - if (r_size == align) + if (r_size <= align) aligns[order] += align; if (order > max_order) max_order = order; -- cgit v1.2.3-59-g8ed1b From ea00353f36b64375518662a8ad15e39218a1f324 Mon Sep 17 00:00:00 2001 From: Lukas Wunner Date: Tue, 18 Apr 2017 20:44:30 +0200 Subject: PCI: Freeze PME scan before suspending devices MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Laurent Pinchart reported that the Renesas R-Car H2 Lager board (r8a7790) crashes during suspend tests. Geert Uytterhoeven managed to reproduce the issue on an M2-W Koelsch board (r8a7791): It occurs when the PME scan runs, once per second. During PME scan, the PCI host bridge (rcar-pci) registers are accessed while its module clock has already been disabled, leading to the crash. One reproducer is to configure s2ram to use "s2idle" instead of "deep" suspend: # echo 0 > /sys/module/printk/parameters/console_suspend # echo s2idle > /sys/power/mem_sleep # echo mem > /sys/power/state Another reproducer is to write either "platform" or "processors" to /sys/power/pm_test. It does not (or is less likely) to happen during full system suspend ("core" or "none") because system suspend also disables timers, and thus the workqueue handling PME scans no longer runs. Geert believes the issue may still happen in the small window between disabling module clocks and disabling timers: # echo 0 > /sys/module/printk/parameters/console_suspend # echo platform > /sys/power/pm_test # Or "processors" # echo mem > /sys/power/state (Make sure CONFIG_PCI_RCAR_GEN2 and CONFIG_USB_OHCI_HCD_PCI are enabled.) Rafael Wysocki agrees that PME scans should be suspended before the host bridge registers become inaccessible. To that end, queue the task on a workqueue that gets frozen before devices suspend. Rafael notes however that as a result, some wakeup events may be missed if they are delivered via PME from a device without working IRQ (which hence must be polled) and occur after the workqueue has been frozen. If that turns out to be an issue in practice, it may be possible to solve it by calling pci_pme_list_scan() once directly from one of the host bridge's pm_ops callbacks. Stacktrace for posterity: PM: Syncing filesystems ... [ 38.566237] done. PM: Preparing system for sleep (mem) Freezing user space processes ... [ 38.579813] (elapsed 0.001 seconds) done. Freezing remaining freezable tasks ... (elapsed 0.001 seconds) done. PM: Suspending system (mem) PM: suspend of devices complete after 152.456 msecs PM: late suspend of devices complete after 2.809 msecs PM: noirq suspend of devices complete after 29.863 msecs suspend debug: Waiting for 5 second(s). Unhandled fault: asynchronous external abort (0x1211) at 0x00000000 pgd = c0003000 [00000000] *pgd=80000040004003, *pmd=00000000 Internal error: : 1211 [#1] SMP ARM Modules linked in: CPU: 1 PID: 20 Comm: kworker/1:1 Not tainted 4.9.0-rc1-koelsch-00011-g68db9bc814362e7f #3383 Hardware name: Generic R8A7791 (Flattened Device Tree) Workqueue: events pci_pme_list_scan task: eb56e140 task.stack: eb58e000 PC is at pci_generic_config_read+0x64/0x6c LR is at rcar_pci_cfg_base+0x64/0x84 pc : [] lr : [] psr: 600d0093 sp : eb58fe98 ip : c041d750 fp : 00000008 r10: c0e2283c r9 : 00000000 r8 : 600d0013 r7 : 00000008 r6 : eb58fed6 r5 : 00000002 r4 : eb58feb4 r3 : 00000000 r2 : 00000044 r1 : 00000008 r0 : 00000000 Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user Control: 30c5387d Table: 6a9f6c80 DAC: 55555555 Process kworker/1:1 (pid: 20, stack limit = 0xeb58e210) Stack: (0xeb58fe98 to 0xeb590000) fe80: 00000002 00000044 fea0: eb6f5800 c041d9b0 eb58feb4 00000008 00000044 00000000 eb78a000 eb78a000 fec0: 00000044 00000000 eb9aff00 c0424bf0 eb78a000 00000000 eb78a000 c0e22830 fee0: ea8a6fc0 c0424c5c eaae79c0 c0424ce0 eb55f380 c0e22838 eb9a9800 c0235fbc ff00: eb55f380 c0e22838 eb55f380 eb9a9800 eb9a9800 eb58e000 eb9a9824 c0e02100 ff20: eb55f398 c02366c4 eb56e140 eb5631c0 00000000 eb55f380 c023641c 00000000 ff40: 00000000 00000000 00000000 c023a928 cd105598 00000000 40506a34 eb55f380 ff60: 00000000 00000000 dead4ead ffffffff ffffffff eb58ff74 eb58ff74 00000000 ff80: 00000000 dead4ead ffffffff ffffffff eb58ff90 eb58ff90 eb58ffac eb5631c0 ffa0: c023a844 00000000 00000000 c0206d68 00000000 00000000 00000000 00000000 ffc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 ffe0: 00000000 00000000 00000000 00000000 00000013 00000000 3a81336c 10ccd1dd [] (pci_generic_config_read) from [] (pci_bus_read_config_word+0x58/0x80) [] (pci_bus_read_config_word) from [] (pci_check_pme_status+0x34/0x78) [] (pci_check_pme_status) from [] (pci_pme_wakeup+0x28/0x54) [] (pci_pme_wakeup) from [] (pci_pme_list_scan+0x58/0xb4) [] (pci_pme_list_scan) from [] (process_one_work+0x1bc/0x308) [] (process_one_work) from [] (worker_thread+0x2a8/0x3e0) [] (worker_thread) from [] (kthread+0xe4/0xfc) [] (kthread) from [] (ret_from_fork+0x14/0x2c) Code: ea000000 e5903000 f57ff04f e3a00000 (e5843000) ---[ end trace 667d43ba3aa9e589 ]--- Fixes: df17e62e5bff ("PCI: Add support for polling PME state on suspended legacy PCI devices") Reported-and-tested-by: Laurent Pinchart Reported-and-tested-by: Geert Uytterhoeven Signed-off-by: Lukas Wunner Signed-off-by: Bjorn Helgaas Reviewed-by: Laurent Pinchart Acked-by: Rafael J. Wysocki Cc: stable@vger.kernel.org # 2.6.37+ Cc: Mika Westerberg Cc: Niklas Söderlund Cc: Simon Horman Cc: Yinghai Lu Cc: Matthew Garrett --- drivers/pci/pci.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 9779483f81fd..6541cba8eef0 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -1784,8 +1784,8 @@ static void pci_pme_list_scan(struct work_struct *work) } } if (!list_empty(&pci_pme_list)) - schedule_delayed_work(&pci_pme_work, - msecs_to_jiffies(PME_TIMEOUT)); + queue_delayed_work(system_freezable_wq, &pci_pme_work, + msecs_to_jiffies(PME_TIMEOUT)); mutex_unlock(&pci_pme_list_mutex); } @@ -1850,8 +1850,9 @@ void pci_pme_active(struct pci_dev *dev, bool enable) mutex_lock(&pci_pme_list_mutex); list_add(&pme_dev->list, &pci_pme_list); if (list_is_singular(&pci_pme_list)) - schedule_delayed_work(&pci_pme_work, - msecs_to_jiffies(PME_TIMEOUT)); + queue_delayed_work(system_freezable_wq, + &pci_pme_work, + msecs_to_jiffies(PME_TIMEOUT)); mutex_unlock(&pci_pme_list_mutex); } else { mutex_lock(&pci_pme_list_mutex); -- cgit v1.2.3-59-g8ed1b From 9b70ae4951746146838b474c3a90722666edf4c1 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 19 Apr 2017 07:44:51 -0500 Subject: PCI: Include PCI-to-PCIe bridges as "Downstream Ports" A PCI/PCI-X to PCI Express bridge, sometimes referred to as a "reverse bridge", is a bridge with conventional PCI or PCI-X on its primary side and a PCI Express Port on its secondary (downstream) side. That PCIe Port is a Downstream Port and could be connected to a slot, just like a Root Port or a Switch Downstream Port. Make pcie_downstream_port() return true for them, so we can access the Slot registers in the PCIe capability. Signed-off-by: Bjorn Helgaas --- drivers/pci/access.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pci/access.c b/drivers/pci/access.c index 1c8051003e67..9b09cd31158c 100644 --- a/drivers/pci/access.c +++ b/drivers/pci/access.c @@ -700,7 +700,8 @@ static bool pcie_downstream_port(const struct pci_dev *dev) int type = pci_pcie_type(dev); return type == PCI_EXP_TYPE_ROOT_PORT || - type == PCI_EXP_TYPE_DOWNSTREAM; + type == PCI_EXP_TYPE_DOWNSTREAM || + type == PCI_EXP_TYPE_PCIE_BRIDGE; } bool pcie_cap_has_lnkctl(const struct pci_dev *dev) -- cgit v1.2.3-59-g8ed1b From 0a701aa6378496ea54fb065c68b41d918e372e94 Mon Sep 17 00:00:00 2001 From: Yongji Xie Date: Mon, 10 Apr 2017 19:58:12 +0800 Subject: PCI: Add pcibios_default_alignment() for arch-specific alignment control When VFIO passes through a PCI device to a guest, it does not allow the guest to mmap BARs that are smaller than PAGE_SIZE unless it can reserve the rest of the page (see vfio_pci_probe_mmaps()). This is because a page might contain several small BARs for unrelated devices and a guest should not be able to access all of them. VFIO emulates guest accesses to non-mappable BARs, which is functional but slow. On systems with large page sizes, e.g., PowerNV with 64K pages, BARs are more likely to share a page and performance is more likely to be a problem. Add a weak function to set default alignment for all PCI devices. An arch can override it to force the PCI core to place memory BARs on their own pages. Signed-off-by: Yongji Xie Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 679af2a253ad..d22fb35be26e 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -4947,6 +4947,11 @@ void pci_ignore_hotplug(struct pci_dev *dev) } EXPORT_SYMBOL_GPL(pci_ignore_hotplug); +resource_size_t __weak pcibios_default_alignment(void) +{ + return 0; +} + #define RESOURCE_ALIGNMENT_PARAM_SIZE COMMAND_LINE_SIZE static char resource_alignment_param[RESOURCE_ALIGNMENT_PARAM_SIZE] = {0}; static DEFINE_SPINLOCK(resource_alignment_lock); @@ -4962,14 +4967,15 @@ static resource_size_t pci_specified_resource_alignment(struct pci_dev *dev) { int seg, bus, slot, func, align_order, count; unsigned short vendor, device, subsystem_vendor, subsystem_device; - resource_size_t align = 0; + resource_size_t align = pcibios_default_alignment(); char *p; spin_lock(&resource_alignment_lock); p = resource_alignment_param; - if (!*p) + if (!*p && !align) goto out; if (pci_has_flag(PCI_PROBE_ONLY)) { + align = 0; pr_info_once("PCI: Ignoring requested alignments (PCI_PROBE_ONLY)\n"); goto out; } -- cgit v1.2.3-59-g8ed1b From 382746376993cfa6d6c4e546c67384201c0f3a82 Mon Sep 17 00:00:00 2001 From: Yongji Xie Date: Mon, 10 Apr 2017 19:58:13 +0800 Subject: powerpc/powernv: Override pcibios_default_alignment() to force PCI devices to be page aligned Override pcibios_default_alignment() to set default alignment to PAGE_SIZE for all PCI devices on PowerNV platform. Thus sub-page BARs would not share a page and could be mapped into guest when VFIO passthrough them. Signed-off-by: Yongji Xie Signed-off-by: Bjorn Helgaas --- arch/powerpc/include/asm/machdep.h | 2 ++ arch/powerpc/kernel/pci-common.c | 8 ++++++++ arch/powerpc/platforms/powernv/pci-ioda.c | 7 +++++++ 3 files changed, 17 insertions(+) diff --git a/arch/powerpc/include/asm/machdep.h b/arch/powerpc/include/asm/machdep.h index 5011b69107a7..f90b22c722e1 100644 --- a/arch/powerpc/include/asm/machdep.h +++ b/arch/powerpc/include/asm/machdep.h @@ -173,6 +173,8 @@ struct machdep_calls { /* Called after scan and before resource survey */ void (*pcibios_fixup_phb)(struct pci_controller *hose); + resource_size_t (*pcibios_default_alignment)(void); + #ifdef CONFIG_PCI_IOV void (*pcibios_fixup_sriov)(struct pci_dev *pdev); resource_size_t (*pcibios_iov_resource_alignment)(struct pci_dev *, int resno); diff --git a/arch/powerpc/kernel/pci-common.c b/arch/powerpc/kernel/pci-common.c index ffda24a38dda..03a7c9f58126 100644 --- a/arch/powerpc/kernel/pci-common.c +++ b/arch/powerpc/kernel/pci-common.c @@ -233,6 +233,14 @@ void pcibios_reset_secondary_bus(struct pci_dev *dev) pci_reset_secondary_bus(dev); } +resource_size_t pcibios_default_alignment(void) +{ + if (ppc_md.pcibios_default_alignment) + return ppc_md.pcibios_default_alignment(); + + return 0; +} + #ifdef CONFIG_PCI_IOV resource_size_t pcibios_iov_resource_alignment(struct pci_dev *pdev, int resno) { diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c index 6901a06da2f9..0529d5630afd 100644 --- a/arch/powerpc/platforms/powernv/pci-ioda.c +++ b/arch/powerpc/platforms/powernv/pci-ioda.c @@ -3287,6 +3287,11 @@ static void pnv_pci_setup_bridge(struct pci_bus *bus, unsigned long type) } } +static resource_size_t pnv_pci_default_alignment(void) +{ + return PAGE_SIZE; +} + #ifdef CONFIG_PCI_IOV static resource_size_t pnv_pci_iov_resource_alignment(struct pci_dev *pdev, int resno) @@ -3820,6 +3825,8 @@ static void __init pnv_pci_init_ioda_phb(struct device_node *np, hose->controller_ops = pnv_pci_ioda_controller_ops; } + ppc_md.pcibios_default_alignment = pnv_pci_default_alignment; + #ifdef CONFIG_PCI_IOV ppc_md.pcibios_fixup_sriov = pnv_pci_ioda_fixup_iov_resources; ppc_md.pcibios_iov_resource_alignment = pnv_pci_iov_resource_alignment; -- cgit v1.2.3-59-g8ed1b From 81a5e70e0de55707c3184d186b2103b5bb9377ec Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Fri, 14 Apr 2017 14:12:06 -0500 Subject: PCI: Factor pci_reassigndev_resource_alignment() Pull the BAR size adjustment out into a new function, pci_request_resource_alignment(), and add a comment about how and why we increase the resource size and alignment. Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 69 +++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 47 insertions(+), 22 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index d22fb35be26e..1d8bb090a40b 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -5049,6 +5049,48 @@ out: return align; } +static void pci_request_resource_alignment(struct pci_dev *dev, int bar, + resource_size_t align) +{ + struct resource *r = &dev->resource[bar]; + resource_size_t size; + + if (!(r->flags & IORESOURCE_MEM)) + return; + + if (r->flags & IORESOURCE_PCI_FIXED) { + dev_info(&dev->dev, "BAR%d %pR: ignoring requested alignment %#llx\n", + bar, r, (unsigned long long)align); + return; + } + + size = resource_size(r); + if (size < align) { + + /* + * Increase the size of the resource. BARs are aligned on + * their size, so when we reallocate space for this + * resource, we'll allocate it with the larger alignment. + * It also prevents assignment of any other BARs inside the + * size. If we're requesting page alignment, this means no + * other BARs will share the page. + * + * This makes the resource larger than the hardware BAR, + * which may break drivers that compute things based on the + * resource size, e.g., to find registers at a fixed offset + * before the end of the BAR. We hope users don't request + * alignment for such devices. + */ + size = align; + dev_info(&dev->dev, "BAR%d %pR: requesting alignment to %#llx\n", + bar, r, (unsigned long long)align); + + } + r->flags |= IORESOURCE_UNSET; + r->end = size - 1; + r->start = 0; +} + /* * This function disables memory decoding and releases memory resources * of the device specified by kernel's boot parameter 'pci=resource_alignment='. @@ -5060,7 +5102,7 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) { int i; struct resource *r; - resource_size_t align, size; + resource_size_t align; u16 command; /* @@ -5090,28 +5132,11 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) command &= ~PCI_COMMAND_MEMORY; pci_write_config_word(dev, PCI_COMMAND, command); - for (i = 0; i <= PCI_ROM_RESOURCE; i++) { - r = &dev->resource[i]; - if (!(r->flags & IORESOURCE_MEM)) - continue; - if (r->flags & IORESOURCE_PCI_FIXED) { - dev_info(&dev->dev, "Ignoring requested alignment for BAR%d: %pR\n", - i, r); - continue; - } + for (i = 0; i <= PCI_ROM_RESOURCE; i++) + pci_request_resource_alignment(dev, i, align); - size = resource_size(r); - if (size < align) { - size = align; - dev_info(&dev->dev, - "Rounding up size of resource #%d to %#llx.\n", - i, (unsigned long long)size); - } - r->flags |= IORESOURCE_UNSET; - r->end = size - 1; - r->start = 0; - } - /* Need to disable bridge's resource window, + /* + * Need to disable bridge's resource window, * to enable the kernel to reassign new resource * window later on. */ -- cgit v1.2.3-59-g8ed1b From 0dde1c08d1b9dea01cefb327dba8a6e3ae795214 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Mon, 17 Apr 2017 15:20:58 -0500 Subject: PCI: Don't reassign resources that are already aligned The "pci=resource_alignment=" kernel argument designates devices for which we want alignment greater than is required by the PCI specs. Previously we set IORESOURCE_UNSET for every MEM resource of those devices, even if the resource was *already* sufficiently aligned. If a resource is already sufficiently aligned, leave it alone and don't try to reassign it. Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 40 +++++++++++++++++++--------------------- 1 file changed, 19 insertions(+), 21 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 1d8bb090a40b..7eb6eb384e36 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -5065,30 +5065,28 @@ static void pci_request_resource_alignment(struct pci_dev *dev, int bar, } size = resource_size(r); - if (size < align) { + if (size >= align) + return; - /* - * Increase the size of the resource. BARs are aligned on - * their size, so when we reallocate space for this - * resource, we'll allocate it with the larger alignment. - * It also prevents assignment of any other BARs inside the - * size. If we're requesting page alignment, this means no - * other BARs will share the page. - * - * This makes the resource larger than the hardware BAR, - * which may break drivers that compute things based on the - * resource size, e.g., to find registers at a fixed offset - * before the end of the BAR. We hope users don't request - * alignment for such devices. - */ - size = align; - dev_info(&dev->dev, "BAR%d %pR: requesting alignment to %#llx\n", - bar, r, (unsigned long long)align); + /* + * Increase the size of the resource. BARs are aligned on their + * size, so when we reallocate space for this resource, we'll + * allocate it with the larger alignment. It also prevents + * assignment of any other BARs inside the size. If we're + * requesting page alignment, this means no other BARs will share + * the page. + * + * This makes the resource larger than the hardware BAR, which may + * break drivers that compute things based on the resource size, + * e.g., to find registers at a fixed offset before the end of the + * BAR. We hope users don't request alignment for such devices. + */ + dev_info(&dev->dev, "BAR%d %pR: requesting alignment to %#llx\n", + bar, r, (unsigned long long)align); - } - r->flags |= IORESOURCE_UNSET; - r->end = size - 1; r->start = 0; + r->end = align - 1; + r->flags |= IORESOURCE_UNSET; } /* -- cgit v1.2.3-59-g8ed1b From e3adec72a3c50b733eb277a9f93c044f59ff55eb Mon Sep 17 00:00:00 2001 From: Yongji Xie Date: Mon, 10 Apr 2017 19:58:14 +0800 Subject: PCI: Don't resize resources when realigning all devices in system The "pci=resource_alignment" argument aligns BARs of designated devices by artificially increasing their size. Increasing the size increases the alignment and prevents other resources from being assigned in the same alignment region, e.g., in the same page, but it can break drivers that use the BAR size to locate things, e.g., ilo_map_device() does this: off = pci_resource_len(pdev, bar) - 0x2000; The new pcibios_default_alignment() interface allows an arch to request that *all* BARs in the system be aligned to a larger size. In this case, we don't need to artificially increase the resource size because we know every BAR of every device will be realigned, so nothing will share the same alignment region. Use IORESOURCE_STARTALIGN to request realignment of PCI BARs when we know we're realigning all BARs in the system. [bhelgaas: comment, changelog] Signed-off-by: Yongji Xie Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 59 ++++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 43 insertions(+), 16 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7eb6eb384e36..c140cb29ee7e 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -4959,11 +4959,13 @@ static DEFINE_SPINLOCK(resource_alignment_lock); /** * pci_specified_resource_alignment - get resource alignment specified by user. * @dev: the PCI device to get + * @resize: whether or not to change resources' size when reassigning alignment * * RETURNS: Resource alignment if it is specified. * Zero if it is not specified. */ -static resource_size_t pci_specified_resource_alignment(struct pci_dev *dev) +static resource_size_t pci_specified_resource_alignment(struct pci_dev *dev, + bool *resize) { int seg, bus, slot, func, align_order, count; unsigned short vendor, device, subsystem_vendor, subsystem_device; @@ -5005,6 +5007,7 @@ static resource_size_t pci_specified_resource_alignment(struct pci_dev *dev) (!device || (device == dev->device)) && (!subsystem_vendor || (subsystem_vendor == dev->subsystem_vendor)) && (!subsystem_device || (subsystem_device == dev->subsystem_device))) { + *resize = true; if (align_order == -1) align = PAGE_SIZE; else @@ -5030,6 +5033,7 @@ static resource_size_t pci_specified_resource_alignment(struct pci_dev *dev) bus == dev->bus->number && slot == PCI_SLOT(dev->devfn) && func == PCI_FUNC(dev->devfn)) { + *resize = true; if (align_order == -1) align = PAGE_SIZE; else @@ -5050,7 +5054,7 @@ out: } static void pci_request_resource_alignment(struct pci_dev *dev, int bar, - resource_size_t align) + resource_size_t align, bool resize) { struct resource *r = &dev->resource[bar]; resource_size_t size; @@ -5069,23 +5073,45 @@ static void pci_request_resource_alignment(struct pci_dev *dev, int bar, return; /* - * Increase the size of the resource. BARs are aligned on their - * size, so when we reallocate space for this resource, we'll - * allocate it with the larger alignment. It also prevents - * assignment of any other BARs inside the size. If we're - * requesting page alignment, this means no other BARs will share - * the page. + * Increase the alignment of the resource. There are two ways we + * can do this: * - * This makes the resource larger than the hardware BAR, which may - * break drivers that compute things based on the resource size, - * e.g., to find registers at a fixed offset before the end of the - * BAR. We hope users don't request alignment for such devices. + * 1) Increase the size of the resource. BARs are aligned on their + * size, so when we reallocate space for this resource, we'll + * allocate it with the larger alignment. This also prevents + * assignment of any other BARs inside the alignment region, so + * if we're requesting page alignment, this means no other BARs + * will share the page. + * + * The disadvantage is that this makes the resource larger than + * the hardware BAR, which may break drivers that compute things + * based on the resource size, e.g., to find registers at a + * fixed offset before the end of the BAR. + * + * 2) Retain the resource size, but use IORESOURCE_STARTALIGN and + * set r->start to the desired alignment. By itself this + * doesn't prevent other BARs being put inside the alignment + * region, but if we realign *every* resource of every device in + * the system, none of them will share an alignment region. + * + * When the user has requested alignment for only some devices via + * the "pci=resource_alignment" argument, "resize" is true and we + * use the first method. Otherwise we assume we're aligning all + * devices and we use the second. */ + dev_info(&dev->dev, "BAR%d %pR: requesting alignment to %#llx\n", bar, r, (unsigned long long)align); - r->start = 0; - r->end = align - 1; + if (resize) { + r->start = 0; + r->end = align - 1; + } else { + r->flags &= ~IORESOURCE_SIZEALIGN; + r->flags |= IORESOURCE_STARTALIGN; + r->start = align; + r->end = r->start + size - 1; + } r->flags |= IORESOURCE_UNSET; } @@ -5102,6 +5128,7 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) struct resource *r; resource_size_t align; u16 command; + bool resize = false; /* * VF BARs are read-only zero according to SR-IOV spec r1.1, sec @@ -5113,7 +5140,7 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) return; /* check if specified PCI is target device to reassign */ - align = pci_specified_resource_alignment(dev); + align = pci_specified_resource_alignment(dev, &resize); if (!align) return; @@ -5131,7 +5158,7 @@ void pci_reassigndev_resource_alignment(struct pci_dev *dev) pci_write_config_word(dev, PCI_COMMAND, command); for (i = 0; i <= PCI_ROM_RESOURCE; i++) - pci_request_resource_alignment(dev, i, align); + pci_request_resource_alignment(dev, i, align, resize); /* * Need to disable bridge's resource window, -- cgit v1.2.3-59-g8ed1b From 7b309aef0463340d3ad5449d1f605d14e10a4225 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:50 +0100 Subject: PCI: Remove __weak tag from pci_remap_iospace() pci_remap_iospace() is marked as a weak symbol even though no architecture is currently overriding it; given that its implementation internals have already code paths that are arch specific (ie PCI_IOBASE and ioremap_page_range() attributes) there is no need to leave the weak symbol in the kernel since the same functionality can be achieved by customizing per-arch the corresponding functionality. Remove the __weak symbol from pci_remap_iospace(). Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Acked-by: Arnd Bergmann --- drivers/pci/pci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..bd98674a0419 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -3363,7 +3363,7 @@ unsigned long __weak pci_address_to_pio(phys_addr_t address) * Only architectures that have memory mapped IO functions defined * (and the PCI_IOBASE value defined) should call this function. */ -int __weak pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr) +int pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr) { #if defined(PCI_IOBASE) && defined(CONFIG_MMU) unsigned long vaddr = (unsigned long)PCI_IOBASE + res->start; -- cgit v1.2.3-59-g8ed1b From cf9ea8ca4a0bea7eda12f8fb04dc34146839a215 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:51 +0100 Subject: linux/io.h: Add pci_remap_cfgspace() interface The PCI specifications (Rev 3.0, 3.2.5 "Transaction Ordering and Posting") mandate non-posted configuration transactions. As further highlighted in the PCIe specifications (4.0 - Rev0.3, "Ordering Considerations for the Enhanced Configuration Access Mechanism"), through ECAM and ECAM-derivative configuration mechanism, the memory mapped transactions from the host CPU into Configuration Requests on the PCI express fabric may create ordering problems for software because writes to memory address are typically posted transactions (unless the architecture can enforce through virtual address mapping non-posted write transactions behaviour) but writes to Configuration Space are not posted on the PCI express fabric. Current DT and ACPI host bridge controllers map PCI configuration space (ECAM and ECAM-derivative) into the virtual address space through ioremap() calls, that are non-cacheable device accesses on most architectures, but may provide "bufferable" or "posted" write semantics in architecture like eg ARM/ARM64 that allow ioremap'ed regions writes to be buffered in the bus connecting the host CPU to the PCI fabric; this behaviour, as underlined in the PCIe specifications, may trigger transactions ordering rules and must be prevented. Introduce a new generic and explicit API to create a memory mapping for ECAM and ECAM-derivative config space area that defaults to ioremap_nocache() (which should provide a sane default behaviour) but still allowing architectures on which ioremap_nocache() results in posted write transactions to override the function call with an arch specific implementation that complies with the PCI specifications for configuration transactions. [bhelgaas: fold in #ifdef CONFIG_PCI wrapper] Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Arnd Bergmann Cc: Will Deacon Cc: Russell King Cc: Catalin Marinas --- include/linux/io.h | 21 +++++++++++++++++++++ 1 file changed, 21 insertions(+) diff --git a/include/linux/io.h b/include/linux/io.h index 82ef36eac8a1..2195d9ea4aaa 100644 --- a/include/linux/io.h +++ b/include/linux/io.h @@ -90,6 +90,27 @@ void devm_memunmap(struct device *dev, void *addr); void *__devm_memremap_pages(struct device *dev, struct resource *res); +#ifdef CONFIG_PCI +/* + * The PCI specifications (Rev 3.0, 3.2.5 "Transaction Ordering and + * Posting") mandate non-posted configuration transactions. There is + * no ioremap API in the kernel that can guarantee non-posted write + * semantics across arches so provide a default implementation for + * mapping PCI config space that defaults to ioremap_nocache(); arches + * should override it if they have memory mapping implementations that + * guarantee non-posted writes semantics to make the memory mapping + * compliant with the PCI specification. + */ +#ifndef pci_remap_cfgspace +#define pci_remap_cfgspace pci_remap_cfgspace +static inline void __iomem *pci_remap_cfgspace(phys_addr_t offset, + size_t size) +{ + return ioremap_nocache(offset, size); +} +#endif +#endif + /* * Some systems do not have legacy ISA devices. * /dev/port is not a valid interface on these systems. -- cgit v1.2.3-59-g8ed1b From dca40b186b757c7f9a05f870f69e0dfaeca59d7b Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:57 +0100 Subject: PCI: Use BAR index in sysfs attr->private instead of resource pointer We store the pointer, and then on *every* use of it we loop over the device's resources to find out the index. That's kind of silly. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- drivers/pci/pci-sysfs.c | 38 ++++++++++++++------------------------ 1 file changed, 14 insertions(+), 24 deletions(-) diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index cb04bc29943f..534844df8bf5 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -1017,26 +1017,20 @@ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, struct vm_area_struct *vma, int write_combine) { struct pci_dev *pdev = to_pci_dev(kobj_to_dev(kobj)); - struct resource *res = attr->private; + int bar = (unsigned long)attr->private; enum pci_mmap_state mmap_type; resource_size_t start, end; - int i; - - for (i = 0; i < PCI_ROM_RESOURCE; i++) - if (res == &pdev->resource[i]) - break; - if (i >= PCI_ROM_RESOURCE) - return -ENODEV; + struct resource *res = &pdev->resource[bar]; if (res->flags & IORESOURCE_MEM && iomem_is_exclusive(res->start)) return -EINVAL; - if (!pci_mmap_fits(pdev, i, vma, PCI_MMAP_SYSFS)) { + if (!pci_mmap_fits(pdev, bar, vma, PCI_MMAP_SYSFS)) { WARN(1, "process \"%s\" tried to map 0x%08lx bytes at page 0x%08lx on %s BAR %d (start 0x%16Lx, size 0x%16Lx)\n", current->comm, vma->vm_end-vma->vm_start, vma->vm_pgoff, - pci_name(pdev), i, - (u64)pci_resource_start(pdev, i), - (u64)pci_resource_len(pdev, i)); + pci_name(pdev), bar, + (u64)pci_resource_start(pdev, bar), + (u64)pci_resource_len(pdev, bar)); return -EINVAL; } @@ -1044,7 +1038,7 @@ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, * from /proc/bus/pci/ which is a "user visible" value. If this is * different from the resource itself, arch will do necessary fixup. */ - pci_resource_to_user(pdev, i, res, &start, &end); + pci_resource_to_user(pdev, bar, res, &start, &end); vma->vm_pgoff += start >> PAGE_SHIFT; mmap_type = res->flags & IORESOURCE_MEM ? pci_mmap_mem : pci_mmap_io; return pci_mmap_page_range(pdev, vma, mmap_type, write_combine); @@ -1069,22 +1063,18 @@ static ssize_t pci_resource_io(struct file *filp, struct kobject *kobj, loff_t off, size_t count, bool write) { struct pci_dev *pdev = to_pci_dev(kobj_to_dev(kobj)); - struct resource *res = attr->private; + int bar = (unsigned long)attr->private; + struct resource *res; unsigned long port = off; - int i; - for (i = 0; i < PCI_ROM_RESOURCE; i++) - if (res == &pdev->resource[i]) - break; - if (i >= PCI_ROM_RESOURCE) - return -ENODEV; + res = &pdev->resource[bar]; - port += pci_resource_start(pdev, i); + port += pci_resource_start(pdev, bar); - if (port > pci_resource_end(pdev, i)) + if (port > pci_resource_end(pdev, bar)) return 0; - if (port + count - 1 > pci_resource_end(pdev, i)) + if (port + count - 1 > pci_resource_end(pdev, bar)) return -EINVAL; switch (count) { @@ -1186,7 +1176,7 @@ static int pci_create_attr(struct pci_dev *pdev, int num, int write_combine) res_attr->attr.name = res_attr_name; res_attr->attr.mode = S_IRUSR | S_IWUSR; res_attr->size = pci_resource_len(pdev, num); - res_attr->private = &pdev->resource[num]; + res_attr->private = (void *)(unsigned long)num; retval = sysfs_create_bin_file(&pdev->dev.kobj, res_attr); if (retval) kfree(res_attr); -- cgit v1.2.3-59-g8ed1b From f66e225828c1b046c7db1db65b0dd2d135f6a2da Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:58 +0100 Subject: PCI: Add BAR index argument to pci_mmap_page_range() In all cases we know which BAR it is. Passing it in means that arch code (or generic code; watch this space) won't have to go looking for it again. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/arm/kernel/bios32.c | 3 ++- arch/cris/arch-v32/drivers/pci/bios.c | 3 ++- arch/ia64/pci/pci.c | 3 ++- arch/microblaze/pci/pci-common.c | 2 +- arch/mips/pci/pci.c | 3 ++- arch/mn10300/unit-asb2305/pci-asb2305.c | 3 ++- arch/parisc/kernel/pci.c | 3 ++- arch/powerpc/kernel/pci-common.c | 3 ++- arch/sh/drivers/pci/pci.c | 3 ++- arch/sparc/kernel/pci.c | 6 +++--- arch/unicore32/kernel/pci.c | 3 ++- arch/x86/pci/i386.c | 3 ++- arch/xtensa/kernel/pci.c | 3 ++- drivers/pci/pci-sysfs.c | 2 +- drivers/pci/proc.c | 2 +- include/linux/pci.h | 3 ++- 16 files changed, 30 insertions(+), 18 deletions(-) diff --git a/arch/arm/kernel/bios32.c b/arch/arm/kernel/bios32.c index 2f0e07735d1d..a4fc3f46eeae 100644 --- a/arch/arm/kernel/bios32.c +++ b/arch/arm/kernel/bios32.c @@ -597,7 +597,8 @@ resource_size_t pcibios_align_resource(void *data, const struct resource *res, return start; } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { if (mmap_state == pci_mmap_io) diff --git a/arch/cris/arch-v32/drivers/pci/bios.c b/arch/cris/arch-v32/drivers/pci/bios.c index 212266a2c5d9..a589686d5448 100644 --- a/arch/cris/arch-v32/drivers/pci/bios.c +++ b/arch/cris/arch-v32/drivers/pci/bios.c @@ -14,7 +14,8 @@ void pcibios_set_master(struct pci_dev *dev) pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat); } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 8f6ac2f8ae4c..053c688b15a5 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -419,7 +419,8 @@ pcibios_align_resource (void *data, const struct resource *res, } int -pci_mmap_page_range (struct pci_dev *dev, struct vm_area_struct *vma, +pci_mmap_page_range (struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long size = vma->vm_end - vma->vm_start; diff --git a/arch/microblaze/pci/pci-common.c b/arch/microblaze/pci/pci-common.c index 13bc93242c0c..404fb38d06b7 100644 --- a/arch/microblaze/pci/pci-common.c +++ b/arch/microblaze/pci/pci-common.c @@ -278,7 +278,7 @@ pgprot_t pci_phys_mem_access_prot(struct file *file, * * Returns a negative error code on failure, zero on success. */ -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { resource_size_t offset = diff --git a/arch/mips/pci/pci.c b/arch/mips/pci/pci.c index f6325fa657fb..f189502041a6 100644 --- a/arch/mips/pci/pci.c +++ b/arch/mips/pci/pci.c @@ -58,7 +58,8 @@ void pci_resource_to_user(const struct pci_dev *dev, int bar, *end = rsrc->start + size; } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; diff --git a/arch/mn10300/unit-asb2305/pci-asb2305.c b/arch/mn10300/unit-asb2305/pci-asb2305.c index b7ab8378964c..4abbbd54ac7d 100644 --- a/arch/mn10300/unit-asb2305/pci-asb2305.c +++ b/arch/mn10300/unit-asb2305/pci-asb2305.c @@ -211,7 +211,8 @@ void __init pcibios_resource_survey(void) pcibios_allocate_resources(1); } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; diff --git a/arch/parisc/kernel/pci.c b/arch/parisc/kernel/pci.c index 0903c6abd7a4..653877553acd 100644 --- a/arch/parisc/kernel/pci.c +++ b/arch/parisc/kernel/pci.c @@ -228,7 +228,8 @@ resource_size_t pcibios_align_resource(void *data, const struct resource *res, } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; diff --git a/arch/powerpc/kernel/pci-common.c b/arch/powerpc/kernel/pci-common.c index ffda24a38dda..6dda4a20de6e 100644 --- a/arch/powerpc/kernel/pci-common.c +++ b/arch/powerpc/kernel/pci-common.c @@ -513,7 +513,8 @@ pgprot_t pci_phys_mem_access_prot(struct file *file, * * Returns a negative error code on failure, zero on success. */ -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { resource_size_t offset = diff --git a/arch/sh/drivers/pci/pci.c b/arch/sh/drivers/pci/pci.c index 84563e39a5b8..c8b36b7ceb98 100644 --- a/arch/sh/drivers/pci/pci.c +++ b/arch/sh/drivers/pci/pci.c @@ -269,7 +269,8 @@ void __ref pcibios_report_status(unsigned int status_mask, int warn) } } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { /* diff --git a/arch/sparc/kernel/pci.c b/arch/sparc/kernel/pci.c index 015e55a7495d..7eceaa10836f 100644 --- a/arch/sparc/kernel/pci.c +++ b/arch/sparc/kernel/pci.c @@ -862,9 +862,9 @@ static void __pci_mmap_set_pgprot(struct pci_dev *dev, struct vm_area_struct *vm * * Returns a negative error code on failure, zero on success. */ -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, - int write_combine) +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine) { int ret; diff --git a/arch/unicore32/kernel/pci.c b/arch/unicore32/kernel/pci.c index 62137d13c6f9..1b438857e91f 100644 --- a/arch/unicore32/kernel/pci.c +++ b/arch/unicore32/kernel/pci.c @@ -357,7 +357,8 @@ int pcibios_enable_device(struct pci_dev *dev, int mask) return 0; } -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long phys; diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index 0a9f2caf358f..8ca5e5d9f251 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -411,7 +411,8 @@ static const struct vm_operations_struct pci_mmap_ops = { .access = generic_access_phys, }; -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { unsigned long prot; diff --git a/arch/xtensa/kernel/pci.c b/arch/xtensa/kernel/pci.c index 5b73fc2f076c..903963ee495d 100644 --- a/arch/xtensa/kernel/pci.c +++ b/arch/xtensa/kernel/pci.c @@ -343,7 +343,8 @@ __pci_mmap_make_offset(struct pci_dev *dev, struct vm_area_struct *vma, * * Returns a negative error code on failure, zero on success. */ -int pci_mmap_page_range(struct pci_dev *dev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 534844df8bf5..bfd9efe637c4 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -1041,7 +1041,7 @@ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, pci_resource_to_user(pdev, bar, res, &start, &end); vma->vm_pgoff += start >> PAGE_SHIFT; mmap_type = res->flags & IORESOURCE_MEM ? pci_mmap_mem : pci_mmap_io; - return pci_mmap_page_range(pdev, vma, mmap_type, write_combine); + return pci_mmap_page_range(pdev, bar, vma, mmap_type, write_combine); } static int pci_mmap_resource_uc(struct file *filp, struct kobject *kobj, diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c index 45e5cf7e9193..098360d7ff81 100644 --- a/drivers/pci/proc.c +++ b/drivers/pci/proc.c @@ -262,7 +262,7 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma) else return -EINVAL; } - ret = pci_mmap_page_range(dev, vma, + ret = pci_mmap_page_range(dev, i, vma, fpriv->mmap_state, write_combine); if (ret < 0) return ret; diff --git a/include/linux/pci.h b/include/linux/pci.h index 590cfcf6acf5..7173a677d6dd 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1630,7 +1630,8 @@ static inline int pci_get_new_domain_nr(void) { return -ENOSYS; } * Architectures provide this function if they set HAVE_PCI_MMAP, and * it accepts the 'write_combine' argument when arch_can_pci_mmap_wc() * evaluates to nonzero. */ -int pci_mmap_page_range(struct pci_dev *pdev, struct vm_area_struct *vma, +int pci_mmap_page_range(struct pci_dev *pdev, int bar, + struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine); #ifndef arch_can_pci_mmap_wc -- cgit v1.2.3-59-g8ed1b From f719582435afe9c7985206e42d804ea6aa315d33 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:25:59 +0100 Subject: PCI: Add pci_mmap_resource_range() and use it for ARM64 Starting to leave behind the legacy of the pci_mmap_page_range() interface which takes "user-visible" BAR addresses. This takes just the resource and offset. For now, both APIs coexist and depending on the platform, one is implemented as a wrapper around the other. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- Documentation/filesystems/sysfs-pci.txt | 10 ++-- arch/arm64/include/asm/pci.h | 2 + drivers/pci/Makefile | 2 +- drivers/pci/mmap.c | 95 +++++++++++++++++++++++++++++++++ drivers/pci/pci-sysfs.c | 13 ++--- drivers/pci/pci.h | 4 +- include/linux/pci.h | 19 +++++-- 7 files changed, 125 insertions(+), 20 deletions(-) create mode 100644 drivers/pci/mmap.c diff --git a/Documentation/filesystems/sysfs-pci.txt b/Documentation/filesystems/sysfs-pci.txt index 46b95d82c4fd..06f1d64c6f70 100644 --- a/Documentation/filesystems/sysfs-pci.txt +++ b/Documentation/filesystems/sysfs-pci.txt @@ -113,9 +113,13 @@ Supporting PCI access on new platforms -------------------------------------- In order to support PCI resource mapping as described above, Linux platform -code must define HAVE_PCI_MMAP and provide a pci_mmap_page_range function. -Platforms are free to only support subsets of the mmap functionality, but -useful return codes should be provided. +code should ideally define ARCH_GENERIC_PCI_MMAP_RESOURCE and use the generic +implementation of that functionality. To support the historical interface of +mmap() through files in /proc/bus/pci, platforms may also set HAVE_PCI_MMAP. + +Alternatively, platforms which set HAVE_PCI_MMAP may provide their own +implementation of pci_mmap_page_range() instead of defining +ARCH_GENERIC_PCI_MMAP_RESOURCE. Platforms which support write-combining maps of PCI resources must define arch_can_pci_mmap_wc() which shall evaluate to non-zero at runtime when diff --git a/arch/arm64/include/asm/pci.h b/arch/arm64/include/asm/pci.h index b9a7ba9ca44c..1fc19744ffe9 100644 --- a/arch/arm64/include/asm/pci.h +++ b/arch/arm64/include/asm/pci.h @@ -22,6 +22,8 @@ */ #define PCI_DMA_BUS_IS_PHYS (0) +#define ARCH_GENERIC_PCI_MMAP_RESOURCE 1 + extern int isa_dma_bridge_buggy; #ifdef CONFIG_PCI diff --git a/drivers/pci/Makefile b/drivers/pci/Makefile index 8db5079f09a7..3d40e415a171 100644 --- a/drivers/pci/Makefile +++ b/drivers/pci/Makefile @@ -4,7 +4,7 @@ obj-y += access.o bus.o probe.o host-bridge.o remove.o pci.o \ pci-driver.o search.o pci-sysfs.o rom.o setup-res.o \ - irq.o vpd.o setup-bus.o vc.o + irq.o vpd.o setup-bus.o vc.o mmap.o obj-$(CONFIG_PROC_FS) += proc.o obj-$(CONFIG_SYSFS) += slot.o diff --git a/drivers/pci/mmap.c b/drivers/pci/mmap.c new file mode 100644 index 000000000000..b7aca9c89c31 --- /dev/null +++ b/drivers/pci/mmap.c @@ -0,0 +1,95 @@ +/* + * mmap.c — generic PCI resource mmap helper + * + * Copyright © 2017 Amazon.com, Inc. or its affiliates. + * + * Author: David Woodhouse + * + * 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 +#include +#include + +#ifdef ARCH_GENERIC_PCI_MMAP_RESOURCE + +/* + * Modern setup: generic pci_mmap_resource_range(), and implement the legacy + * pci_mmap_page_range() (if needed) as a wrapper round it. + */ + +#ifdef HAVE_PCI_MMAP +int pci_mmap_page_range(struct pci_dev *pdev, int bar, + struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine) +{ + resource_size_t start, end; + + pci_resource_to_user(pdev, bar, &pdev->resource[bar], &start, &end); + + /* Adjust vm_pgoff to be the offset within the resource */ + vma->vm_pgoff -= start >> PAGE_SHIFT; + return pci_mmap_resource_range(pdev, bar, vma, mmap_state, + write_combine); +} +#endif + +static const struct vm_operations_struct pci_phys_vm_ops = { +#ifdef CONFIG_HAVE_IOREMAP_PROT + .access = generic_access_phys, +#endif +}; + +int pci_mmap_resource_range(struct pci_dev *pdev, int bar, + struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine) +{ + unsigned long size; + + if (mmap_state == pci_mmap_io) + return -EINVAL; + + size = ((pci_resource_len(pdev, bar) - 1) >> PAGE_SHIFT) + 1; + if (vma->vm_pgoff + vma_pages(vma) > size) + return -EINVAL; + + if (write_combine) + vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); + else + vma->vm_page_prot = pgprot_device(vma->vm_page_prot); + + vma->vm_pgoff += (pci_resource_start(pdev, bar) >> PAGE_SHIFT); + vma->vm_ops = &pci_phys_vm_ops; + + return io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, + vma->vm_end - vma->vm_start, + vma->vm_page_prot); +} + +#elif defined(HAVE_PCI_MMAP) /* && !ARCH_GENERIC_PCI_MMAP_RESOURCE */ + +/* + * Legacy setup: Impement pci_mmap_resource_range() as a wrapper around + * the architecture's pci_mmap_page_range(), converting to "user visible" + * addresses as necessary. + */ + +int pci_mmap_resource_range(struct pci_dev *pdev, int bar, + struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine) +{ + resource_size_t start, end; + + /* + * pci_mmap_page_range() expects the same kind of entry as coming + * from /proc/bus/pci/ which is a "user visible" value. If this is + * different from the resource itself, arch will do necessary fixup. + */ + pci_resource_to_user(pdev, bar, &pdev->resource[bar], &start, &end); + vma->vm_pgoff += start >> PAGE_SHIFT; + return pci_mmap_page_range(pdev, bar, vma, mmap_state, write_combine); +} +#endif diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index bfd9efe637c4..10feb98a2b1d 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -980,7 +980,7 @@ void pci_remove_legacy_files(struct pci_bus *b) } #endif /* HAVE_PCI_LEGACY */ -#ifdef HAVE_PCI_MMAP +#if defined(HAVE_PCI_MMAP) || defined(ARCH_GENERIC_PCI_MMAP_RESOURCE) int pci_mmap_fits(struct pci_dev *pdev, int resno, struct vm_area_struct *vma, enum pci_mmap_api mmap_api) @@ -1019,7 +1019,6 @@ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, struct pci_dev *pdev = to_pci_dev(kobj_to_dev(kobj)); int bar = (unsigned long)attr->private; enum pci_mmap_state mmap_type; - resource_size_t start, end; struct resource *res = &pdev->resource[bar]; if (res->flags & IORESOURCE_MEM && iomem_is_exclusive(res->start)) @@ -1033,15 +1032,9 @@ static int pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr, (u64)pci_resource_len(pdev, bar)); return -EINVAL; } - - /* pci_mmap_page_range() expects the same kind of entry as coming - * from /proc/bus/pci/ which is a "user visible" value. If this is - * different from the resource itself, arch will do necessary fixup. - */ - pci_resource_to_user(pdev, bar, res, &start, &end); - vma->vm_pgoff += start >> PAGE_SHIFT; mmap_type = res->flags & IORESOURCE_MEM ? pci_mmap_mem : pci_mmap_io; - return pci_mmap_page_range(pdev, bar, vma, mmap_type, write_combine); + + return pci_mmap_resource_range(pdev, bar, vma, mmap_type, write_combine); } static int pci_mmap_resource_uc(struct file *filp, struct kobject *kobj, diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 8dd38e69d6f2..8e5ca2dec7e7 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -21,14 +21,14 @@ void pci_create_firmware_label_files(struct pci_dev *pdev); void pci_remove_firmware_label_files(struct pci_dev *pdev); #endif void pci_cleanup_rom(struct pci_dev *dev); -#ifdef HAVE_PCI_MMAP + enum pci_mmap_api { PCI_MMAP_SYSFS, /* mmap on /sys/bus/pci/devices//resource */ PCI_MMAP_PROCFS /* mmap on /proc/bus/pci/ */ }; int pci_mmap_fits(struct pci_dev *pdev, int resno, struct vm_area_struct *vmai, enum pci_mmap_api mmap_api); -#endif + int pci_probe_reset_function(struct pci_dev *dev); /** diff --git a/include/linux/pci.h b/include/linux/pci.h index 7173a677d6dd..98a72abcf361 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1626,10 +1626,21 @@ static inline int pci_get_new_domain_nr(void) { return -ENOSYS; } #include -/* Map a range of PCI memory or I/O space for a device into user space. - * Architectures provide this function if they set HAVE_PCI_MMAP, and - * it accepts the 'write_combine' argument when arch_can_pci_mmap_wc() - * evaluates to nonzero. */ +/* These two functions provide almost identical functionality. Depennding + * on the architecture, one will be implemented as a wrapper around the + * other (in drivers/pci/mmap.c). + * + * pci_mmap_resource_range() maps a specific BAR, and vm->vm_pgoff + * is expected to be an offset within that region. + * + * pci_mmap_page_range() is the legacy architecture-specific interface, + * which accepts a "user visible" resource address converted by + * pci_resource_to_user(), as used in the legacy mmap() interface in + * /proc/bus/pci/. + */ +int pci_mmap_resource_range(struct pci_dev *dev, int bar, + struct vm_area_struct *vma, + enum pci_mmap_state mmap_state, int write_combine); int pci_mmap_page_range(struct pci_dev *pdev, int bar, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine); -- cgit v1.2.3-59-g8ed1b From 00d2904ffeac067d7fe29c04edcfa0216102b4c2 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:00 +0100 Subject: ARM/PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/arm/include/asm/pci.h | 1 + arch/arm/kernel/bios32.c | 20 -------------------- 2 files changed, 1 insertion(+), 20 deletions(-) diff --git a/arch/arm/include/asm/pci.h b/arch/arm/include/asm/pci.h index 51118a0cabbc..396c92bcc0cf 100644 --- a/arch/arm/include/asm/pci.h +++ b/arch/arm/include/asm/pci.h @@ -29,6 +29,7 @@ static inline int pci_proc_domain(struct pci_bus *bus) #define PCI_DMA_BUS_IS_PHYS (1) #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel) { diff --git a/arch/arm/kernel/bios32.c b/arch/arm/kernel/bios32.c index a4fc3f46eeae..b259956365a0 100644 --- a/arch/arm/kernel/bios32.c +++ b/arch/arm/kernel/bios32.c @@ -597,26 +597,6 @@ resource_size_t pcibios_align_resource(void *data, const struct resource *res, return start; } -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - if (mmap_state == pci_mmap_io) - return -EINVAL; - - /* - * Mark this as IO - */ - vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - - if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, - vma->vm_page_prot)) - return -EAGAIN; - - return 0; -} - void __init pci_map_io_early(unsigned long pfn) { struct map_desc pci_io_desc = { -- cgit v1.2.3-59-g8ed1b From 3efdd5fe0a363e308ed61d020f14371dfb5aab4a Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:01 +0100 Subject: cris/PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas Acked-by: Jesper Nilsson --- arch/cris/arch-v32/drivers/pci/bios.c | 23 ----------------------- arch/cris/include/asm/pci.h | 1 + 2 files changed, 1 insertion(+), 23 deletions(-) diff --git a/arch/cris/arch-v32/drivers/pci/bios.c b/arch/cris/arch-v32/drivers/pci/bios.c index a589686d5448..394c2a73d5e2 100644 --- a/arch/cris/arch-v32/drivers/pci/bios.c +++ b/arch/cris/arch-v32/drivers/pci/bios.c @@ -14,29 +14,6 @@ void pcibios_set_master(struct pci_dev *dev) pci_write_config_byte(dev, PCI_LATENCY_TIMER, lat); } -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long prot; - - /* Leave vm_pgoff as-is, the PCI space address is the physical - * address on this platform. - */ - prot = pgprot_val(vma->vm_page_prot); - vma->vm_page_prot = __pgprot(prot); - - /* Write-combine setting is ignored, it is changed via the mtrr - * interfaces on this platform. - */ - if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, - vma->vm_page_prot)) - return -EAGAIN; - - return 0; -} - resource_size_t pcibios_align_resource(void *data, const struct resource *res, resource_size_t size, resource_size_t align) diff --git a/arch/cris/include/asm/pci.h b/arch/cris/include/asm/pci.h index 65198cb46c85..6e505332b3e3 100644 --- a/arch/cris/include/asm/pci.h +++ b/arch/cris/include/asm/pci.h @@ -42,6 +42,7 @@ struct pci_dev; #define PCI_DMA_BUS_IS_PHYS (1) #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #endif /* __KERNEL__ */ -- cgit v1.2.3-59-g8ed1b From 35368f80403d32a993587c7f26cd3bcede3ed202 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:02 +0100 Subject: MIPS: PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/mips/include/asm/pci.h | 2 +- arch/mips/pci/pci.c | 25 ------------------------- 2 files changed, 1 insertion(+), 26 deletions(-) diff --git a/arch/mips/include/asm/pci.h b/arch/mips/include/asm/pci.h index 3141e2a8113a..1000c1b4c875 100644 --- a/arch/mips/include/asm/pci.h +++ b/arch/mips/include/asm/pci.h @@ -110,7 +110,7 @@ extern unsigned long PCIBIOS_MIN_MEM; extern void pcibios_set_master(struct pci_dev *dev); #define HAVE_PCI_MMAP - +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #define HAVE_ARCH_PCI_RESOURCE_TO_USER /* diff --git a/arch/mips/pci/pci.c b/arch/mips/pci/pci.c index f189502041a6..bd67ac74fe2d 100644 --- a/arch/mips/pci/pci.c +++ b/arch/mips/pci/pci.c @@ -57,28 +57,3 @@ void pci_resource_to_user(const struct pci_dev *dev, int bar, *start = fixup_bigphys_addr(rsrc->start, size); *end = rsrc->start + size; } - -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long prot; - - /* - * I/O space can be accessed via normal processor loads and stores on - * this platform but for now we elect not to do this and portable - * drivers should not do this anyway. - */ - if (mmap_state == pci_mmap_io) - return -EINVAL; - - /* - * Ignore write-combine; for now only return uncached mappings. - */ - prot = pgprot_val(vma->vm_page_prot); - prot = (prot & ~_CACHE_MASK) | _CACHE_UNCACHED; - vma->vm_page_prot = __pgprot(prot); - - return remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, vma->vm_page_prot); -} -- cgit v1.2.3-59-g8ed1b From 903bdbe271217f656d13767d8a3e74b8aabc54c2 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:03 +0100 Subject: mn10300/PCI: Use generic pci_mmap_resource_range() This was setting vma->vm_flags |= VM_LOCKED. Not sure why... Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas Reviewed-by: David Howells --- arch/mn10300/include/asm/pci.h | 1 + arch/mn10300/unit-asb2305/pci-asb2305.c | 24 ------------------------ 2 files changed, 1 insertion(+), 24 deletions(-) diff --git a/arch/mn10300/include/asm/pci.h b/arch/mn10300/include/asm/pci.h index 082b6de90936..d27654902f28 100644 --- a/arch/mn10300/include/asm/pci.h +++ b/arch/mn10300/include/asm/pci.h @@ -74,6 +74,7 @@ static inline int pci_controller_num(struct pci_dev *dev) } #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #endif /* __KERNEL__ */ diff --git a/arch/mn10300/unit-asb2305/pci-asb2305.c b/arch/mn10300/unit-asb2305/pci-asb2305.c index 4abbbd54ac7d..e0f4617c0c7a 100644 --- a/arch/mn10300/unit-asb2305/pci-asb2305.c +++ b/arch/mn10300/unit-asb2305/pci-asb2305.c @@ -210,27 +210,3 @@ void __init pcibios_resource_survey(void) pcibios_allocate_resources(0); pcibios_allocate_resources(1); } - -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long prot; - - /* Leave vm_pgoff as-is, the PCI space address is the physical - * address on this platform. - */ - vma->vm_flags |= VM_LOCKED; - - prot = pgprot_val(vma->vm_page_prot); - prot &= ~_PAGE_CACHE; - vma->vm_page_prot = __pgprot(prot); - - /* Write-combine setting is ignored */ - if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, - vma->vm_page_prot)) - return -EAGAIN; - - return 0; -} -- cgit v1.2.3-59-g8ed1b From 6a94ca14c389b00515b026cd35bcd074dec6f7ff Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:04 +0100 Subject: parisc: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/parisc/include/asm/pci.h | 1 + arch/parisc/kernel/pci.c | 29 ----------------------------- 2 files changed, 1 insertion(+), 29 deletions(-) diff --git a/arch/parisc/include/asm/pci.h b/arch/parisc/include/asm/pci.h index bb9ea9003e08..1de1a3f412ec 100644 --- a/arch/parisc/include/asm/pci.h +++ b/arch/parisc/include/asm/pci.h @@ -200,5 +200,6 @@ static inline int pci_get_legacy_ide_irq(struct pci_dev *dev, int channel) } #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #endif /* __ASM_PARISC_PCI_H */ diff --git a/arch/parisc/kernel/pci.c b/arch/parisc/kernel/pci.c index 653877553acd..13ee3569959a 100644 --- a/arch/parisc/kernel/pci.c +++ b/arch/parisc/kernel/pci.c @@ -227,35 +227,6 @@ resource_size_t pcibios_align_resource(void *data, const struct resource *res, return start; } - -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long prot; - - /* - * I/O space can be accessed via normal processor loads and stores on - * this platform but for now we elect not to do this and portable - * drivers should not do this anyway. - */ - if (mmap_state == pci_mmap_io) - return -EINVAL; - - if (write_combine) - return -EINVAL; - - /* - * Ignore write-combine; for now only return uncached mappings. - */ - prot = pgprot_val(vma->vm_page_prot); - prot |= _PAGE_NO_CACHE; - vma->vm_page_prot = __pgprot(prot); - - return remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, vma->vm_page_prot); -} - /* * A driver is enabling the device. We make sure that all the appropriate * bits are set to allow the device to operate as the driver is expecting. -- cgit v1.2.3-59-g8ed1b From adf7bde63f049ab0c3284a67898055bb11614c15 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:05 +0100 Subject: sh/PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/sh/drivers/pci/pci.c | 22 ---------------------- arch/sh/include/asm/pci.h | 1 + 2 files changed, 1 insertion(+), 22 deletions(-) diff --git a/arch/sh/drivers/pci/pci.c b/arch/sh/drivers/pci/pci.c index c8b36b7ceb98..c99ee286b69f 100644 --- a/arch/sh/drivers/pci/pci.c +++ b/arch/sh/drivers/pci/pci.c @@ -269,28 +269,6 @@ void __ref pcibios_report_status(unsigned int status_mask, int warn) } } -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - /* - * I/O space can be accessed via normal processor loads and stores on - * this platform but for now we elect not to do this and portable - * drivers should not do this anyway. - */ - if (mmap_state == pci_mmap_io) - return -EINVAL; - - /* - * Ignore write-combine; for now only return uncached mappings. - */ - vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - - return remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, - vma->vm_page_prot); -} - #ifndef CONFIG_GENERIC_IOMAP void __iomem *__pci_ioport_map(struct pci_dev *dev, diff --git a/arch/sh/include/asm/pci.h b/arch/sh/include/asm/pci.h index 46abbc9983c9..17fa69bc814d 100644 --- a/arch/sh/include/asm/pci.h +++ b/arch/sh/include/asm/pci.h @@ -66,6 +66,7 @@ extern unsigned long PCIBIOS_MIN_IO, PCIBIOS_MIN_MEM; struct pci_dev; #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE extern void pcibios_set_master(struct pci_dev *dev); -- cgit v1.2.3-59-g8ed1b From 69c701ec1f97315dc0f482699f370887448e150e Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:06 +0100 Subject: unicore32/PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/unicore32/include/asm/pci.h | 1 + arch/unicore32/kernel/pci.c | 24 ------------------------ 2 files changed, 1 insertion(+), 24 deletions(-) diff --git a/arch/unicore32/include/asm/pci.h b/arch/unicore32/include/asm/pci.h index a51290862acd..ac5acdf4c4d0 100644 --- a/arch/unicore32/include/asm/pci.h +++ b/arch/unicore32/include/asm/pci.h @@ -17,6 +17,7 @@ #include /* for PCIBIOS_MIN_* */ #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #endif /* __KERNEL__ */ #endif diff --git a/arch/unicore32/kernel/pci.c b/arch/unicore32/kernel/pci.c index 1b438857e91f..1053bca1f8aa 100644 --- a/arch/unicore32/kernel/pci.c +++ b/arch/unicore32/kernel/pci.c @@ -356,27 +356,3 @@ int pcibios_enable_device(struct pci_dev *dev, int mask) } return 0; } - -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long phys; - - if (mmap_state == pci_mmap_io) - return -EINVAL; - - phys = vma->vm_pgoff; - - /* - * Mark this as IO - */ - vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - - if (remap_pfn_range(vma, vma->vm_start, phys, - vma->vm_end - vma->vm_start, - vma->vm_page_prot)) - return -EAGAIN; - - return 0; -} -- cgit v1.2.3-59-g8ed1b From 5c2d5ce2ab4bd716b2125f15b2e0879143ac0563 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:07 +0100 Subject: x86/PCI: Use generic pci_mmap_resource_range() Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- arch/x86/include/asm/pci.h | 1 + arch/x86/pci/i386.c | 48 ---------------------------------------------- 2 files changed, 1 insertion(+), 48 deletions(-) diff --git a/arch/x86/include/asm/pci.h b/arch/x86/include/asm/pci.h index 734cc9411b3a..f513cc231151 100644 --- a/arch/x86/include/asm/pci.h +++ b/arch/x86/include/asm/pci.h @@ -104,6 +104,7 @@ int pcibios_set_irq_routing(struct pci_dev *dev, int pin, int irq); #define HAVE_PCI_MMAP #define arch_can_pci_mmap_wc() pat_enabled() +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #ifdef CONFIG_PCI extern void early_quirks(void); diff --git a/arch/x86/pci/i386.c b/arch/x86/pci/i386.c index 8ca5e5d9f251..68499cf82ccf 100644 --- a/arch/x86/pci/i386.c +++ b/arch/x86/pci/i386.c @@ -406,51 +406,3 @@ void __init pcibios_resource_survey(void) */ ioapic_insert_resources(); } - -static const struct vm_operations_struct pci_mmap_ops = { - .access = generic_access_phys, -}; - -int pci_mmap_page_range(struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - unsigned long prot; - - /* I/O space cannot be accessed via normal processor loads and - * stores on this platform. - */ - if (mmap_state == pci_mmap_io) - return -EINVAL; - - prot = pgprot_val(vma->vm_page_prot); - - /* - * Return error if pat is not enabled and write_combine is requested. - * Caller can followup with UC MINUS request and add a WC mtrr if there - * is a free mtrr slot. - */ - if (!pat_enabled() && write_combine) - return -EINVAL; - - if (pat_enabled() && write_combine) - prot |= cachemode2protval(_PAGE_CACHE_MODE_WC); - else if (pat_enabled() || boot_cpu_data.x86 > 3) - /* - * ioremap() and ioremap_nocache() defaults to UC MINUS for now. - * To avoid attribute conflicts, request UC MINUS here - * as well. - */ - prot |= cachemode2protval(_PAGE_CACHE_MODE_UC_MINUS); - - vma->vm_page_prot = __pgprot(prot); - - if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, - vma->vm_page_prot)) - return -EAGAIN; - - vma->vm_ops = &pci_mmap_ops; - - return 0; -} -- cgit v1.2.3-59-g8ed1b From 2bea36fd1af440e1853e431459a0bf929266cd52 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Wed, 12 Apr 2017 13:26:08 +0100 Subject: PCI: Add I/O BAR support to generic pci_mmap_resource_range() This will need to call into an arch-provided pci_iobar_pfn() function. Signed-off-by: David Woodhouse Signed-off-by: Bjorn Helgaas --- drivers/pci/mmap.c | 12 ++++++++---- include/linux/pci.h | 4 ++++ 2 files changed, 12 insertions(+), 4 deletions(-) diff --git a/drivers/pci/mmap.c b/drivers/pci/mmap.c index b7aca9c89c31..9a5e5a9055eb 100644 --- a/drivers/pci/mmap.c +++ b/drivers/pci/mmap.c @@ -48,9 +48,7 @@ int pci_mmap_resource_range(struct pci_dev *pdev, int bar, enum pci_mmap_state mmap_state, int write_combine) { unsigned long size; - - if (mmap_state == pci_mmap_io) - return -EINVAL; + int ret; size = ((pci_resource_len(pdev, bar) - 1) >> PAGE_SHIFT) + 1; if (vma->vm_pgoff + vma_pages(vma) > size) @@ -61,7 +59,13 @@ int pci_mmap_resource_range(struct pci_dev *pdev, int bar, else vma->vm_page_prot = pgprot_device(vma->vm_page_prot); - vma->vm_pgoff += (pci_resource_start(pdev, bar) >> PAGE_SHIFT); + if (mmap_state == pci_mmap_io) { + ret = pci_iobar_pfn(pdev, bar, vma); + if (ret) + return ret; + } else + vma->vm_pgoff += (pci_resource_start(pdev, bar) >> PAGE_SHIFT); + vma->vm_ops = &pci_phys_vm_ops; return io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, diff --git a/include/linux/pci.h b/include/linux/pci.h index 98a72abcf361..9f302444d4ac 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1648,8 +1648,12 @@ int pci_mmap_page_range(struct pci_dev *pdev, int bar, #ifndef arch_can_pci_mmap_wc #define arch_can_pci_mmap_wc() 0 #endif + #ifndef arch_can_pci_mmap_io #define arch_can_pci_mmap_io() 0 +#define pci_iobar_pfn(pdev, bar, vma) (-EINVAL) +#else +int pci_iobar_pfn(struct pci_dev *pdev, int bar, struct vm_area_struct *vma); #endif #ifndef pci_root_bus_fwnode -- cgit v1.2.3-59-g8ed1b From 61eee41ae135517392d05cbf1177523a02f12727 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Fri, 7 Apr 2017 11:22:42 +0200 Subject: ia64: Remove redundant valid_mmap_phys_addr_range() from pci_mmap_page_range() We know we are within a valid MMIO BAR by the time this function gets called; there's no need to check. Signed-off-by: David Woodhouse Tested-by: Tony Luck --- arch/ia64/pci/pci.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 053c688b15a5..27020f30caa6 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -439,9 +439,6 @@ pci_mmap_page_range (struct pci_dev *dev, int bar, */ return -EINVAL; - if (!valid_mmap_phys_addr_range(vma->vm_pgoff, size)) - return -EINVAL; - prot = phys_mem_access_prot(NULL, vma->vm_pgoff, size, vma->vm_page_prot); -- cgit v1.2.3-59-g8ed1b From fcdb10d6b179fde41ca94d032afda8f4ed796b8e Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Fri, 7 Apr 2017 12:01:00 +0200 Subject: ia64: Remove redundant checks for WC in pci_mmap_page_range() For a PCI MMIO BAR, phys_mem_access_prot() should always return UC or WC. And while a mixture of cached and uncached mappings is forbidden, we were already mixing WC and UC, which is OK. Just do as we're asked. Signed-off-by: David Woodhouse Tested-by: Tony Luck --- arch/ia64/pci/pci.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-) diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 27020f30caa6..7438e8c84cdd 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -423,9 +423,6 @@ pci_mmap_page_range (struct pci_dev *dev, int bar, struct vm_area_struct *vma, enum pci_mmap_state mmap_state, int write_combine) { - unsigned long size = vma->vm_end - vma->vm_start; - pgprot_t prot; - /* * I/O space cannot be accessed via normal processor loads and * stores on this platform. @@ -439,21 +436,10 @@ pci_mmap_page_range (struct pci_dev *dev, int bar, */ return -EINVAL; - prot = phys_mem_access_prot(NULL, vma->vm_pgoff, size, - vma->vm_page_prot); - - /* - * If the user requested WC, the kernel uses UC or WC for this region, - * and the chipset supports WC, we can use WC. Otherwise, we have to - * use the same attribute the kernel uses. - */ - if (write_combine && - ((pgprot_val(prot) & _PAGE_MA_MASK) == _PAGE_MA_UC || - (pgprot_val(prot) & _PAGE_MA_MASK) == _PAGE_MA_WC) && - efi_range_is_wc(vma->vm_start, vma->vm_end - vma->vm_start)) + if (write_combine) vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); else - vma->vm_page_prot = prot; + vma->vm_page_prot = pgprot_device(vma->vm_page_prot); if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, vma->vm_end - vma->vm_start, vma->vm_page_prot)) -- cgit v1.2.3-59-g8ed1b From d9c102de2caa733c3e718e1b032cd154a9994326 Mon Sep 17 00:00:00 2001 From: David Woodhouse Date: Fri, 7 Apr 2017 12:10:53 +0200 Subject: ia64: Use generic pci_mmap_resource_range() Now that we eliminated the different behaviour in separately-reviewable commits, we can switch IA64 to the generic implementation. Signed-off-by: David Woodhouse Tested-by: Tony Luck --- arch/ia64/include/asm/pci.h | 1 + arch/ia64/pci/pci.c | 30 ------------------------------ 2 files changed, 1 insertion(+), 30 deletions(-) diff --git a/arch/ia64/include/asm/pci.h b/arch/ia64/include/asm/pci.h index fc60b3d139f4..6459f2d46200 100644 --- a/arch/ia64/include/asm/pci.h +++ b/arch/ia64/include/asm/pci.h @@ -51,6 +51,7 @@ extern unsigned long ia64_max_iommu_merge_mask; #define PCI_DMA_BUS_IS_PHYS (ia64_max_iommu_merge_mask == ~0UL) #define HAVE_PCI_MMAP +#define ARCH_GENERIC_PCI_MMAP_RESOURCE #define arch_can_pci_mmap_wc() 1 #define HAVE_PCI_LEGACY diff --git a/arch/ia64/pci/pci.c b/arch/ia64/pci/pci.c index 7438e8c84cdd..4068bde623dc 100644 --- a/arch/ia64/pci/pci.c +++ b/arch/ia64/pci/pci.c @@ -418,36 +418,6 @@ pcibios_align_resource (void *data, const struct resource *res, return res->start; } -int -pci_mmap_page_range (struct pci_dev *dev, int bar, - struct vm_area_struct *vma, - enum pci_mmap_state mmap_state, int write_combine) -{ - /* - * I/O space cannot be accessed via normal processor loads and - * stores on this platform. - */ - if (mmap_state == pci_mmap_io) - /* - * XXX we could relax this for I/O spaces for which ACPI - * indicates that the space is 1-to-1 mapped. But at the - * moment, we don't support multiple PCI address spaces and - * the legacy I/O space is not 1-to-1 mapped, so this is moot. - */ - return -EINVAL; - - if (write_combine) - vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot); - else - vma->vm_page_prot = pgprot_device(vma->vm_page_prot); - - if (remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - vma->vm_end - vma->vm_start, vma->vm_page_prot)) - return -EAGAIN; - - return 0; -} - /** * ia64_pci_get_legacy_mem - generic legacy mem routine * @bus: bus to get legacy memory base address for -- cgit v1.2.3-59-g8ed1b From 0e7df22401a3dfd403b26dea62dd00e0598b538b Mon Sep 17 00:00:00 2001 From: Bodong Wang Date: Thu, 13 Apr 2017 01:51:40 +0300 Subject: PCI: Add sysfs sriov_drivers_autoprobe to control VF driver binding Sometimes it is not desirable to bind SR-IOV VFs to drivers. This can save host side resource usage by VF instances that will be assigned to VMs. Add a new PCI sysfs interface "sriov_drivers_autoprobe" to control that from the PF. To modify it, echo 0/n/N (disable probe) or 1/y/Y (enable probe) to: /sys/bus/pci/devices//sriov_drivers_autoprobe Note that this must be done before enabling VFs. The change will not take effect if VFs are already enabled. Simply, one can disable VFs by setting sriov_numvfs to 0, choose whether to probe or not, and then re-enable the VFs by restoring sriov_numvfs. [bhelgaas: changelog, ABI doc] Signed-off-by: Bodong Wang Signed-off-by: Eli Cohen Signed-off-by: Bjorn Helgaas Reviewed-by: Gavin Shan Reviewed-by: Alex Williamson --- Documentation/ABI/testing/sysfs-bus-pci | 22 ++++++++++++++++++++++ Documentation/PCI/pci-iov-howto.txt | 12 ++++++++++++ drivers/pci/iov.c | 1 + drivers/pci/pci-driver.c | 22 ++++++++++++++++++---- drivers/pci/pci-sysfs.c | 28 ++++++++++++++++++++++++++++ drivers/pci/pci.h | 1 + 6 files changed, 82 insertions(+), 4 deletions(-) diff --git a/Documentation/ABI/testing/sysfs-bus-pci b/Documentation/ABI/testing/sysfs-bus-pci index 5a1732b78707..b728f700340d 100644 --- a/Documentation/ABI/testing/sysfs-bus-pci +++ b/Documentation/ABI/testing/sysfs-bus-pci @@ -301,3 +301,25 @@ Contact: Emil Velikov Description: This file contains the revision field of the the PCI device. The value comes from device config space. The file is read only. + +What: /sys/bus/pci/devices/.../sriov_drivers_autoprobe +Date: April 2017 +Contact: Bodong Wang +Description: + This file is associated with the PF of a device that + supports SR-IOV. It determines whether newly-enabled VFs + are immediately bound to a driver. It initially contains + 1, which means the kernel automatically binds VFs to a + compatible driver immediately after they are enabled. If + an application writes 0 to the file before enabling VFs, + the kernel will not bind VFs to a driver. + + A typical use case is to write 0 to this file, then enable + VFs, then assign the newly-created VFs to virtual machines. + Note that changing this file does not affect already- + enabled VFs. In this scenario, the user must first disable + the VFs, write 0 to sriov_drivers_autoprobe, then re-enable + the VFs. + + This is similar to /sys/bus/pci/drivers_autoprobe, but + affects only the VFs associated with a specific PF. diff --git a/Documentation/PCI/pci-iov-howto.txt b/Documentation/PCI/pci-iov-howto.txt index 2d91ae251982..d2a84151e99c 100644 --- a/Documentation/PCI/pci-iov-howto.txt +++ b/Documentation/PCI/pci-iov-howto.txt @@ -68,6 +68,18 @@ To disable SR-IOV capability: echo 0 > \ /sys/bus/pci/devices//sriov_numvfs +To enable auto probing VFs by a compatible driver on the host, run +command below before enabling SR-IOV capabilities. This is the +default behavior. + echo 1 > \ + /sys/bus/pci/devices//sriov_drivers_autoprobe + +To disable auto probing VFs by a compatible driver on the host, run +command below before enabling SR-IOV capabilities. Updating this +entry will not affect VFs which are already probed. + echo 0 > \ + /sys/bus/pci/devices//sriov_drivers_autoprobe + 3.2 Usage example Following piece of code illustrates the usage of the SR-IOV API. diff --git a/drivers/pci/iov.c b/drivers/pci/iov.c index 2479ae876482..d9dc7363ac77 100644 --- a/drivers/pci/iov.c +++ b/drivers/pci/iov.c @@ -450,6 +450,7 @@ found: iov->total_VFs = total; iov->pgsz = pgsz; iov->self = dev; + iov->drivers_autoprobe = true; pci_read_config_dword(dev, pos + PCI_SRIOV_CAP, &iov->cap); pci_read_config_byte(dev, pos + PCI_SRIOV_FUNC_LINK, &iov->link); if (pci_pcie_type(dev) == PCI_EXP_TYPE_RC_END) diff --git a/drivers/pci/pci-driver.c b/drivers/pci/pci-driver.c index afa72717a979..f99f7fe749b8 100644 --- a/drivers/pci/pci-driver.c +++ b/drivers/pci/pci-driver.c @@ -394,6 +394,18 @@ void __weak pcibios_free_irq(struct pci_dev *dev) { } +#ifdef CONFIG_PCI_IOV +static inline bool pci_device_can_probe(struct pci_dev *pdev) +{ + return (!pdev->is_virtfn || pdev->physfn->sriov->drivers_autoprobe); +} +#else +static inline bool pci_device_can_probe(struct pci_dev *pdev) +{ + return true; +} +#endif + static int pci_device_probe(struct device *dev) { int error; @@ -405,10 +417,12 @@ static int pci_device_probe(struct device *dev) return error; pci_dev_get(pci_dev); - error = __pci_device_probe(drv, pci_dev); - if (error) { - pcibios_free_irq(pci_dev); - pci_dev_put(pci_dev); + if (pci_device_can_probe(pci_dev)) { + error = __pci_device_probe(drv, pci_dev); + if (error) { + pcibios_free_irq(pci_dev); + pci_dev_put(pci_dev); + } } return error; diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c index 25d010d449a3..369c999683ca 100644 --- a/drivers/pci/pci-sysfs.c +++ b/drivers/pci/pci-sysfs.c @@ -526,10 +526,37 @@ exit: return count; } +static ssize_t sriov_drivers_autoprobe_show(struct device *dev, + struct device_attribute *attr, + char *buf) +{ + struct pci_dev *pdev = to_pci_dev(dev); + + return sprintf(buf, "%u\n", pdev->sriov->drivers_autoprobe); +} + +static ssize_t sriov_drivers_autoprobe_store(struct device *dev, + struct device_attribute *attr, + const char *buf, size_t count) +{ + struct pci_dev *pdev = to_pci_dev(dev); + bool drivers_autoprobe; + + if (kstrtobool(buf, &drivers_autoprobe) < 0) + return -EINVAL; + + pdev->sriov->drivers_autoprobe = drivers_autoprobe; + + return count; +} + static struct device_attribute sriov_totalvfs_attr = __ATTR_RO(sriov_totalvfs); static struct device_attribute sriov_numvfs_attr = __ATTR(sriov_numvfs, (S_IRUGO|S_IWUSR|S_IWGRP), sriov_numvfs_show, sriov_numvfs_store); +static struct device_attribute sriov_drivers_autoprobe_attr = + __ATTR(sriov_drivers_autoprobe, (S_IRUGO|S_IWUSR|S_IWGRP), + sriov_drivers_autoprobe_show, sriov_drivers_autoprobe_store); #endif /* CONFIG_PCI_IOV */ static ssize_t driver_override_store(struct device *dev, @@ -1549,6 +1576,7 @@ static struct attribute_group pci_dev_hp_attr_group = { static struct attribute *sriov_dev_attrs[] = { &sriov_totalvfs_attr.attr, &sriov_numvfs_attr.attr, + &sriov_drivers_autoprobe_attr.attr, NULL, }; diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h index 8dd38e69d6f2..3ba7d586f522 100644 --- a/drivers/pci/pci.h +++ b/drivers/pci/pci.h @@ -272,6 +272,7 @@ struct pci_sriov { struct pci_dev *self; /* this PF */ struct mutex lock; /* lock for setting sriov_numvfs in sysfs */ resource_size_t barsz[PCI_SRIOV_NUM_BARS]; /* VF BAR size */ + bool drivers_autoprobe; /* auto probing of VFs by driver */ }; #ifdef CONFIG_PCI_ATS -- cgit v1.2.3-59-g8ed1b From a60a2b73ba69abca26653fff157b0fd8947bc498 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 14 Apr 2017 21:11:25 +0200 Subject: PCI: Export pcie_flr() Currently we opencode the FLR sequence in lots of place; export a core helper instead. We split out the probing for FLR support as all the non-core callers already know their hardware. Note that in the new pci_has_flr() function the quirk check has been moved before the capability check as there is no point in reading the capability in this case. Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas --- drivers/pci/pci.c | 39 ++++++++++++++++++++++++++++----------- include/linux/pci.h | 1 + 2 files changed, 29 insertions(+), 11 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index bef14777bb30..957a11a6a840 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -3773,27 +3773,41 @@ static void pci_flr_wait(struct pci_dev *dev) (i - 1) * 100); } -static int pcie_flr(struct pci_dev *dev, int probe) +/** + * pcie_has_flr - check if a device supports function level resets + * @dev: device to check + * + * Returns true if the device advertises support for PCIe function level + * resets. + */ +static bool pcie_has_flr(struct pci_dev *dev) { u32 cap; - pcie_capability_read_dword(dev, PCI_EXP_DEVCAP, &cap); - if (!(cap & PCI_EXP_DEVCAP_FLR)) - return -ENOTTY; - if (dev->dev_flags & PCI_DEV_FLAGS_NO_FLR_RESET) - return -ENOTTY; + return false; - if (probe) - return 0; + pcie_capability_read_dword(dev, PCI_EXP_DEVCAP, &cap); + return cap & PCI_EXP_DEVCAP_FLR; +} +/** + * pcie_flr - initiate a PCIe function level reset + * @dev: device to reset + * + * Initiate a function level reset on @dev. The caller should ensure the + * device supports FLR before calling this function, e.g. by using the + * pcie_has_flr() helper. + */ +void pcie_flr(struct pci_dev *dev) +{ if (!pci_wait_for_pending_transaction(dev)) dev_err(&dev->dev, "timed out waiting for pending transaction; performing function level reset anyway\n"); pcie_capability_set_word(dev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); pci_flr_wait(dev); - return 0; } +EXPORT_SYMBOL_GPL(pcie_flr); static int pci_af_flr(struct pci_dev *dev, int probe) { @@ -3977,9 +3991,12 @@ static int __pci_dev_reset(struct pci_dev *dev, int probe) if (rc != -ENOTTY) goto done; - rc = pcie_flr(dev, probe); - if (rc != -ENOTTY) + if (pcie_has_flr(dev)) { + if (!probe) + pcie_flr(dev); + rc = 0; goto done; + } rc = pci_af_flr(dev, probe); if (rc != -ENOTTY) diff --git a/include/linux/pci.h b/include/linux/pci.h index 22cad2c66d59..a9ff99c91601 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1054,6 +1054,7 @@ int pcie_get_mps(struct pci_dev *dev); int pcie_set_mps(struct pci_dev *dev, int mps); int pcie_get_minimum_link(struct pci_dev *dev, enum pci_bus_speed *speed, enum pcie_link_width *width); +void pcie_flr(struct pci_dev *dev); int __pci_reset_function(struct pci_dev *dev); int __pci_reset_function_locked(struct pci_dev *dev); int pci_reset_function(struct pci_dev *dev); -- cgit v1.2.3-59-g8ed1b From c8d8096abc84b544dd4f37f12e929d5f8aadd160 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 14 Apr 2017 21:11:26 +0200 Subject: PCI: Call pcie_flr() from reset_intel_82599_sfp_virtfn() The 82599 quirk contained an outdated copy of the FLR code. Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas --- drivers/pci/quirks.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index 823271b13d12..fd5ec86816aa 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -3641,19 +3641,11 @@ static int reset_intel_82599_sfp_virtfn(struct pci_dev *dev, int probe) * * The 82599 supports FLR on VFs, but FLR support is reported only * in the PF DEVCAP (sec 9.3.10.4), not in the VF DEVCAP (sec 9.5). - * Therefore, we can't use pcie_flr(), which checks the VF DEVCAP. + * Thus we must call pcie_flr() directly without first checking if it is + * supported. */ - - if (probe) - return 0; - - if (!pci_wait_for_pending_transaction(dev)) - dev_err(&dev->dev, "transaction is not cleared; proceeding with reset anyway\n"); - - pcie_capability_set_word(dev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); - - msleep(100); - + if (!probe) + pcie_flr(dev); return 0; } -- cgit v1.2.3-59-g8ed1b From 48f52d1a8ca9f1aff3ffad8e35b6c56eafa17b9b Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 14 Apr 2017 21:11:27 +0200 Subject: PCI: Call pcie_flr() from reset_chelsio_generic_dev() Instead of copy & pasting and old version of the code. Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas --- drivers/pci/quirks.c | 15 +-------------- 1 file changed, 1 insertion(+), 14 deletions(-) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index fd5ec86816aa..a3a0b5a0b72c 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -3750,20 +3750,7 @@ static int reset_chelsio_generic_dev(struct pci_dev *dev, int probe) PCI_MSIX_FLAGS_ENABLE | PCI_MSIX_FLAGS_MASKALL); - /* - * Start of pcie_flr() code sequence. This reset code is a copy of - * the guts of pcie_flr() because that's not an exported function. - */ - - if (!pci_wait_for_pending_transaction(dev)) - dev_err(&dev->dev, "transaction is not cleared; proceeding with reset anyway\n"); - - pcie_capability_set_word(dev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); - msleep(100); - - /* - * End of pcie_flr() code sequence. - */ + pcie_flr(dev); /* * Restore the configuration information (BAR values, etc.) including -- cgit v1.2.3-59-g8ed1b From 073d3dbe9a7c38888d8dafe09df421b174a6cffa Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 9 Mar 2017 18:46:15 -0800 Subject: PCI: rockchip: Add remove() support Currently, if we try to unbind the platform device, the remove will succeed, but the removal won't undo most of the registration, leaving partially-configured PCI devices in the system. This allows, for example, a simple 'lspci' to crash the system, as it will try to touch the freed (via devm_*) driver structures, e.g., on RK3399: # echo f8000000.pcie > /sys/bus/platform/drivers/rockchip-pcie/unbind # lspci So let's implement device remove(). Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas Acked-by: Shawn Lin --- drivers/pci/host/pcie-rockchip.c | 36 ++++++++++++++++++++++++++++++++++-- 1 file changed, 34 insertions(+), 2 deletions(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index 94feb7dfd8f4..76f2edc93663 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -223,9 +223,11 @@ struct rockchip_pcie { int link_gen; struct device *dev; struct irq_domain *irq_domain; - u32 io_size; int offset; + struct pci_bus *root_bus; + struct resource *io; phys_addr_t io_bus_addr; + u32 io_size; void __iomem *msg_region; u32 mem_size; phys_addr_t msg_bus_addr; @@ -1366,6 +1368,7 @@ static int rockchip_pcie_probe(struct platform_device *pdev) err, io); continue; } + rockchip->io = io; break; case IORESOURCE_MEM: mem = win->res; @@ -1397,6 +1400,7 @@ static int rockchip_pcie_probe(struct platform_device *pdev) err = -ENOMEM; goto err_free_res; } + rockchip->root_bus = bus; pci_bus_size_bridges(bus); pci_bus_assign_resources(bus); @@ -1427,6 +1431,34 @@ err_aclk_pcie: return err; } +static int rockchip_pcie_remove(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct rockchip_pcie *rockchip = dev_get_drvdata(dev); + + pci_stop_root_bus(rockchip->root_bus); + pci_remove_root_bus(rockchip->root_bus); + pci_unmap_iospace(rockchip->io); + irq_domain_remove(rockchip->irq_domain); + + phy_power_off(rockchip->phy); + phy_exit(rockchip->phy); + + clk_disable_unprepare(rockchip->clk_pcie_pm); + clk_disable_unprepare(rockchip->hclk_pcie); + clk_disable_unprepare(rockchip->aclk_perf_pcie); + clk_disable_unprepare(rockchip->aclk_pcie); + + if (!IS_ERR(rockchip->vpcie3v3)) + regulator_disable(rockchip->vpcie3v3); + if (!IS_ERR(rockchip->vpcie1v8)) + regulator_disable(rockchip->vpcie1v8); + if (!IS_ERR(rockchip->vpcie0v9)) + regulator_disable(rockchip->vpcie0v9); + + return 0; +} + static const struct dev_pm_ops rockchip_pcie_pm_ops = { SET_NOIRQ_SYSTEM_SLEEP_PM_OPS(rockchip_pcie_suspend_noirq, rockchip_pcie_resume_noirq) @@ -1444,6 +1476,6 @@ static struct platform_driver rockchip_pcie_driver = { .pm = &rockchip_pcie_pm_ops, }, .probe = rockchip_pcie_probe, - + .remove = rockchip_pcie_remove, }; builtin_platform_driver(rockchip_pcie_driver); -- cgit v1.2.3-59-g8ed1b From f90b0875463e450c5dcb714fb887720ecb53a600 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 9 Mar 2017 18:46:16 -0800 Subject: PCI: Export pci_remap_iospace() and pci_unmap_iospace() These are useful for PCIe host drivers, and those drivers can be modules. [bhelgaas: don't remove __weak; it's removed elsewhere] Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas Acked-by: Shawn Lin --- drivers/pci/pci.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 7904d02ffdb9..c87d1edf0203 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -3383,6 +3383,7 @@ int __weak pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr) return -ENODEV; #endif } +EXPORT_SYMBOL(pci_remap_iospace); /** * pci_unmap_iospace - Unmap the memory mapped I/O space @@ -3400,6 +3401,7 @@ void pci_unmap_iospace(struct resource *res) unmap_kernel_range(vaddr, resource_size(res)); #endif } +EXPORT_SYMBOL(pci_unmap_iospace); static void __pci_set_master(struct pci_dev *dev, bool enable) { -- cgit v1.2.3-59-g8ed1b From b0308c5419607c77e5be3c9d388b120ed471037b Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 9 Mar 2017 18:46:17 -0800 Subject: PCI: rockchip: Modularize Now that we've exported pci_remap_iospace() and added proper remove() support, there's no reason this can't be a loadable module. Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas Acked-by: Shawn Lin --- drivers/pci/host/Kconfig | 2 +- drivers/pci/host/pcie-rockchip.c | 8 +++++++- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig index f7c1d4d5c665..d2293ed81cf9 100644 --- a/drivers/pci/host/Kconfig +++ b/drivers/pci/host/Kconfig @@ -164,7 +164,7 @@ config PCI_HOST_THUNDER_ECAM Say Y here if you want ECAM support for CN88XX-Pass-1.x Cavium Thunder SoCs. config PCIE_ROCKCHIP - bool "Rockchip PCIe controller" + tristate "Rockchip PCIe controller" depends on ARCH_ROCKCHIP || COMPILE_TEST depends on OF depends on PCI_MSI_IRQ_DOMAIN diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index 76f2edc93663..abc7fd76b181 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include @@ -1468,6 +1469,7 @@ static const struct of_device_id rockchip_pcie_of_match[] = { { .compatible = "rockchip,rk3399-pcie", }, {} }; +MODULE_DEVICE_TABLE(of, rockchip_pcie_of_match); static struct platform_driver rockchip_pcie_driver = { .driver = { @@ -1478,4 +1480,8 @@ static struct platform_driver rockchip_pcie_driver = { .probe = rockchip_pcie_probe, .remove = rockchip_pcie_remove, }; -builtin_platform_driver(rockchip_pcie_driver); +module_platform_driver(rockchip_pcie_driver); + +MODULE_AUTHOR("Rockchip Inc"); +MODULE_DESCRIPTION("Rockchip AXI PCIe driver"); +MODULE_LICENSE("GPL v2"); -- cgit v1.2.3-59-g8ed1b From ced414a14f709fc0af60bd381ba8a566dc566869 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Fri, 21 Apr 2017 11:42:54 -0500 Subject: PCI/ACPI: Tidy up MCFG quirk whitespace With no blank lines, it's not obvious where the macro definitions end and the uses begin. Add some blank lines and reorder the ThunderX definitions. No functional change intended. Signed-off-by: Bjorn Helgaas CC: stable@vger.kernel.org # v4.10+ --- drivers/acpi/pci_mcfg.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c index 2944353253ed..65affd8f29c1 100644 --- a/drivers/acpi/pci_mcfg.c +++ b/drivers/acpi/pci_mcfg.c @@ -54,6 +54,7 @@ static struct mcfg_fixup mcfg_quirks[] = { #define QCOM_ECAM32(seg) \ { "QCOM ", "QDF2432 ", 1, seg, MCFG_BUS_ANY, &pci_32b_ops } + QCOM_ECAM32(0), QCOM_ECAM32(1), QCOM_ECAM32(2), @@ -68,6 +69,7 @@ static struct mcfg_fixup mcfg_quirks[] = { { "HISI ", table_id, 0, (seg) + 1, MCFG_BUS_ANY, ops }, \ { "HISI ", table_id, 0, (seg) + 2, MCFG_BUS_ANY, ops }, \ { "HISI ", table_id, 0, (seg) + 3, MCFG_BUS_ANY, ops } + HISI_QUAD_DOM("HIP05 ", 0, &hisi_pcie_ops), HISI_QUAD_DOM("HIP06 ", 0, &hisi_pcie_ops), HISI_QUAD_DOM("HIP07 ", 0, &hisi_pcie_ops), @@ -77,6 +79,7 @@ static struct mcfg_fixup mcfg_quirks[] = { #define THUNDER_PEM_RES(addr, node) \ DEFINE_RES_MEM((addr) + ((u64) (node) << 44), 0x39 * SZ_16M) + #define THUNDER_PEM_QUIRK(rev, node) \ { "CAVIUM", "THUNDERX", rev, 4 + (10 * (node)), MCFG_BUS_ANY, \ &thunder_pem_ecam_ops, THUNDER_PEM_RES(0x88001f000000UL, node) }, \ @@ -90,13 +93,15 @@ static struct mcfg_fixup mcfg_quirks[] = { &thunder_pem_ecam_ops, THUNDER_PEM_RES(0x894057000000UL, node) }, \ { "CAVIUM", "THUNDERX", rev, 9 + (10 * (node)), MCFG_BUS_ANY, \ &thunder_pem_ecam_ops, THUNDER_PEM_RES(0x89808f000000UL, node) } - /* SoC pass2.x */ - THUNDER_PEM_QUIRK(1, 0), - THUNDER_PEM_QUIRK(1, 1), #define THUNDER_ECAM_QUIRK(rev, seg) \ { "CAVIUM", "THUNDERX", rev, seg, MCFG_BUS_ANY, \ &pci_thunder_ecam_ops } + + /* SoC pass2.x */ + THUNDER_PEM_QUIRK(1, 0), + THUNDER_PEM_QUIRK(1, 1), + /* SoC pass1.x */ THUNDER_PEM_QUIRK(2, 0), /* off-chip devices */ THUNDER_PEM_QUIRK(2, 1), /* off-chip devices */ @@ -112,9 +117,11 @@ static struct mcfg_fixup mcfg_quirks[] = { #define XGENE_V1_ECAM_MCFG(rev, seg) \ {"APM ", "XGENE ", rev, seg, MCFG_BUS_ANY, \ &xgene_v1_pcie_ecam_ops } + #define XGENE_V2_ECAM_MCFG(rev, seg) \ {"APM ", "XGENE ", rev, seg, MCFG_BUS_ANY, \ &xgene_v2_pcie_ecam_ops } + /* X-Gene SoC with v1 PCIe controller */ XGENE_V1_ECAM_MCFG(1, 0), XGENE_V1_ECAM_MCFG(1, 1), -- cgit v1.2.3-59-g8ed1b From cd183740480f045600aa1fa38fe70809b5498f05 Mon Sep 17 00:00:00 2001 From: Tomasz Nowicki Date: Wed, 29 Mar 2017 14:16:13 +0200 Subject: PCI/ACPI: Add ThunderX pass2.x 2nd node MCFG quirk Currently SoCs pass2.x do not emulate EA headers for ACPI boot method at all. However, for pass2.x some devices (like EDAC) advertise incorrect base addresses in their BARs which results in driver probe failure during resource request. Since all problematic blocks are on 2nd NUMA node under domain 10 add necessary quirk entry to obtain BAR addresses correction using EA header emulation. Fixes: 44f22bd91e88 ("PCI: Add MCFG quirks for Cavium ThunderX pass2.x host controller") Signed-off-by: Tomasz Nowicki Signed-off-by: Bjorn Helgaas Acked-by: Robert Richter CC: stable@vger.kernel.org # v4.10+ --- drivers/acpi/pci_mcfg.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/acpi/pci_mcfg.c b/drivers/acpi/pci_mcfg.c index 65affd8f29c1..a4e8432fc2fb 100644 --- a/drivers/acpi/pci_mcfg.c +++ b/drivers/acpi/pci_mcfg.c @@ -101,6 +101,7 @@ static struct mcfg_fixup mcfg_quirks[] = { /* SoC pass2.x */ THUNDER_PEM_QUIRK(1, 0), THUNDER_PEM_QUIRK(1, 1), + THUNDER_ECAM_QUIRK(1, 10), /* SoC pass1.x */ THUNDER_PEM_QUIRK(2, 0), /* off-chip devices */ -- cgit v1.2.3-59-g8ed1b From f1e209b7f80288a711268af9054a04d3f6900a6b Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:52 +0100 Subject: ARM64: Implement pci_remap_cfgspace() interface The PCI bus specification (rev 3.0, 3.2.5 "Transaction Ordering and Posting") defines rules for PCI configuration space transactions ordering and posting, that state that configuration writes are non-posted transactions. This rule is reinforced by the ARM v8 architecture reference manual (issue A.k, Early Write Acknowledgment) that explicitly recommends that No Early Write Acknowledgment attribute should be used to map PCI configuration (write) transactions. Current ioremap interface on ARM64 implements mapping functions where the Early Write Acknowledgment hint is enabled, so they cannot be used to map PCI configuration space in a PCI specs compliant way. Implement an ARM64 specific pci_remap_cfgspace() interface that allows to map PCI config region with nGnRnE attributes, providing a remap function that complies with PCI specifications and the ARMv8 architecture reference manual recommendations. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Acked-by: Catalin Marinas Cc: Will Deacon --- arch/arm64/include/asm/io.h | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/arch/arm64/include/asm/io.h b/arch/arm64/include/asm/io.h index 0c00c87bb9dd..35b2e50f17fb 100644 --- a/arch/arm64/include/asm/io.h +++ b/arch/arm64/include/asm/io.h @@ -172,6 +172,16 @@ extern void __iomem *ioremap_cache(phys_addr_t phys_addr, size_t size); #define ioremap_wt(addr, size) __ioremap((addr), (size), __pgprot(PROT_DEVICE_nGnRE)) #define iounmap __iounmap +/* + * PCI configuration space mapping function. + * + * The PCI specification disallows posted write configuration transactions. + * Add an arch specific pci_remap_cfgspace() definition that is implemented + * through nGnRnE device memory attribute as recommended by the ARM v8 + * Architecture reference manual Issue A.k B2.8.2 "Device memory". + */ +#define pci_remap_cfgspace(addr, size) __ioremap((addr), (size), __pgprot(PROT_DEVICE_nGnRnE)) + /* * io{read,write}{16,32,64}be() macros */ -- cgit v1.2.3-59-g8ed1b From b9cdbe6e39351f0ba6cc0c5bc218443f0898e123 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:53 +0100 Subject: ARM: Implement pci_remap_cfgspace() interface The PCI bus specification (rev 3.0, 3.2.5 "Transaction Ordering and Posting") defines rules for PCI configuration space transactions ordering and posting, that state that configuration writes have to be non-posted transactions. Current ioremap interface on ARM provides mapping functions that provide "bufferable" writes transactions (ie ioremap uses MT_DEVICE memory type) aka posted writes, so PCI host controller drivers have no arch interface to remap PCI configuration space with memory attributes that comply with the PCI specifications for configuration space. Implement an ARM specific pci_remap_cfgspace() interface that allows to map PCI config memory regions with MT_UNCACHED memory type (ie strongly ordered - non-posted writes), providing a remap function that complies with PCI specifications for config space transactions. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Arnd Bergmann Cc: Russell King --- arch/arm/include/asm/io.h | 10 ++++++++++ arch/arm/mm/ioremap.c | 7 +++++++ arch/arm/mm/nommu.c | 12 ++++++++++++ 3 files changed, 29 insertions(+) diff --git a/arch/arm/include/asm/io.h b/arch/arm/include/asm/io.h index 42871fb8340e..2cfbc531f63b 100644 --- a/arch/arm/include/asm/io.h +++ b/arch/arm/include/asm/io.h @@ -186,6 +186,16 @@ static inline void pci_ioremap_set_mem_type(int mem_type) {} extern int pci_ioremap_io(unsigned int offset, phys_addr_t phys_addr); +/* + * PCI configuration space mapping function. + * + * The PCI specification does not allow configuration write + * transactions to be posted. Add an arch specific + * pci_remap_cfgspace() definition that is implemented + * through strongly ordered memory mappings. + */ +#define pci_remap_cfgspace pci_remap_cfgspace +void __iomem *pci_remap_cfgspace(resource_size_t res_cookie, size_t size); /* * Now, pick up the machine-defined IO definitions */ diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c index ff0eed23ddf1..fc91205ff46c 100644 --- a/arch/arm/mm/ioremap.c +++ b/arch/arm/mm/ioremap.c @@ -481,6 +481,13 @@ int pci_ioremap_io(unsigned int offset, phys_addr_t phys_addr) __pgprot(get_mem_type(pci_ioremap_mem_type)->prot_pte)); } EXPORT_SYMBOL_GPL(pci_ioremap_io); + +void __iomem *pci_remap_cfgspace(resource_size_t res_cookie, size_t size) +{ + return arch_ioremap_caller(res_cookie, size, MT_UNCACHED, + __builtin_return_address(0)); +} +EXPORT_SYMBOL_GPL(pci_remap_cfgspace); #endif /* diff --git a/arch/arm/mm/nommu.c b/arch/arm/mm/nommu.c index 3b5c7aaf9c76..2a3f6002efbc 100644 --- a/arch/arm/mm/nommu.c +++ b/arch/arm/mm/nommu.c @@ -433,6 +433,18 @@ void __iomem *ioremap_wc(resource_size_t res_cookie, size_t size) } EXPORT_SYMBOL(ioremap_wc); +#ifdef CONFIG_PCI + +#include + +void __iomem *pci_remap_cfgspace(resource_size_t res_cookie, size_t size) +{ + return arch_ioremap_caller(res_cookie, size, MT_UNCACHED, + __builtin_return_address(0)); +} +EXPORT_SYMBOL_GPL(pci_remap_cfgspace); +#endif + void *arch_memremap_wb(phys_addr_t phys_addr, size_t size) { return (void *)phys_addr; -- cgit v1.2.3-59-g8ed1b From 6524754eff305760d0b31dac468e78057a56f47e Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:54 +0100 Subject: devres: fix devm_ioremap_*() offset parameter kerneldoc description The offset parameter in the devres devm_ioremap_*() functions kerneldoc entries is erroneously defined as BUS offset whereas it is actually a resource address. Since it is actually misleading, fix the devres devm_ioremap_* offset parameter kerneldoc entry by replacing BUS offset with a more suitable description (ie Resource address). Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Tejun Heo --- lib/devres.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/lib/devres.c b/lib/devres.c index cb1464c411a2..78eca713b1d9 100644 --- a/lib/devres.c +++ b/lib/devres.c @@ -17,7 +17,7 @@ static int devm_ioremap_match(struct device *dev, void *res, void *match_data) /** * devm_ioremap - Managed ioremap() * @dev: Generic device to remap IO address for - * @offset: BUS offset to map + * @offset: Resource address to map * @size: Size of map * * Managed ioremap(). Map is automatically unmapped on driver detach. @@ -45,7 +45,7 @@ EXPORT_SYMBOL(devm_ioremap); /** * devm_ioremap_nocache - Managed ioremap_nocache() * @dev: Generic device to remap IO address for - * @offset: BUS offset to map + * @offset: Resource address to map * @size: Size of map * * Managed ioremap_nocache(). Map is automatically unmapped on driver @@ -74,7 +74,7 @@ EXPORT_SYMBOL(devm_ioremap_nocache); /** * devm_ioremap_wc - Managed ioremap_wc() * @dev: Generic device to remap IO address for - * @offset: BUS offset to map + * @offset: Resource address to map * @size: Size of map * * Managed ioremap_wc(). Map is automatically unmapped on driver detach. -- cgit v1.2.3-59-g8ed1b From 490cb6ddb17df5ef5f5eb33c9a34f3033b31c204 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:55 +0100 Subject: PCI: Implement devm_pci_remap_cfgspace() The introduction of the pci_remap_cfgspace() interface allows PCI host controller drivers to map PCI config space through a dedicated kernel interface. Current PCI host controller drivers use the devm_ioremap_*() devres interfaces to map PCI configuration space regions so in order to update them to the new pci_remap_cfgspace() mapping interface a new set of devres interfaces should be implemented so that PCI host controller drivers can make use of them. Introduce two new functions in the PCI kernel layer and Devres documentation: - devm_pci_remap_cfgspace() - devm_pci_remap_cfg_resource() so that PCI host controller drivers can make use of them to map PCI configuration space regions. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Jonathan Corbet --- Documentation/driver-model/devres.txt | 6 ++- drivers/pci/pci.c | 82 +++++++++++++++++++++++++++++++++++ include/linux/pci.h | 5 +++ 3 files changed, 91 insertions(+), 2 deletions(-) diff --git a/Documentation/driver-model/devres.txt b/Documentation/driver-model/devres.txt index bf34d5b3a733..e72587fe477d 100644 --- a/Documentation/driver-model/devres.txt +++ b/Documentation/driver-model/devres.txt @@ -342,8 +342,10 @@ PER-CPU MEM devm_free_percpu() PCI - pcim_enable_device() : after success, all PCI ops become managed - pcim_pin_device() : keep PCI device enabled after release + devm_pci_remap_cfgspace() : ioremap PCI configuration space + devm_pci_remap_cfg_resource() : ioremap PCI configuration space resource + pcim_enable_device() : after success, all PCI ops become managed + pcim_pin_device() : keep PCI device enabled after release PHY devm_usb_get_phy() diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index bd98674a0419..4129f9453861 100644 --- a/drivers/pci/pci.c +++ b/drivers/pci/pci.c @@ -3401,6 +3401,88 @@ void pci_unmap_iospace(struct resource *res) #endif } +/** + * devm_pci_remap_cfgspace - Managed pci_remap_cfgspace() + * @dev: Generic device to remap IO address for + * @offset: Resource address to map + * @size: Size of map + * + * Managed pci_remap_cfgspace(). Map is automatically unmapped on driver + * detach. + */ +void __iomem *devm_pci_remap_cfgspace(struct device *dev, + resource_size_t offset, + resource_size_t size) +{ + void __iomem **ptr, *addr; + + ptr = devres_alloc(devm_ioremap_release, sizeof(*ptr), GFP_KERNEL); + if (!ptr) + return NULL; + + addr = pci_remap_cfgspace(offset, size); + if (addr) { + *ptr = addr; + devres_add(dev, ptr); + } else + devres_free(ptr); + + return addr; +} +EXPORT_SYMBOL(devm_pci_remap_cfgspace); + +/** + * devm_pci_remap_cfg_resource - check, request region and ioremap cfg resource + * @dev: generic device to handle the resource for + * @res: configuration space resource to be handled + * + * Checks that a resource is a valid memory region, requests the memory + * region and ioremaps with pci_remap_cfgspace() API that ensures the + * proper PCI configuration space memory attributes are guaranteed. + * + * All operations are managed and will be undone on driver detach. + * + * Returns a pointer to the remapped memory or an ERR_PTR() encoded error code + * on failure. Usage example: + * + * res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + * base = devm_pci_remap_cfg_resource(&pdev->dev, res); + * if (IS_ERR(base)) + * return PTR_ERR(base); + */ +void __iomem *devm_pci_remap_cfg_resource(struct device *dev, + struct resource *res) +{ + resource_size_t size; + const char *name; + void __iomem *dest_ptr; + + BUG_ON(!dev); + + if (!res || resource_type(res) != IORESOURCE_MEM) { + dev_err(dev, "invalid resource\n"); + return IOMEM_ERR_PTR(-EINVAL); + } + + size = resource_size(res); + name = res->name ?: dev_name(dev); + + if (!devm_request_mem_region(dev, res->start, size, name)) { + dev_err(dev, "can't request region for resource %pR\n", res); + return IOMEM_ERR_PTR(-EBUSY); + } + + dest_ptr = devm_pci_remap_cfgspace(dev, res->start, size); + if (!dest_ptr) { + dev_err(dev, "ioremap failed for resource %pR\n", res); + devm_release_mem_region(dev, res->start, size); + dest_ptr = IOMEM_ERR_PTR(-ENOMEM); + } + + return dest_ptr; +} +EXPORT_SYMBOL(devm_pci_remap_cfg_resource); + static void __pci_set_master(struct pci_dev *dev, bool enable) { u16 old_cmd, cmd; diff --git a/include/linux/pci.h b/include/linux/pci.h index eb3da1a04e6c..70534d66d18a 100644 --- a/include/linux/pci.h +++ b/include/linux/pci.h @@ -1199,6 +1199,11 @@ unsigned long pci_address_to_pio(phys_addr_t addr); phys_addr_t pci_pio_to_address(unsigned long pio); int pci_remap_iospace(const struct resource *res, phys_addr_t phys_addr); void pci_unmap_iospace(struct resource *res); +void __iomem *devm_pci_remap_cfgspace(struct device *dev, + resource_size_t offset, + resource_size_t size); +void __iomem *devm_pci_remap_cfg_resource(struct device *dev, + struct resource *res); static inline pci_bus_addr_t pci_bus_address(struct pci_dev *pdev, int bar) { -- cgit v1.2.3-59-g8ed1b From 053497cec7daecd50313190786872191586ff68c Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:56 +0100 Subject: PCI: ECAM: Map config region with pci_remap_cfgspace() The current ECAM kernel implementation uses ioremap() to map the ECAM configuration space memory region; this is not safe in that on some architectures the ioremap interface provides mappings that allow posted write transactions. This, as highlighted in the PCIe specifications (4.0 - Rev0.3, "Ordering Considerations for the Enhanced Configuration Address Mechanism"), can create ordering issues for software because posted writes transactions on the CPU host bus are non posted in the PCI express fabric. Update the ioremap() interface to use pci_remap_cfgspace() whose mapping attributes guarantee that non-posted writes transactions are issued for memory writes within the ECAM memory mapped address region. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Jayachandran C --- drivers/pci/ecam.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/pci/ecam.c b/drivers/pci/ecam.c index 2fee61bb6559..c228a2eb7faa 100644 --- a/drivers/pci/ecam.c +++ b/drivers/pci/ecam.c @@ -84,12 +84,14 @@ struct pci_config_window *pci_ecam_create(struct device *dev, if (!cfg->winp) goto err_exit_malloc; for (i = 0; i < bus_range; i++) { - cfg->winp[i] = ioremap(cfgres->start + i * bsz, bsz); + cfg->winp[i] = + pci_remap_cfgspace(cfgres->start + i * bsz, + bsz); if (!cfg->winp[i]) goto err_exit_iomap; } } else { - cfg->win = ioremap(cfgres->start, bus_range * bsz); + cfg->win = pci_remap_cfgspace(cfgres->start, bus_range * bsz); if (!cfg->win) goto err_exit_iomap; } -- cgit v1.2.3-59-g8ed1b From eb21f9e82009b8bce01aa242f15c1acbebbd76e2 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:57 +0100 Subject: PCI: xilinx: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Bharat Kumar Gogada Cc: Michal Simek --- drivers/pci/host/pcie-xilinx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-xilinx.c b/drivers/pci/host/pcie-xilinx.c index 7f030f5d750b..2fe2df51f9f8 100644 --- a/drivers/pci/host/pcie-xilinx.c +++ b/drivers/pci/host/pcie-xilinx.c @@ -606,7 +606,7 @@ static int xilinx_pcie_parse_dt(struct xilinx_pcie_port *port) return err; } - port->reg_base = devm_ioremap_resource(dev, ®s); + port->reg_base = devm_pci_remap_cfg_resource(dev, ®s); if (IS_ERR(port->reg_base)) return PTR_ERR(port->reg_base); -- cgit v1.2.3-59-g8ed1b From cd00f084ed1d501c8cb391c6d98b4f31b2a9a3e2 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:58 +0100 Subject: PCI: xilinx-nwl: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Bharat Kumar Gogada Cc: Michal Simek --- drivers/pci/host/pcie-xilinx-nwl.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-xilinx-nwl.c b/drivers/pci/host/pcie-xilinx-nwl.c index 4c3e0ab35496..4b16b26ae909 100644 --- a/drivers/pci/host/pcie-xilinx-nwl.c +++ b/drivers/pci/host/pcie-xilinx-nwl.c @@ -761,7 +761,7 @@ static int nwl_pcie_parse_dt(struct nwl_pcie *pcie, pcie->phys_pcie_reg_base = res->start; res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "cfg"); - pcie->ecam_base = devm_ioremap_resource(dev, res); + pcie->ecam_base = devm_pci_remap_cfg_resource(dev, res); if (IS_ERR(pcie->ecam_base)) return PTR_ERR(pcie->ecam_base); pcie->phys_ecam_base = res->start; -- cgit v1.2.3-59-g8ed1b From 05043c89ef067ec3c067e89afad8c98974e7a234 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:48:59 +0100 Subject: PCI: spear13xx: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generate on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Pratyush Anand --- drivers/pci/dwc/pcie-spear13xx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/dwc/pcie-spear13xx.c b/drivers/pci/dwc/pcie-spear13xx.c index eaa4ea8e2ea4..3ae59de5c043 100644 --- a/drivers/pci/dwc/pcie-spear13xx.c +++ b/drivers/pci/dwc/pcie-spear13xx.c @@ -273,7 +273,7 @@ static int spear13xx_pcie_probe(struct platform_device *pdev) } dbi_base = platform_get_resource_byname(pdev, IORESOURCE_MEM, "dbi"); - pci->dbi_base = devm_ioremap_resource(dev, dbi_base); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, dbi_base); if (IS_ERR(pci->dbi_base)) { dev_err(dev, "couldn't remap dbi base %p\n", dbi_base); ret = PTR_ERR(pci->dbi_base); -- cgit v1.2.3-59-g8ed1b From 995b76e48fe3cba4820ca42b4e0726cc5715a952 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:00 +0100 Subject: PCI: rockchip: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Wenrui Li Cc: Shawn Lin --- drivers/pci/host/pcie-rockchip.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-rockchip.c b/drivers/pci/host/pcie-rockchip.c index 26ddd3535272..690a7b508d64 100644 --- a/drivers/pci/host/pcie-rockchip.c +++ b/drivers/pci/host/pcie-rockchip.c @@ -822,7 +822,7 @@ static int rockchip_pcie_parse_dt(struct rockchip_pcie *rockchip) regs = platform_get_resource_byname(pdev, IORESOURCE_MEM, "axi-base"); - rockchip->reg_base = devm_ioremap_resource(dev, regs); + rockchip->reg_base = devm_pci_remap_cfg_resource(dev, regs); if (IS_ERR(rockchip->reg_base)) return PTR_ERR(rockchip->reg_base); -- cgit v1.2.3-59-g8ed1b From 10c736f784cb0dad0dce6ce938878a754676bc89 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:01 +0100 Subject: PCI: qcom: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Stanimir Varbanov --- drivers/pci/dwc/pcie-qcom.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/dwc/pcie-qcom.c b/drivers/pci/dwc/pcie-qcom.c index 67eb7f5926dd..5bf23d432fdb 100644 --- a/drivers/pci/dwc/pcie-qcom.c +++ b/drivers/pci/dwc/pcie-qcom.c @@ -700,7 +700,7 @@ static int qcom_pcie_probe(struct platform_device *pdev) return PTR_ERR(pcie->parf); res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "dbi"); - pci->dbi_base = devm_ioremap_resource(dev, res); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, res); if (IS_ERR(pci->dbi_base)) return PTR_ERR(pci->dbi_base); -- cgit v1.2.3-59-g8ed1b From 868564da658b58ab8864124428fa8fff970bf2bd Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:02 +0100 Subject: PCI: iproc-platform: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Ray Jui Cc: Jon Mason --- drivers/pci/host/pcie-iproc-platform.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pci/host/pcie-iproc-platform.c b/drivers/pci/host/pcie-iproc-platform.c index f4909bb0b2ad..b48d0dbd7620 100644 --- a/drivers/pci/host/pcie-iproc-platform.c +++ b/drivers/pci/host/pcie-iproc-platform.c @@ -67,7 +67,8 @@ static int iproc_pcie_pltfm_probe(struct platform_device *pdev) return ret; } - pcie->base = devm_ioremap(dev, reg.start, resource_size(®)); + pcie->base = devm_pci_remap_cfgspace(dev, reg.start, + resource_size(®)); if (!pcie->base) { dev_err(dev, "unable to map controller registers\n"); return -ENOMEM; -- cgit v1.2.3-59-g8ed1b From cc7b0d495589e03607a925a60a3c969debbc4746 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:03 +0100 Subject: PCI: designware: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Jingoo Han Cc: Joao Pinto --- drivers/pci/dwc/pcie-designware-host.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/pci/dwc/pcie-designware-host.c b/drivers/pci/dwc/pcie-designware-host.c index 5ba334938b52..2b789af5dbbc 100644 --- a/drivers/pci/dwc/pcie-designware-host.c +++ b/drivers/pci/dwc/pcie-designware-host.c @@ -338,8 +338,9 @@ int dw_pcie_host_init(struct pcie_port *pp) } if (!pci->dbi_base) { - pci->dbi_base = devm_ioremap(dev, pp->cfg->start, - resource_size(pp->cfg)); + pci->dbi_base = devm_pci_remap_cfgspace(dev, + pp->cfg->start, + resource_size(pp->cfg)); if (!pci->dbi_base) { dev_err(dev, "error with ioremap\n"); ret = -ENOMEM; @@ -350,8 +351,8 @@ int dw_pcie_host_init(struct pcie_port *pp) pp->mem_base = pp->mem->start; if (!pp->va_cfg0_base) { - pp->va_cfg0_base = devm_ioremap(dev, pp->cfg0_base, - pp->cfg0_size); + pp->va_cfg0_base = devm_pci_remap_cfgspace(dev, + pp->cfg0_base, pp->cfg0_size); if (!pp->va_cfg0_base) { dev_err(dev, "error with ioremap in function\n"); ret = -ENOMEM; @@ -360,7 +361,8 @@ int dw_pcie_host_init(struct pcie_port *pp) } if (!pp->va_cfg1_base) { - pp->va_cfg1_base = devm_ioremap(dev, pp->cfg1_base, + pp->va_cfg1_base = devm_pci_remap_cfgspace(dev, + pp->cfg1_base, pp->cfg1_size); if (!pp->va_cfg1_base) { dev_err(dev, "error with ioremap\n"); -- cgit v1.2.3-59-g8ed1b From 53dfa17dfed84fe1adf00a4ff817789737ead410 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:04 +0100 Subject: PCI: armada8k: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Thomas Petazzoni --- drivers/pci/dwc/pcie-armada8k.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/dwc/pcie-armada8k.c b/drivers/pci/dwc/pcie-armada8k.c index f110e3b24a26..3ff31301323b 100644 --- a/drivers/pci/dwc/pcie-armada8k.c +++ b/drivers/pci/dwc/pcie-armada8k.c @@ -230,7 +230,7 @@ static int armada8k_pcie_probe(struct platform_device *pdev) /* Get the dw-pcie unit configuration/control registers base. */ base = platform_get_resource_byname(pdev, IORESOURCE_MEM, "ctrl"); - pci->dbi_base = devm_ioremap_resource(dev, base); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, base); if (IS_ERR(pci->dbi_base)) { dev_err(dev, "couldn't remap regs base %p\n", base); ret = PTR_ERR(pci->dbi_base); -- cgit v1.2.3-59-g8ed1b From 26b758f775a368653b9ecb3f7c06b9ec34ccb1be Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:05 +0100 Subject: PCI: xgene: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Tanmay Inamdar --- drivers/pci/host/pci-xgene.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/host/pci-xgene.c b/drivers/pci/host/pci-xgene.c index 1a6108788f6f..de198980432e 100644 --- a/drivers/pci/host/pci-xgene.c +++ b/drivers/pci/host/pci-xgene.c @@ -248,7 +248,7 @@ static int xgene_pcie_ecam_init(struct pci_config_window *cfg, u32 ipversion) dev_err(dev, "can't get CSR resource\n"); return ret; } - port->csr_base = devm_ioremap_resource(dev, &csr); + port->csr_base = devm_pci_remap_cfg_resource(dev, &csr); if (IS_ERR(port->csr_base)) return PTR_ERR(port->csr_base); @@ -359,7 +359,7 @@ static int xgene_pcie_map_reg(struct xgene_pcie_port *port, struct resource *res; res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "csr"); - port->csr_base = devm_ioremap_resource(dev, res); + port->csr_base = devm_pci_remap_cfg_resource(dev, res); if (IS_ERR(port->csr_base)) return PTR_ERR(port->csr_base); -- cgit v1.2.3-59-g8ed1b From 3e02dc439b2bcb9911abb1632029a15590ba3b97 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:06 +0100 Subject: PCI: tegra: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use correct memory mapping attributes to map config space regions to enforce configuration space non-posted writes behaviour. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Thierry Reding --- drivers/pci/host/pci-tegra.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/host/pci-tegra.c b/drivers/pci/host/pci-tegra.c index ed8a93f2bfb5..2618f875a600 100644 --- a/drivers/pci/host/pci-tegra.c +++ b/drivers/pci/host/pci-tegra.c @@ -380,7 +380,7 @@ static struct tegra_pcie_bus *tegra_pcie_bus_alloc(struct tegra_pcie *pcie, unsigned int busnr) { struct device *dev = pcie->dev; - pgprot_t prot = pgprot_device(PAGE_KERNEL); + pgprot_t prot = pgprot_noncached(PAGE_KERNEL); phys_addr_t cs = pcie->cs->start; struct tegra_pcie_bus *bus; unsigned int i; @@ -1962,7 +1962,7 @@ static int tegra_pcie_parse_dt(struct tegra_pcie *pcie) rp->pcie = pcie; rp->np = port; - rp->base = devm_ioremap_resource(dev, &rp->regs); + rp->base = devm_pci_remap_cfg_resource(dev, &rp->regs); if (IS_ERR(rp->base)) return PTR_ERR(rp->base); -- cgit v1.2.3-59-g8ed1b From e313a447e73527ef99bd54bc58071956970ca491 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:07 +0100 Subject: PCI: hisi: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Gabriele Paoloni Signed-off-by: Bjorn Helgaas Cc: Zhou Wang --- drivers/pci/dwc/pcie-hisi.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/pci/dwc/pcie-hisi.c b/drivers/pci/dwc/pcie-hisi.c index fd66a3199db7..1606de5bf757 100644 --- a/drivers/pci/dwc/pcie-hisi.c +++ b/drivers/pci/dwc/pcie-hisi.c @@ -99,7 +99,7 @@ static int hisi_pcie_init(struct pci_config_window *cfg) return -ENOMEM; } - reg_base = devm_ioremap(dev, res->start, resource_size(res)); + reg_base = devm_pci_remap_cfgspace(dev, res->start, resource_size(res)); if (!reg_base) return -ENOMEM; @@ -296,10 +296,9 @@ static int hisi_pcie_probe(struct platform_device *pdev) } reg = platform_get_resource_byname(pdev, IORESOURCE_MEM, "rc_dbi"); - pci->dbi_base = devm_ioremap_resource(dev, reg); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, reg); if (IS_ERR(pci->dbi_base)) return PTR_ERR(pci->dbi_base); - platform_set_drvdata(pdev, hisi_pcie); ret = hisi_add_pcie_port(hisi_pcie, pdev); @@ -360,7 +359,7 @@ static int hisi_pcie_platform_init(struct pci_config_window *cfg) return -EINVAL; } - reg_base = devm_ioremap(dev, res->start, resource_size(res)); + reg_base = devm_pci_remap_cfgspace(dev, res->start, resource_size(res)); if (!reg_base) return -ENOMEM; -- cgit v1.2.3-59-g8ed1b From 01bd489dba01bcc58b23c669b437330039639c75 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:08 +0100 Subject: PCI: layerscape: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Mingkai Hu Cc: Minghuan Lian Cc: Roy Zang --- drivers/pci/dwc/pci-layerscape.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/dwc/pci-layerscape.c b/drivers/pci/dwc/pci-layerscape.c index c32e392a0ae6..8f0ee0d8129c 100644 --- a/drivers/pci/dwc/pci-layerscape.c +++ b/drivers/pci/dwc/pci-layerscape.c @@ -283,7 +283,7 @@ static int __init ls_pcie_probe(struct platform_device *pdev) pcie->pci = pci; dbi_base = platform_get_resource_byname(pdev, IORESOURCE_MEM, "regs"); - pci->dbi_base = devm_ioremap_resource(dev, dbi_base); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, dbi_base); if (IS_ERR(pci->dbi_base)) return PTR_ERR(pci->dbi_base); -- cgit v1.2.3-59-g8ed1b From 89874a1a6e64b6be8e442b1d52a447e2b730ebbd Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:09 +0100 Subject: PCI: keystone-dw: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_pci_remap_cfg* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Murali Karicheri --- drivers/pci/dwc/pci-keystone-dw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/dwc/pci-keystone-dw.c b/drivers/pci/dwc/pci-keystone-dw.c index 6b396f6b4615..8bc626e640c8 100644 --- a/drivers/pci/dwc/pci-keystone-dw.c +++ b/drivers/pci/dwc/pci-keystone-dw.c @@ -543,7 +543,7 @@ int __init ks_dw_pcie_host_init(struct keystone_pcie *ks_pcie, /* Index 0 is the config reg. space address */ res = platform_get_resource(pdev, IORESOURCE_MEM, 0); - pci->dbi_base = devm_ioremap_resource(dev, res); + pci->dbi_base = devm_pci_remap_cfg_resource(dev, res); if (IS_ERR(pci->dbi_base)) return PTR_ERR(pci->dbi_base); -- cgit v1.2.3-59-g8ed1b From bc636ee94501aedef744f96c57bb2e29cf21aa39 Mon Sep 17 00:00:00 2001 From: Lorenzo Pieralisi Date: Wed, 19 Apr 2017 17:49:10 +0100 Subject: PCI: versatile: Update PCI config space remap function PCI configuration space should be mapped with a memory region type that generates on the CPU host bus non-posted write transations. Update the driver to use the devm_ioremap_nopost* interface to make sure the correct memory mappings for PCI configuration space are used. Signed-off-by: Lorenzo Pieralisi Signed-off-by: Bjorn Helgaas Cc: Rob Herring --- drivers/pci/host/pci-versatile.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/pci/host/pci-versatile.c b/drivers/pci/host/pci-versatile.c index 5ebee7d37ff5..85e773661bc8 100644 --- a/drivers/pci/host/pci-versatile.c +++ b/drivers/pci/host/pci-versatile.c @@ -138,7 +138,8 @@ static int versatile_pci_probe(struct platform_device *pdev) return PTR_ERR(versatile_cfg_base[0]); res = platform_get_resource(pdev, IORESOURCE_MEM, 2); - versatile_cfg_base[1] = devm_ioremap_resource(&pdev->dev, res); + versatile_cfg_base[1] = devm_pci_remap_cfg_resource(&pdev->dev, + res); if (IS_ERR(versatile_cfg_base[1])) return PTR_ERR(versatile_cfg_base[1]); -- cgit v1.2.3-59-g8ed1b From c4e649b09f55595e6df6da5465a5b3cfc93557c1 Mon Sep 17 00:00:00 2001 From: Stefan Assmann Date: Wed, 19 Apr 2017 09:22:45 +0200 Subject: PCI: Disable boot interrupt quirk for ASUS M2N-LR The ASUS M2N-LR should not trigger boot interrupt quirks although it carries an Intel 6702PXH. On this board the boot interrupt quirks cause incorrect IRQ assignments and should be disabled. Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=43074 Tested-by: Solomon Peachy Signed-off-by: Stefan Assmann Signed-off-by: Bjorn Helgaas --- drivers/pci/quirks.c | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c index f754453fe754..a1920b5e6d31 100644 --- a/drivers/pci/quirks.c +++ b/drivers/pci/quirks.c @@ -1685,6 +1685,29 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x260a, quirk_intel_pcie_pm); DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x260b, quirk_intel_pcie_pm); #ifdef CONFIG_X86_IO_APIC +static int dmi_disable_ioapicreroute(const struct dmi_system_id *d) +{ + noioapicreroute = 1; + pr_info("%s detected: disable boot interrupt reroute\n", d->ident); + + return 0; +} + +static struct dmi_system_id boot_interrupt_dmi_table[] = { + /* + * Systems to exclude from boot interrupt reroute quirks + */ + { + .callback = dmi_disable_ioapicreroute, + .ident = "ASUSTek Computer INC. M2N-LR", + .matches = { + DMI_MATCH(DMI_SYS_VENDOR, "ASUSTek Computer INC."), + DMI_MATCH(DMI_PRODUCT_NAME, "M2N-LR"), + }, + }, + {} +}; + /* * Boot interrupts on some chipsets cannot be turned off. For these chipsets, * remap the original interrupt in the linux kernel to the boot interrupt, so @@ -1693,6 +1716,7 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_INTEL, 0x260b, quirk_intel_pcie_pm); */ static void quirk_reroute_to_boot_interrupts_intel(struct pci_dev *dev) { + dmi_check_system(boot_interrupt_dmi_table); if (noioapicquirk || noioapicreroute) return; -- cgit v1.2.3-59-g8ed1b From 7221547e55b7929e4d46983f6f3ca15f36ee4dac Mon Sep 17 00:00:00 2001 From: Colin Ian King Date: Fri, 21 Apr 2017 08:02:30 +0100 Subject: PCI: imx6: Fix spelling mistake: "contol" -> "control" Trivial fix to spelling mistake in dev_err message Signed-off-by: Colin Ian King Signed-off-by: Bjorn Helgaas Acked-by: Richard Zhu --- drivers/pci/dwc/pci-imx6.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index 102edcf1e261..129717ae5022 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -726,13 +726,13 @@ static int imx6_pcie_probe(struct platform_device *pdev) imx6_pcie->pciephy_reset = devm_reset_control_get(dev, "pciephy"); if (IS_ERR(imx6_pcie->pciephy_reset)) { - dev_err(dev, "Failed to get PCIEPHY reset contol\n"); + dev_err(dev, "Failed to get PCIEPHY reset control\n"); return PTR_ERR(imx6_pcie->pciephy_reset); } imx6_pcie->apps_reset = devm_reset_control_get(dev, "apps"); if (IS_ERR(imx6_pcie->apps_reset)) { - dev_err(dev, "Failed to get PCIE APPS reset contol\n"); + dev_err(dev, "Failed to get PCIE APPS reset control\n"); return PTR_ERR(imx6_pcie->apps_reset); } break; -- cgit v1.2.3-59-g8ed1b From 21c433a74b6b94538d01ea18d66195018136efd6 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 25 Apr 2017 14:36:19 -0500 Subject: IB/hfi1: Use pcie_flr() instead of duplicating it Tested-by: Jakub Byczkowski Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Acked-by: Doug Ledford --- drivers/infiniband/hw/hfi1/chip.c | 4 ++-- drivers/infiniband/hw/hfi1/hfi.h | 1 - drivers/infiniband/hw/hfi1/pcie.c | 30 ------------------------------ 3 files changed, 2 insertions(+), 33 deletions(-) diff --git a/drivers/infiniband/hw/hfi1/chip.c b/drivers/infiniband/hw/hfi1/chip.c index 121a4c920f1b..d037f72e4d96 100644 --- a/drivers/infiniband/hw/hfi1/chip.c +++ b/drivers/infiniband/hw/hfi1/chip.c @@ -13610,14 +13610,14 @@ static void init_chip(struct hfi1_devdata *dd) dd_dev_info(dd, "Resetting CSRs with FLR\n"); /* do the FLR, the DC reset will remain */ - hfi1_pcie_flr(dd); + pcie_flr(dd->pcidev); /* restore command and BARs */ restore_pci_variables(dd); if (is_ax(dd)) { dd_dev_info(dd, "Resetting CSRs with FLR\n"); - hfi1_pcie_flr(dd); + pcie_flr(dd->pcidev); restore_pci_variables(dd); } } else { diff --git a/drivers/infiniband/hw/hfi1/hfi.h b/drivers/infiniband/hw/hfi1/hfi.h index 0808e3c3ba39..40d7559fa723 100644 --- a/drivers/infiniband/hw/hfi1/hfi.h +++ b/drivers/infiniband/hw/hfi1/hfi.h @@ -1764,7 +1764,6 @@ int hfi1_pcie_init(struct pci_dev *, const struct pci_device_id *); void hfi1_pcie_cleanup(struct pci_dev *); int hfi1_pcie_ddinit(struct hfi1_devdata *, struct pci_dev *); void hfi1_pcie_ddcleanup(struct hfi1_devdata *); -void hfi1_pcie_flr(struct hfi1_devdata *); int pcie_speeds(struct hfi1_devdata *); void request_msix(struct hfi1_devdata *, u32 *, struct hfi1_msix_entry *); void hfi1_enable_intx(struct pci_dev *); diff --git a/drivers/infiniband/hw/hfi1/pcie.c b/drivers/infiniband/hw/hfi1/pcie.c index 0829fce06172..c81556e84831 100644 --- a/drivers/infiniband/hw/hfi1/pcie.c +++ b/drivers/infiniband/hw/hfi1/pcie.c @@ -240,36 +240,6 @@ void hfi1_pcie_ddcleanup(struct hfi1_devdata *dd) iounmap(dd->piobase); } -/* - * Do a Function Level Reset (FLR) on the device. - * Based on static function drivers/pci/pci.c:pcie_flr(). - */ -void hfi1_pcie_flr(struct hfi1_devdata *dd) -{ - int i; - u16 status; - - /* no need to check for the capability - we know the device has it */ - - /* wait for Transaction Pending bit to clear, at most a few ms */ - for (i = 0; i < 4; i++) { - if (i) - msleep((1 << (i - 1)) * 100); - - pcie_capability_read_word(dd->pcidev, PCI_EXP_DEVSTA, &status); - if (!(status & PCI_EXP_DEVSTA_TRPND)) - goto clear; - } - - dd_dev_err(dd, "Transaction Pending bit is not clearing, proceeding with reset anyway\n"); - -clear: - pcie_capability_set_word(dd->pcidev, PCI_EXP_DEVCTL, - PCI_EXP_DEVCTL_BCR_FLR); - /* PCIe spec requires the function to be back within 100ms */ - msleep(100); -} - static void msix_setup(struct hfi1_devdata *dd, int pos, u32 *msixcnt, struct hfi1_msix_entry *hfi1_msix_entry) { -- cgit v1.2.3-59-g8ed1b From 63af8f7a8293b94787763c933abfd21ba852807f Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 14 Apr 2017 21:11:28 +0200 Subject: ixgbe: Use pcie_flr() instead of duplicating it Signed-off-by: Christoph Hellwig Signed-off-by: Bjorn Helgaas Acked-by: Jeff Kirsher --- drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 16 ++-------------- 1 file changed, 2 insertions(+), 14 deletions(-) diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c index a7a430a7be2c..543ddde5f8e2 100644 --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c @@ -7112,18 +7112,6 @@ static void ixgbe_watchdog_flush_tx(struct ixgbe_adapter *adapter) } #ifdef CONFIG_PCI_IOV -static inline void ixgbe_issue_vf_flr(struct ixgbe_adapter *adapter, - struct pci_dev *vfdev) -{ - if (!pci_wait_for_pending_transaction(vfdev)) - e_dev_warn("Issuing VFLR with pending transactions\n"); - - e_dev_err("Issuing VFLR for VF %s\n", pci_name(vfdev)); - pcie_capability_set_word(vfdev, PCI_EXP_DEVCTL, PCI_EXP_DEVCTL_BCR_FLR); - - msleep(100); -} - static void ixgbe_check_for_bad_vf(struct ixgbe_adapter *adapter) { struct ixgbe_hw *hw = &adapter->hw; @@ -7156,7 +7144,7 @@ static void ixgbe_check_for_bad_vf(struct ixgbe_adapter *adapter) pci_read_config_word(vfdev, PCI_STATUS, &status_reg); if (status_reg != IXGBE_FAILED_READ_CFG_WORD && status_reg & PCI_STATUS_REC_MASTER_ABORT) - ixgbe_issue_vf_flr(adapter, vfdev); + pcie_flr(vfdev); } } @@ -10244,7 +10232,7 @@ static pci_ers_result_t ixgbe_io_error_detected(struct pci_dev *pdev, * VFLR. Just clean up the AER in that case. */ if (vfdev) { - ixgbe_issue_vf_flr(adapter, vfdev); + pcie_flr(vfdev); /* Free device reference count */ pci_dev_put(vfdev); } -- cgit v1.2.3-59-g8ed1b From 13570533813cf111bb49954392cca0c2c409e775 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:04 +0530 Subject: Documentation: PCI: Add binding documentation for pci-test endpoint function Add binding documentation for pci-test endpoint function that helps in adding and configuring pci-test endpoint function. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- Documentation/PCI/00-INDEX | 2 ++ .../PCI/endpoint/function/binding/pci-test.txt | 17 +++++++++++++++++ 2 files changed, 19 insertions(+) create mode 100644 Documentation/PCI/endpoint/function/binding/pci-test.txt diff --git a/Documentation/PCI/00-INDEX b/Documentation/PCI/00-INDEX index ab35e4bbdb1c..2fc901a1c32e 100644 --- a/Documentation/PCI/00-INDEX +++ b/Documentation/PCI/00-INDEX @@ -18,3 +18,5 @@ endpoint/pci-endpoint-cfs.txt - guide to use configfs to configure the PCI endpoint function. endpoint/pci-test-function.txt - specification of *PCI test* function device. +endpoint/function/binding/ + - binding documentation for PCI endpoint function diff --git a/Documentation/PCI/endpoint/function/binding/pci-test.txt b/Documentation/PCI/endpoint/function/binding/pci-test.txt new file mode 100644 index 000000000000..3b68b955fb50 --- /dev/null +++ b/Documentation/PCI/endpoint/function/binding/pci-test.txt @@ -0,0 +1,17 @@ +PCI TEST ENDPOINT FUNCTION + +name: Should be "pci_epf_test" to bind to the pci_epf_test driver. + +Configurable Fields: +vendorid : should be 0x104c +deviceid : should be 0xb500 for DRA74x and 0xb501 for DRA72x +revid : don't care +progif_code : don't care +subclass_code : don't care +baseclass_code : should be 0xff +cache_line_size : don't care +subsys_vendor_id : don't care +subsys_id : don't care +interrupt_pin : Should be 1 - INTA, 2 - INTB, 3 - INTC, 4 -INTD +msi_interrupts : Should be 1 to 32 depending on the number of MSI interrupts + to test -- cgit v1.2.3-59-g8ed1b From f8aed6ec624fb436877a1a552393fd22510a5ff7 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:05 +0530 Subject: PCI: dwc: designware: Add EP mode support Add endpoint mode support to designware driver. This uses the EP Core layer introduced recently to add endpoint mode support. *Any* function driver can now use this designware device in order to achieve the EP functionality. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/Kconfig | 5 + drivers/pci/dwc/Makefile | 1 + drivers/pci/dwc/pcie-designware-ep.c | 342 +++++++++++++++++++++++++++++++++++ drivers/pci/dwc/pcie-designware.c | 125 +++++++++++++ drivers/pci/dwc/pcie-designware.h | 105 +++++++++++ 5 files changed, 578 insertions(+) create mode 100644 drivers/pci/dwc/pcie-designware-ep.c diff --git a/drivers/pci/dwc/Kconfig b/drivers/pci/dwc/Kconfig index d2d2ba5b8a68..d37ea72a846a 100644 --- a/drivers/pci/dwc/Kconfig +++ b/drivers/pci/dwc/Kconfig @@ -9,6 +9,11 @@ config PCIE_DW_HOST depends on PCI_MSI_IRQ_DOMAIN select PCIE_DW +config PCIE_DW_EP + bool + depends on PCI_ENDPOINT + select PCIE_DW + config PCI_DRA7XX bool "TI DRA7xx PCIe controller" depends on PCI diff --git a/drivers/pci/dwc/Makefile b/drivers/pci/dwc/Makefile index a2df13c28798..b38425d36200 100644 --- a/drivers/pci/dwc/Makefile +++ b/drivers/pci/dwc/Makefile @@ -1,5 +1,6 @@ obj-$(CONFIG_PCIE_DW) += pcie-designware.o obj-$(CONFIG_PCIE_DW_HOST) += pcie-designware-host.o +obj-$(CONFIG_PCIE_DW_EP) += pcie-designware-ep.o obj-$(CONFIG_PCIE_DW_PLAT) += pcie-designware-plat.o obj-$(CONFIG_PCI_DRA7XX) += pci-dra7xx.o obj-$(CONFIG_PCI_EXYNOS) += pci-exynos.o diff --git a/drivers/pci/dwc/pcie-designware-ep.c b/drivers/pci/dwc/pcie-designware-ep.c new file mode 100644 index 000000000000..398406393f37 --- /dev/null +++ b/drivers/pci/dwc/pcie-designware-ep.c @@ -0,0 +1,342 @@ +/** + * Synopsys Designware PCIe Endpoint controller driver + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include + +#include "pcie-designware.h" +#include +#include + +void dw_pcie_ep_linkup(struct dw_pcie_ep *ep) +{ + struct pci_epc *epc = ep->epc; + + pci_epc_linkup(epc); +} + +static void dw_pcie_ep_reset_bar(struct dw_pcie *pci, enum pci_barno bar) +{ + u32 reg; + + reg = PCI_BASE_ADDRESS_0 + (4 * bar); + dw_pcie_writel_dbi2(pci, reg, 0x0); + dw_pcie_writel_dbi(pci, reg, 0x0); +} + +static int dw_pcie_ep_write_header(struct pci_epc *epc, + struct pci_epf_header *hdr) +{ + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + dw_pcie_writew_dbi(pci, PCI_VENDOR_ID, hdr->vendorid); + dw_pcie_writew_dbi(pci, PCI_DEVICE_ID, hdr->deviceid); + dw_pcie_writeb_dbi(pci, PCI_REVISION_ID, hdr->revid); + dw_pcie_writeb_dbi(pci, PCI_CLASS_PROG, hdr->progif_code); + dw_pcie_writew_dbi(pci, PCI_CLASS_DEVICE, + hdr->subclass_code | hdr->baseclass_code << 8); + dw_pcie_writeb_dbi(pci, PCI_CACHE_LINE_SIZE, + hdr->cache_line_size); + dw_pcie_writew_dbi(pci, PCI_SUBSYSTEM_VENDOR_ID, + hdr->subsys_vendor_id); + dw_pcie_writew_dbi(pci, PCI_SUBSYSTEM_ID, hdr->subsys_id); + dw_pcie_writeb_dbi(pci, PCI_INTERRUPT_PIN, + hdr->interrupt_pin); + + return 0; +} + +static int dw_pcie_ep_inbound_atu(struct dw_pcie_ep *ep, enum pci_barno bar, + dma_addr_t cpu_addr, + enum dw_pcie_as_type as_type) +{ + int ret; + u32 free_win; + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + free_win = find_first_zero_bit(&ep->ib_window_map, + sizeof(ep->ib_window_map)); + if (free_win >= ep->num_ib_windows) { + dev_err(pci->dev, "no free inbound window\n"); + return -EINVAL; + } + + ret = dw_pcie_prog_inbound_atu(pci, free_win, bar, cpu_addr, + as_type); + if (ret < 0) { + dev_err(pci->dev, "Failed to program IB window\n"); + return ret; + } + + ep->bar_to_atu[bar] = free_win; + set_bit(free_win, &ep->ib_window_map); + + return 0; +} + +static int dw_pcie_ep_outbound_atu(struct dw_pcie_ep *ep, phys_addr_t phys_addr, + u64 pci_addr, size_t size) +{ + u32 free_win; + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + free_win = find_first_zero_bit(&ep->ob_window_map, + sizeof(ep->ob_window_map)); + if (free_win >= ep->num_ob_windows) { + dev_err(pci->dev, "no free outbound window\n"); + return -EINVAL; + } + + dw_pcie_prog_outbound_atu(pci, free_win, PCIE_ATU_TYPE_MEM, + phys_addr, pci_addr, size); + + set_bit(free_win, &ep->ob_window_map); + ep->outbound_addr[free_win] = phys_addr; + + return 0; +} + +static void dw_pcie_ep_clear_bar(struct pci_epc *epc, enum pci_barno bar) +{ + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + u32 atu_index = ep->bar_to_atu[bar]; + + dw_pcie_ep_reset_bar(pci, bar); + + dw_pcie_disable_atu(pci, atu_index, DW_PCIE_REGION_INBOUND); + clear_bit(atu_index, &ep->ib_window_map); +} + +static int dw_pcie_ep_set_bar(struct pci_epc *epc, enum pci_barno bar, + dma_addr_t bar_phys, size_t size, int flags) +{ + int ret; + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + enum dw_pcie_as_type as_type; + u32 reg = PCI_BASE_ADDRESS_0 + (4 * bar); + + if (!(flags & PCI_BASE_ADDRESS_SPACE)) + as_type = DW_PCIE_AS_MEM; + else + as_type = DW_PCIE_AS_IO; + + ret = dw_pcie_ep_inbound_atu(ep, bar, bar_phys, as_type); + if (ret) + return ret; + + dw_pcie_writel_dbi2(pci, reg, size - 1); + dw_pcie_writel_dbi(pci, reg, flags); + + return 0; +} + +static int dw_pcie_find_index(struct dw_pcie_ep *ep, phys_addr_t addr, + u32 *atu_index) +{ + u32 index; + + for (index = 0; index < ep->num_ob_windows; index++) { + if (ep->outbound_addr[index] != addr) + continue; + *atu_index = index; + return 0; + } + + return -EINVAL; +} + +static void dw_pcie_ep_unmap_addr(struct pci_epc *epc, phys_addr_t addr) +{ + int ret; + u32 atu_index; + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + ret = dw_pcie_find_index(ep, addr, &atu_index); + if (ret < 0) + return; + + dw_pcie_disable_atu(pci, atu_index, DW_PCIE_REGION_OUTBOUND); + clear_bit(atu_index, &ep->ob_window_map); +} + +static int dw_pcie_ep_map_addr(struct pci_epc *epc, phys_addr_t addr, + u64 pci_addr, size_t size) +{ + int ret; + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + ret = dw_pcie_ep_outbound_atu(ep, addr, pci_addr, size); + if (ret) { + dev_err(pci->dev, "failed to enable address\n"); + return ret; + } + + return 0; +} + +static int dw_pcie_ep_get_msi(struct pci_epc *epc) +{ + int val; + u32 lower_addr; + u32 upper_addr; + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + val = dw_pcie_readb_dbi(pci, MSI_MESSAGE_CONTROL); + val = (val & MSI_CAP_MME_MASK) >> MSI_CAP_MME_SHIFT; + + lower_addr = dw_pcie_readl_dbi(pci, MSI_MESSAGE_ADDR_L32); + upper_addr = dw_pcie_readl_dbi(pci, MSI_MESSAGE_ADDR_U32); + + if (!(lower_addr || upper_addr)) + return -EINVAL; + + return val; +} + +static int dw_pcie_ep_set_msi(struct pci_epc *epc, u8 encode_int) +{ + int val; + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + val = (encode_int << MSI_CAP_MMC_SHIFT); + dw_pcie_writew_dbi(pci, MSI_MESSAGE_CONTROL, val); + + return 0; +} + +static int dw_pcie_ep_raise_irq(struct pci_epc *epc, + enum pci_epc_irq_type type, u8 interrupt_num) +{ + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + + if (!ep->ops->raise_irq) + return -EINVAL; + + return ep->ops->raise_irq(ep, type, interrupt_num); +} + +static void dw_pcie_ep_stop(struct pci_epc *epc) +{ + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + if (!pci->ops->stop_link) + return; + + pci->ops->stop_link(pci); +} + +static int dw_pcie_ep_start(struct pci_epc *epc) +{ + struct dw_pcie_ep *ep = epc_get_drvdata(epc); + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + + if (!pci->ops->start_link) + return -EINVAL; + + return pci->ops->start_link(pci); +} + +static const struct pci_epc_ops epc_ops = { + .write_header = dw_pcie_ep_write_header, + .set_bar = dw_pcie_ep_set_bar, + .clear_bar = dw_pcie_ep_clear_bar, + .map_addr = dw_pcie_ep_map_addr, + .unmap_addr = dw_pcie_ep_unmap_addr, + .set_msi = dw_pcie_ep_set_msi, + .get_msi = dw_pcie_ep_get_msi, + .raise_irq = dw_pcie_ep_raise_irq, + .start = dw_pcie_ep_start, + .stop = dw_pcie_ep_stop, +}; + +void dw_pcie_ep_exit(struct dw_pcie_ep *ep) +{ + struct pci_epc *epc = ep->epc; + + pci_epc_mem_exit(epc); +} + +int dw_pcie_ep_init(struct dw_pcie_ep *ep) +{ + int ret; + void *addr; + enum pci_barno bar; + struct pci_epc *epc; + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + struct device *dev = pci->dev; + struct device_node *np = dev->of_node; + + if (!pci->dbi_base || !pci->dbi_base2) { + dev_err(dev, "dbi_base/deb_base2 is not populated\n"); + return -EINVAL; + } + + ret = of_property_read_u32(np, "num-ib-windows", &ep->num_ib_windows); + if (ret < 0) { + dev_err(dev, "unable to read *num-ib-windows* property\n"); + return ret; + } + + ret = of_property_read_u32(np, "num-ob-windows", &ep->num_ob_windows); + if (ret < 0) { + dev_err(dev, "unable to read *num-ob-windows* property\n"); + return ret; + } + + addr = devm_kzalloc(dev, sizeof(phys_addr_t) * ep->num_ob_windows, + GFP_KERNEL); + if (!addr) + return -ENOMEM; + ep->outbound_addr = addr; + + for (bar = BAR_0; bar <= BAR_5; bar++) + dw_pcie_ep_reset_bar(pci, bar); + + if (ep->ops->ep_init) + ep->ops->ep_init(ep); + + epc = devm_pci_epc_create(dev, &epc_ops); + if (IS_ERR(epc)) { + dev_err(dev, "failed to create epc device\n"); + return PTR_ERR(epc); + } + + ret = of_property_read_u8(np, "max-functions", &epc->max_functions); + if (ret < 0) + epc->max_functions = 1; + + ret = pci_epc_mem_init(epc, ep->phys_base, ep->addr_size); + if (ret < 0) { + dev_err(dev, "Failed to initialize address space\n"); + return ret; + } + + ep->epc = epc; + epc_set_drvdata(epc, ep); + dw_pcie_setup(pci); + + return 0; +} diff --git a/drivers/pci/dwc/pcie-designware.c b/drivers/pci/dwc/pcie-designware.c index 54de468a745e..0e03af279259 100644 --- a/drivers/pci/dwc/pcie-designware.c +++ b/drivers/pci/dwc/pcie-designware.c @@ -185,6 +185,131 @@ void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, int type, dev_err(pci->dev, "outbound iATU is not being enabled\n"); } +static u32 dw_pcie_readl_ib_unroll(struct dw_pcie *pci, u32 index, u32 reg) +{ + u32 offset = PCIE_GET_ATU_INB_UNR_REG_OFFSET(index); + + return dw_pcie_readl_dbi(pci, offset + reg); +} + +static void dw_pcie_writel_ib_unroll(struct dw_pcie *pci, u32 index, u32 reg, + u32 val) +{ + u32 offset = PCIE_GET_ATU_INB_UNR_REG_OFFSET(index); + + dw_pcie_writel_dbi(pci, offset + reg, val); +} + +int dw_pcie_prog_inbound_atu_unroll(struct dw_pcie *pci, int index, int bar, + u64 cpu_addr, enum dw_pcie_as_type as_type) +{ + int type; + u32 retries, val; + + dw_pcie_writel_ib_unroll(pci, index, PCIE_ATU_UNR_LOWER_TARGET, + lower_32_bits(cpu_addr)); + dw_pcie_writel_ib_unroll(pci, index, PCIE_ATU_UNR_UPPER_TARGET, + upper_32_bits(cpu_addr)); + + switch (as_type) { + case DW_PCIE_AS_MEM: + type = PCIE_ATU_TYPE_MEM; + break; + case DW_PCIE_AS_IO: + type = PCIE_ATU_TYPE_IO; + break; + default: + return -EINVAL; + } + + dw_pcie_writel_ib_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL1, type); + dw_pcie_writel_ib_unroll(pci, index, PCIE_ATU_UNR_REGION_CTRL2, + PCIE_ATU_ENABLE | + PCIE_ATU_BAR_MODE_ENABLE | (bar << 8)); + + /* + * Make sure ATU enable takes effect before any subsequent config + * and I/O accesses. + */ + for (retries = 0; retries < LINK_WAIT_MAX_IATU_RETRIES; retries++) { + val = dw_pcie_readl_ib_unroll(pci, index, + PCIE_ATU_UNR_REGION_CTRL2); + if (val & PCIE_ATU_ENABLE) + return 0; + + usleep_range(LINK_WAIT_IATU_MIN, LINK_WAIT_IATU_MAX); + } + dev_err(pci->dev, "inbound iATU is not being enabled\n"); + + return -EBUSY; +} + +int dw_pcie_prog_inbound_atu(struct dw_pcie *pci, int index, int bar, + u64 cpu_addr, enum dw_pcie_as_type as_type) +{ + int type; + u32 retries, val; + + if (pci->iatu_unroll_enabled) + return dw_pcie_prog_inbound_atu_unroll(pci, index, bar, + cpu_addr, as_type); + + dw_pcie_writel_dbi(pci, PCIE_ATU_VIEWPORT, PCIE_ATU_REGION_INBOUND | + index); + dw_pcie_writel_dbi(pci, PCIE_ATU_LOWER_TARGET, lower_32_bits(cpu_addr)); + dw_pcie_writel_dbi(pci, PCIE_ATU_UPPER_TARGET, upper_32_bits(cpu_addr)); + + switch (as_type) { + case DW_PCIE_AS_MEM: + type = PCIE_ATU_TYPE_MEM; + break; + case DW_PCIE_AS_IO: + type = PCIE_ATU_TYPE_IO; + break; + default: + return -EINVAL; + } + + dw_pcie_writel_dbi(pci, PCIE_ATU_CR1, type); + dw_pcie_writel_dbi(pci, PCIE_ATU_CR2, PCIE_ATU_ENABLE + | PCIE_ATU_BAR_MODE_ENABLE | (bar << 8)); + + /* + * Make sure ATU enable takes effect before any subsequent config + * and I/O accesses. + */ + for (retries = 0; retries < LINK_WAIT_MAX_IATU_RETRIES; retries++) { + val = dw_pcie_readl_dbi(pci, PCIE_ATU_CR2); + if (val & PCIE_ATU_ENABLE) + return 0; + + usleep_range(LINK_WAIT_IATU_MIN, LINK_WAIT_IATU_MAX); + } + dev_err(pci->dev, "inbound iATU is not being enabled\n"); + + return -EBUSY; +} + +void dw_pcie_disable_atu(struct dw_pcie *pci, int index, + enum dw_pcie_region_type type) +{ + int region; + + switch (type) { + case DW_PCIE_REGION_INBOUND: + region = PCIE_ATU_REGION_INBOUND; + break; + case DW_PCIE_REGION_OUTBOUND: + region = PCIE_ATU_REGION_OUTBOUND; + break; + default: + return; + } + + dw_pcie_writel_dbi(pci, PCIE_ATU_VIEWPORT, region | index); + dw_pcie_writel_dbi(pci, PCIE_ATU_CR2, ~PCIE_ATU_ENABLE); +} + int dw_pcie_wait_for_link(struct dw_pcie *pci) { int retries; diff --git a/drivers/pci/dwc/pcie-designware.h b/drivers/pci/dwc/pcie-designware.h index bfaf2b850a88..3cafba40abbc 100644 --- a/drivers/pci/dwc/pcie-designware.h +++ b/drivers/pci/dwc/pcie-designware.h @@ -18,6 +18,9 @@ #include #include +#include +#include + /* Parameters for the waiting for link up routine */ #define LINK_WAIT_MAX_RETRIES 10 #define LINK_WAIT_USLEEP_MIN 90000 @@ -89,6 +92,16 @@ #define PCIE_GET_ATU_OUTB_UNR_REG_OFFSET(region) \ ((0x3 << 20) | ((region) << 9)) +#define PCIE_GET_ATU_INB_UNR_REG_OFFSET(region) \ + ((0x3 << 20) | ((region) << 9) | (0x1 << 8)) + +#define MSI_MESSAGE_CONTROL 0x52 +#define MSI_CAP_MMC_SHIFT 1 +#define MSI_CAP_MME_SHIFT 4 +#define MSI_CAP_MME_MASK (7 << MSI_CAP_MME_SHIFT) +#define MSI_MESSAGE_ADDR_L32 0x54 +#define MSI_MESSAGE_ADDR_U32 0x58 + /* * Maximum number of MSI IRQs can be 256 per controller. But keep * it 32 as of now. Probably we will never need more than 32. If needed, @@ -99,6 +112,13 @@ struct pcie_port; struct dw_pcie; +struct dw_pcie_ep; + +enum dw_pcie_region_type { + DW_PCIE_REGION_UNKNOWN, + DW_PCIE_REGION_INBOUND, + DW_PCIE_REGION_OUTBOUND, +}; struct dw_pcie_host_ops { int (*rd_own_conf)(struct pcie_port *pp, int where, int size, u32 *val); @@ -142,6 +162,31 @@ struct pcie_port { DECLARE_BITMAP(msi_irq_in_use, MAX_MSI_IRQS); }; +enum dw_pcie_as_type { + DW_PCIE_AS_UNKNOWN, + DW_PCIE_AS_MEM, + DW_PCIE_AS_IO, +}; + +struct dw_pcie_ep_ops { + void (*ep_init)(struct dw_pcie_ep *ep); + int (*raise_irq)(struct dw_pcie_ep *ep, enum pci_epc_irq_type type, + u8 interrupt_num); +}; + +struct dw_pcie_ep { + struct pci_epc *epc; + struct dw_pcie_ep_ops *ops; + phys_addr_t phys_base; + size_t addr_size; + u8 bar_to_atu[6]; + phys_addr_t *outbound_addr; + unsigned long ib_window_map; + unsigned long ob_window_map; + u32 num_ib_windows; + u32 num_ob_windows; +}; + struct dw_pcie_ops { u64 (*cpu_addr_fixup)(u64 cpu_addr); u32 (*read_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, @@ -149,19 +194,26 @@ struct dw_pcie_ops { void (*write_dbi)(struct dw_pcie *pcie, void __iomem *base, u32 reg, size_t size, u32 val); int (*link_up)(struct dw_pcie *pcie); + int (*start_link)(struct dw_pcie *pcie); + void (*stop_link)(struct dw_pcie *pcie); }; struct dw_pcie { struct device *dev; void __iomem *dbi_base; + void __iomem *dbi_base2; u32 num_viewport; u8 iatu_unroll_enabled; struct pcie_port pp; + struct dw_pcie_ep ep; const struct dw_pcie_ops *ops; }; #define to_dw_pcie_from_pp(port) container_of((port), struct dw_pcie, pp) +#define to_dw_pcie_from_ep(endpoint) \ + container_of((endpoint), struct dw_pcie, ep) + int dw_pcie_read(void __iomem *addr, int size, u32 *val); int dw_pcie_write(void __iomem *addr, int size, u32 val); @@ -174,6 +226,10 @@ int dw_pcie_wait_for_link(struct dw_pcie *pci); void dw_pcie_prog_outbound_atu(struct dw_pcie *pci, int index, int type, u64 cpu_addr, u64 pci_addr, u32 size); +int dw_pcie_prog_inbound_atu(struct dw_pcie *pci, int index, int bar, + u64 cpu_addr, enum dw_pcie_as_type as_type); +void dw_pcie_disable_atu(struct dw_pcie *pci, int index, + enum dw_pcie_region_type type); void dw_pcie_setup(struct dw_pcie *pci); static inline void dw_pcie_writel_dbi(struct dw_pcie *pci, u32 reg, u32 val) @@ -186,6 +242,36 @@ static inline u32 dw_pcie_readl_dbi(struct dw_pcie *pci, u32 reg) return __dw_pcie_read_dbi(pci, pci->dbi_base, reg, 0x4); } +static inline void dw_pcie_writew_dbi(struct dw_pcie *pci, u32 reg, u16 val) +{ + __dw_pcie_write_dbi(pci, pci->dbi_base, reg, 0x2, val); +} + +static inline u16 dw_pcie_readw_dbi(struct dw_pcie *pci, u32 reg) +{ + return __dw_pcie_read_dbi(pci, pci->dbi_base, reg, 0x2); +} + +static inline void dw_pcie_writeb_dbi(struct dw_pcie *pci, u32 reg, u8 val) +{ + __dw_pcie_write_dbi(pci, pci->dbi_base, reg, 0x1, val); +} + +static inline u8 dw_pcie_readb_dbi(struct dw_pcie *pci, u32 reg) +{ + return __dw_pcie_read_dbi(pci, pci->dbi_base, reg, 0x1); +} + +static inline void dw_pcie_writel_dbi2(struct dw_pcie *pci, u32 reg, u32 val) +{ + __dw_pcie_write_dbi(pci, pci->dbi_base2, reg, 0x4, val); +} + +static inline u32 dw_pcie_readl_dbi2(struct dw_pcie *pci, u32 reg) +{ + return __dw_pcie_read_dbi(pci, pci->dbi_base2, reg, 0x4); +} + #ifdef CONFIG_PCIE_DW_HOST irqreturn_t dw_handle_msi_irq(struct pcie_port *pp); void dw_pcie_msi_init(struct pcie_port *pp); @@ -210,4 +296,23 @@ static inline int dw_pcie_host_init(struct pcie_port *pp) return 0; } #endif + +#ifdef CONFIG_PCIE_DW_EP +void dw_pcie_ep_linkup(struct dw_pcie_ep *ep); +int dw_pcie_ep_init(struct dw_pcie_ep *ep); +void dw_pcie_ep_exit(struct dw_pcie_ep *ep); +#else +static inline void dw_pcie_ep_linkup(struct dw_pcie_ep *ep) +{ +} + +static inline int dw_pcie_ep_init(struct dw_pcie_ep *ep) +{ + return 0; +} + +static inline void dw_pcie_ep_exit(struct dw_pcie_ep *ep) +{ +} +#endif #endif /* _PCIE_DESIGNWARE_H */ -- cgit v1.2.3-59-g8ed1b From b12befecd7de905a05053eb6a266e3b97528c6d4 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:06 +0530 Subject: dt-bindings: PCI: Add DT bindings for PCI designware EP mode Add device tree binding documentation for PCI designware EP mode. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas Acked-by: Rob Herring --- .../devicetree/bindings/pci/designware-pcie.txt | 26 +++++++++++++++------- 1 file changed, 18 insertions(+), 8 deletions(-) diff --git a/Documentation/devicetree/bindings/pci/designware-pcie.txt b/Documentation/devicetree/bindings/pci/designware-pcie.txt index 1392c705ceca..b2480dd38c11 100644 --- a/Documentation/devicetree/bindings/pci/designware-pcie.txt +++ b/Documentation/devicetree/bindings/pci/designware-pcie.txt @@ -6,30 +6,40 @@ Required properties: - reg-names: Must be "config" for the PCIe configuration space. (The old way of getting the configuration address space from "ranges" is deprecated and should be avoided.) +- num-lanes: number of lanes to use +RC mode: - #address-cells: set to <3> - #size-cells: set to <2> - device_type: set to "pci" - ranges: ranges for the PCI memory and I/O regions - #interrupt-cells: set to <1> -- interrupt-map-mask and interrupt-map: standard PCI properties - to define the mapping of the PCIe interface to interrupt +- interrupt-map-mask and interrupt-map: standard PCI + properties to define the mapping of the PCIe interface to interrupt numbers. -- num-lanes: number of lanes to use +EP mode: +- num-ib-windows: number of inbound address translation + windows +- num-ob-windows: number of outbound address translation + windows Optional properties: -- num-viewport: number of view ports configured in hardware. If a platform - does not specify it, the driver assumes 2. - num-lanes: number of lanes to use (this property should be specified unless the link is brought already up in BIOS) - reset-gpio: gpio pin number of power good signal -- bus-range: PCI bus numbers covered (it is recommended for new devicetrees to - specify this property, to keep backwards compatibility a range of 0x00-0xff - is assumed if not present) - clocks: Must contain an entry for each entry in clock-names. See ../clocks/clock-bindings.txt for details. - clock-names: Must include the following entries: - "pcie" - "pcie_bus" +RC mode: +- num-viewport: number of view ports configured in + hardware. If a platform does not specify it, the driver assumes 2. +- bus-range: PCI bus numbers covered (it is recommended + for new devicetrees to specify this property, to keep backwards + compatibility a range of 0x00-0xff is assumed if not present) +EP mode: +- max-functions: maximum number of functions that can be + configured Example configuration: -- cgit v1.2.3-59-g8ed1b From 5ffd90a035a273068a8bbef74a01ae923b80110b Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:07 +0530 Subject: PCI: dwc: dra7xx: Facilitate wrapper and MSI interrupts to be enabled independently No functional change. Split dra7xx_pcie_enable_interrupts() into dra7xx_pcie_enable_wrapper_interrupts() and dra7xx_pcie_enable_msi_interrupts() so that wrapper interrupts and MSI interrupts can be enabled independently. This is in preparation for adding EP mode support to dra7xx driver since EP mode doesn't have to enable msi_interrupts. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pci-dra7xx.c | 24 ++++++++++++++++++------ 1 file changed, 18 insertions(+), 6 deletions(-) diff --git a/drivers/pci/dwc/pci-dra7xx.c b/drivers/pci/dwc/pci-dra7xx.c index 7c9ed6a6675c..d78974d20360 100644 --- a/drivers/pci/dwc/pci-dra7xx.c +++ b/drivers/pci/dwc/pci-dra7xx.c @@ -140,18 +140,30 @@ static int dra7xx_pcie_establish_link(struct dra7xx_pcie *dra7xx) return dw_pcie_wait_for_link(pci); } -static void dra7xx_pcie_enable_interrupts(struct dra7xx_pcie *dra7xx) +static void dra7xx_pcie_enable_msi_interrupts(struct dra7xx_pcie *dra7xx) { - dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQSTATUS_MAIN, - ~INTERRUPTS); - dra7xx_pcie_writel(dra7xx, - PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MAIN, INTERRUPTS); dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQSTATUS_MSI, ~LEG_EP_INTERRUPTS & ~MSI); - dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MSI, + + dra7xx_pcie_writel(dra7xx, + PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MSI, MSI | LEG_EP_INTERRUPTS); } +static void dra7xx_pcie_enable_wrapper_interrupts(struct dra7xx_pcie *dra7xx) +{ + dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQSTATUS_MAIN, + ~INTERRUPTS); + dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MAIN, + INTERRUPTS); +} + +static void dra7xx_pcie_enable_interrupts(struct dra7xx_pcie *dra7xx) +{ + dra7xx_pcie_enable_wrapper_interrupts(dra7xx); + dra7xx_pcie_enable_msi_interrupts(dra7xx); +} + static void dra7xx_pcie_host_init(struct pcie_port *pp) { struct dw_pcie *pci = to_dw_pcie_from_pp(pp); -- cgit v1.2.3-59-g8ed1b From 608793e27b3313b2385af557c95d8e4207126670 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:08 +0530 Subject: PCI: dwc: dra7xx: Add EP mode support The PCIe controller integrated in dra7xx SoCs is capable of operating in endpoint mode. Add endpoint mode support to dra7xx driver. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/Kconfig | 31 +++++- drivers/pci/dwc/Makefile | 4 +- drivers/pci/dwc/pci-dra7xx.c | 197 +++++++++++++++++++++++++++++++++++--- drivers/pci/dwc/pcie-designware.h | 7 ++ 4 files changed, 221 insertions(+), 18 deletions(-) diff --git a/drivers/pci/dwc/Kconfig b/drivers/pci/dwc/Kconfig index d37ea72a846a..b7e15526d676 100644 --- a/drivers/pci/dwc/Kconfig +++ b/drivers/pci/dwc/Kconfig @@ -16,14 +16,37 @@ config PCIE_DW_EP config PCI_DRA7XX bool "TI DRA7xx PCIe controller" - depends on PCI + depends on (PCI && PCI_MSI_IRQ_DOMAIN) || PCI_ENDPOINT depends on OF && HAS_IOMEM && TI_PIPE3 + help + Enables support for the PCIe controller in the DRA7xx SoC. There + are two instances of PCIe controller in DRA7xx. This controller can + work either as EP or RC. In order to enable host-specific features + PCI_DRA7XX_HOST must be selected and in order to enable device- + specific features PCI_DRA7XX_EP must be selected. This uses + the Designware core. + +if PCI_DRA7XX + +config PCI_DRA7XX_HOST + bool "PCI DRA7xx Host Mode" + depends on PCI depends on PCI_MSI_IRQ_DOMAIN select PCIE_DW_HOST + default y help - Enables support for the PCIe controller in the DRA7xx SoC. There - are two instances of PCIe controller in DRA7xx. This controller can - act both as EP and RC. This reuses the Designware core. + Enables support for the PCIe controller in the DRA7xx SoC to work in + host mode. + +config PCI_DRA7XX_EP + bool "PCI DRA7xx Endpoint Mode" + depends on PCI_ENDPOINT + select PCIE_DW_EP + help + Enables support for the PCIe controller in the DRA7xx SoC to work in + endpoint mode. + +endif config PCIE_DW_PLAT bool "Platform bus based DesignWare PCIe Controller" diff --git a/drivers/pci/dwc/Makefile b/drivers/pci/dwc/Makefile index b38425d36200..f31a8596442a 100644 --- a/drivers/pci/dwc/Makefile +++ b/drivers/pci/dwc/Makefile @@ -2,7 +2,9 @@ obj-$(CONFIG_PCIE_DW) += pcie-designware.o obj-$(CONFIG_PCIE_DW_HOST) += pcie-designware-host.o obj-$(CONFIG_PCIE_DW_EP) += pcie-designware-ep.o obj-$(CONFIG_PCIE_DW_PLAT) += pcie-designware-plat.o -obj-$(CONFIG_PCI_DRA7XX) += pci-dra7xx.o +ifneq ($(filter y,$(CONFIG_PCI_DRA7XX_HOST) $(CONFIG_PCI_DRA7XX_EP)),) + obj-$(CONFIG_PCI_DRA7XX) += pci-dra7xx.o +endif obj-$(CONFIG_PCI_EXYNOS) += pci-exynos.o obj-$(CONFIG_PCI_IMX6) += pci-imx6.o obj-$(CONFIG_PCIE_SPEAR13XX) += pcie-spear13xx.o diff --git a/drivers/pci/dwc/pci-dra7xx.c b/drivers/pci/dwc/pci-dra7xx.c index d78974d20360..35c18534469c 100644 --- a/drivers/pci/dwc/pci-dra7xx.c +++ b/drivers/pci/dwc/pci-dra7xx.c @@ -10,12 +10,14 @@ * published by the Free Software Foundation. */ +#include #include #include #include #include #include #include +#include #include #include #include @@ -57,6 +59,11 @@ #define MSI BIT(4) #define LEG_EP_INTERRUPTS (INTA | INTB | INTC | INTD) +#define PCIECTRL_TI_CONF_DEVICE_TYPE 0x0100 +#define DEVICE_TYPE_EP 0x0 +#define DEVICE_TYPE_LEG_EP 0x1 +#define DEVICE_TYPE_RC 0x4 + #define PCIECTRL_DRA7XX_CONF_DEVICE_CMD 0x0104 #define LTSSM_EN 0x1 @@ -66,6 +73,13 @@ #define EXP_CAP_ID_OFFSET 0x70 +#define PCIECTRL_TI_CONF_INTX_ASSERT 0x0124 +#define PCIECTRL_TI_CONF_INTX_DEASSERT 0x0128 + +#define PCIECTRL_TI_CONF_MSI_XMT 0x012c +#define MSI_REQ_GRANT BIT(0) +#define MSI_VECTOR_SHIFT 7 + struct dra7xx_pcie { struct dw_pcie *pci; void __iomem *base; /* DT ti_conf */ @@ -73,6 +87,11 @@ struct dra7xx_pcie { struct phy **phy; int link_gen; struct irq_domain *irq_domain; + enum dw_pcie_device_mode mode; +}; + +struct dra7xx_pcie_of_data { + enum dw_pcie_device_mode mode; }; #define to_dra7xx_pcie(x) dev_get_drvdata((x)->dev) @@ -101,9 +120,19 @@ static int dra7xx_pcie_link_up(struct dw_pcie *pci) return !!(reg & LINK_UP); } -static int dra7xx_pcie_establish_link(struct dra7xx_pcie *dra7xx) +static void dra7xx_pcie_stop_link(struct dw_pcie *pci) { - struct dw_pcie *pci = dra7xx->pci; + struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); + u32 reg; + + reg = dra7xx_pcie_readl(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD); + reg &= ~LTSSM_EN; + dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD, reg); +} + +static int dra7xx_pcie_establish_link(struct dw_pcie *pci) +{ + struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); struct device *dev = pci->dev; u32 reg; u32 exp_cap_off = EXP_CAP_ID_OFFSET; @@ -137,7 +166,7 @@ static int dra7xx_pcie_establish_link(struct dra7xx_pcie *dra7xx) reg |= LTSSM_EN; dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD, reg); - return dw_pcie_wait_for_link(pci); + return 0; } static void dra7xx_pcie_enable_msi_interrupts(struct dra7xx_pcie *dra7xx) @@ -171,7 +200,8 @@ static void dra7xx_pcie_host_init(struct pcie_port *pp) dw_pcie_setup_rc(pp); - dra7xx_pcie_establish_link(dra7xx); + dra7xx_pcie_establish_link(pci); + dw_pcie_wait_for_link(pci); dw_pcie_msi_init(pp); dra7xx_pcie_enable_interrupts(dra7xx); } @@ -249,6 +279,7 @@ static irqreturn_t dra7xx_pcie_irq_handler(int irq, void *arg) struct dra7xx_pcie *dra7xx = arg; struct dw_pcie *pci = dra7xx->pci; struct device *dev = pci->dev; + struct dw_pcie_ep *ep = &pci->ep; u32 reg; reg = dra7xx_pcie_readl(dra7xx, PCIECTRL_DRA7XX_CONF_IRQSTATUS_MAIN); @@ -285,8 +316,11 @@ static irqreturn_t dra7xx_pcie_irq_handler(int irq, void *arg) if (reg & LINK_REQ_RST) dev_dbg(dev, "Link Request Reset\n"); - if (reg & LINK_UP_EVT) + if (reg & LINK_UP_EVT) { + if (dra7xx->mode == DW_PCIE_EP_TYPE) + dw_pcie_ep_linkup(ep); dev_dbg(dev, "Link-up state change\n"); + } if (reg & CFG_BME_EVT) dev_dbg(dev, "CFG 'Bus Master Enable' change\n"); @@ -299,6 +333,94 @@ static irqreturn_t dra7xx_pcie_irq_handler(int irq, void *arg) return IRQ_HANDLED; } +static void dra7xx_pcie_ep_init(struct dw_pcie_ep *ep) +{ + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); + + dra7xx_pcie_enable_wrapper_interrupts(dra7xx); +} + +static void dra7xx_pcie_raise_legacy_irq(struct dra7xx_pcie *dra7xx) +{ + dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_INTX_ASSERT, 0x1); + mdelay(1); + dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_INTX_DEASSERT, 0x1); +} + +static void dra7xx_pcie_raise_msi_irq(struct dra7xx_pcie *dra7xx, + u8 interrupt_num) +{ + u32 reg; + + reg = (interrupt_num - 1) << MSI_VECTOR_SHIFT; + reg |= MSI_REQ_GRANT; + dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_MSI_XMT, reg); +} + +static int dra7xx_pcie_raise_irq(struct dw_pcie_ep *ep, + enum pci_epc_irq_type type, u8 interrupt_num) +{ + struct dw_pcie *pci = to_dw_pcie_from_ep(ep); + struct dra7xx_pcie *dra7xx = to_dra7xx_pcie(pci); + + switch (type) { + case PCI_EPC_IRQ_LEGACY: + dra7xx_pcie_raise_legacy_irq(dra7xx); + break; + case PCI_EPC_IRQ_MSI: + dra7xx_pcie_raise_msi_irq(dra7xx, interrupt_num); + break; + default: + dev_err(pci->dev, "UNKNOWN IRQ type\n"); + } + + return 0; +} + +static struct dw_pcie_ep_ops pcie_ep_ops = { + .ep_init = dra7xx_pcie_ep_init, + .raise_irq = dra7xx_pcie_raise_irq, +}; + +static int __init dra7xx_add_pcie_ep(struct dra7xx_pcie *dra7xx, + struct platform_device *pdev) +{ + int ret; + struct dw_pcie_ep *ep; + struct resource *res; + struct device *dev = &pdev->dev; + struct dw_pcie *pci = dra7xx->pci; + + ep = &pci->ep; + ep->ops = &pcie_ep_ops; + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "ep_dbics"); + pci->dbi_base = devm_ioremap(dev, res->start, resource_size(res)); + if (!pci->dbi_base) + return -ENOMEM; + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "ep_dbics2"); + pci->dbi_base2 = devm_ioremap(dev, res->start, resource_size(res)); + if (!pci->dbi_base2) + return -ENOMEM; + + res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "addr_space"); + if (!res) + return -EINVAL; + + ep->phys_base = res->start; + ep->addr_size = resource_size(res); + + ret = dw_pcie_ep_init(ep); + if (ret) { + dev_err(dev, "failed to initialize endpoint\n"); + return ret; + } + + return 0; +} + static int __init dra7xx_add_pcie_port(struct dra7xx_pcie *dra7xx, struct platform_device *pdev) { @@ -342,6 +464,8 @@ static int __init dra7xx_add_pcie_port(struct dra7xx_pcie *dra7xx, static const struct dw_pcie_ops dw_pcie_ops = { .cpu_addr_fixup = dra7xx_pcie_cpu_addr_fixup, + .start_link = dra7xx_pcie_establish_link, + .stop_link = dra7xx_pcie_stop_link, .link_up = dra7xx_pcie_link_up, }; @@ -384,6 +508,26 @@ err_phy: return ret; } +static const struct dra7xx_pcie_of_data dra7xx_pcie_rc_of_data = { + .mode = DW_PCIE_RC_TYPE, +}; + +static const struct dra7xx_pcie_of_data dra7xx_pcie_ep_of_data = { + .mode = DW_PCIE_EP_TYPE, +}; + +static const struct of_device_id of_dra7xx_pcie_match[] = { + { + .compatible = "ti,dra7-pcie", + .data = &dra7xx_pcie_rc_of_data, + }, + { + .compatible = "ti,dra7-pcie-ep", + .data = &dra7xx_pcie_ep_of_data, + }, + {}, +}; + static int __init dra7xx_pcie_probe(struct platform_device *pdev) { u32 reg; @@ -401,6 +545,16 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) struct device_node *np = dev->of_node; char name[10]; struct gpio_desc *reset; + const struct of_device_id *match; + const struct dra7xx_pcie_of_data *data; + enum dw_pcie_device_mode mode; + + match = of_match_device(of_match_ptr(of_dra7xx_pcie_match), dev); + if (!match) + return -EINVAL; + + data = (struct dra7xx_pcie_of_data *)match->data; + mode = (enum dw_pcie_device_mode)data->mode; dra7xx = devm_kzalloc(dev, sizeof(*dra7xx), GFP_KERNEL); if (!dra7xx) @@ -479,9 +633,25 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2) dra7xx->link_gen = 2; - ret = dra7xx_add_pcie_port(dra7xx, pdev); - if (ret < 0) - goto err_gpio; + switch (mode) { + case DW_PCIE_RC_TYPE: + dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, + DEVICE_TYPE_RC); + ret = dra7xx_add_pcie_port(dra7xx, pdev); + if (ret < 0) + goto err_gpio; + break; + case DW_PCIE_EP_TYPE: + dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, + DEVICE_TYPE_EP); + ret = dra7xx_add_pcie_ep(dra7xx, pdev); + if (ret < 0) + goto err_gpio; + break; + default: + dev_err(dev, "INVALID device type %d\n", mode); + } + dra7xx->mode = mode; ret = devm_request_irq(dev, irq, dra7xx_pcie_irq_handler, IRQF_SHARED, "dra7xx-pcie-main", dra7xx); @@ -509,6 +679,9 @@ static int dra7xx_pcie_suspend(struct device *dev) struct dw_pcie *pci = dra7xx->pci; u32 val; + if (dra7xx->mode != DW_PCIE_RC_TYPE) + return 0; + /* clear MSE */ val = dw_pcie_readl_dbi(pci, PCI_COMMAND); val &= ~PCI_COMMAND_MEMORY; @@ -523,6 +696,9 @@ static int dra7xx_pcie_resume(struct device *dev) struct dw_pcie *pci = dra7xx->pci; u32 val; + if (dra7xx->mode != DW_PCIE_RC_TYPE) + return 0; + /* set MSE */ val = dw_pcie_readl_dbi(pci, PCI_COMMAND); val |= PCI_COMMAND_MEMORY; @@ -561,11 +737,6 @@ static const struct dev_pm_ops dra7xx_pcie_pm_ops = { dra7xx_pcie_resume_noirq) }; -static const struct of_device_id of_dra7xx_pcie_match[] = { - { .compatible = "ti,dra7-pcie", }, - {}, -}; - static struct platform_driver dra7xx_pcie_driver = { .driver = { .name = "dra7-pcie", diff --git a/drivers/pci/dwc/pcie-designware.h b/drivers/pci/dwc/pcie-designware.h index 3cafba40abbc..c6a840575796 100644 --- a/drivers/pci/dwc/pcie-designware.h +++ b/drivers/pci/dwc/pcie-designware.h @@ -120,6 +120,13 @@ enum dw_pcie_region_type { DW_PCIE_REGION_OUTBOUND, }; +enum dw_pcie_device_mode { + DW_PCIE_UNKNOWN_TYPE, + DW_PCIE_EP_TYPE, + DW_PCIE_LEG_EP_TYPE, + DW_PCIE_RC_TYPE, +}; + struct dw_pcie_host_ops { int (*rd_own_conf)(struct pcie_port *pp, int where, int size, u32 *val); int (*wr_own_conf)(struct pcie_port *pp, int where, int size, u32 val); -- cgit v1.2.3-59-g8ed1b From 40cc72e2b51f61956f965510ceba35bddb0c373f Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:09 +0530 Subject: dt-bindings: PCI: dra7xx: Add DT bindings for PCI dra7xx EP mode Add device tree binding documentation for PCI dra7xx EP mode. Acked-by: Rob Herring Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- Documentation/devicetree/bindings/pci/ti-pci.txt | 37 +++++++++++++++++++----- 1 file changed, 30 insertions(+), 7 deletions(-) diff --git a/Documentation/devicetree/bindings/pci/ti-pci.txt b/Documentation/devicetree/bindings/pci/ti-pci.txt index 60e25161f351..60c3cccefabc 100644 --- a/Documentation/devicetree/bindings/pci/ti-pci.txt +++ b/Documentation/devicetree/bindings/pci/ti-pci.txt @@ -1,17 +1,22 @@ TI PCI Controllers PCIe Designware Controller - - compatible: Should be "ti,dra7-pcie"" - - reg : Two register ranges as listed in the reg-names property - - reg-names : The first entry must be "ti-conf" for the TI specific registers - The second entry must be "rc-dbics" for the designware pcie - registers - The third entry must be "config" for the PCIe configuration space + - compatible: Should be "ti,dra7-pcie" for RC + Should be "ti,dra7-pcie-ep" for EP - phys : list of PHY specifiers (used by generic PHY framework) - phy-names : must be "pcie-phy0", "pcie-phy1", "pcie-phyN".. based on the number of PHYs as specified in *phys* property. - ti,hwmods : Name of the hwmod associated to the pcie, "pcie", where is the instance number of the pcie from the HW spec. + - num-lanes as specified in ../designware-pcie.txt + +HOST MODE +========= + - reg : Two register ranges as listed in the reg-names property + - reg-names : The first entry must be "ti-conf" for the TI specific registers + The second entry must be "rc-dbics" for the DesignWare PCIe + registers + The third entry must be "config" for the PCIe configuration space - interrupts : Two interrupt entries must be specified. The first one is for main interrupt line and the second for MSI interrupt line. - #address-cells, @@ -19,13 +24,31 @@ PCIe Designware Controller #interrupt-cells, device_type, ranges, - num-lanes, interrupt-map-mask, interrupt-map : as specified in ../designware-pcie.txt +DEVICE MODE +=========== + - reg : Four register ranges as listed in the reg-names property + - reg-names : "ti-conf" for the TI specific registers + "ep_dbics" for the standard configuration registers as + they are locally accessed within the DIF CS space + "ep_dbics2" for the standard configuration registers as + they are locally accessed within the DIF CS2 space + "addr_space" used to map remote RC address space + - interrupts : one interrupt entries must be specified for main interrupt. + - num-ib-windows : number of inbound address translation windows + - num-ob-windows : number of outbound address translation windows + Optional Property: - gpios : Should be added if a gpio line is required to drive PERST# line +NOTE: Two DT nodes may be added for each PCI controller; one for host +mode and another for device mode. So in order for PCI to +work in host mode, EP mode DT node should be disabled and in order to PCI to +work in EP mode, host mode DT node should be disabled. Host mode and EP +mode are mutually exclusive. + Example: axi { compatible = "simple-bus"; -- cgit v1.2.3-59-g8ed1b From f7a2757f6cd0aa4e1075188a3a0081a583fcee24 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:11 +0530 Subject: PCI: dwc: dra7xx: Workaround for errata id i870 According to errata i870, access to the PCIe slave port that are not 32-bit aligned will result in incorrect mapping to TLP Address and Byte enable fields. Accessing non 32-bit aligned data causes incorrect data in the target buffer if memcpy is used. Implement the workaround for this errata here. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pci-dra7xx.c | 49 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 49 insertions(+) diff --git a/drivers/pci/dwc/pci-dra7xx.c b/drivers/pci/dwc/pci-dra7xx.c index 35c18534469c..8decf46cf525 100644 --- a/drivers/pci/dwc/pci-dra7xx.c +++ b/drivers/pci/dwc/pci-dra7xx.c @@ -26,6 +26,8 @@ #include #include #include +#include +#include #include "pcie-designware.h" @@ -528,6 +530,48 @@ static const struct of_device_id of_dra7xx_pcie_match[] = { {}, }; +/* + * dra7xx_pcie_ep_unaligned_memaccess: workaround for AM572x/AM571x Errata i870 + * @dra7xx: the dra7xx device where the workaround should be applied + * + * Access to the PCIe slave port that are not 32-bit aligned will result + * in incorrect mapping to TLP Address and Byte enable fields. Therefore, + * byte and half-word accesses are not possible to byte offset 0x1, 0x2, or + * 0x3. + * + * To avoid this issue set PCIE_SS1_AXI2OCP_LEGACY_MODE_ENABLE to 1. + */ +static int dra7xx_pcie_ep_unaligned_memaccess(struct device *dev) +{ + int ret; + struct device_node *np = dev->of_node; + struct of_phandle_args args; + struct regmap *regmap; + + regmap = syscon_regmap_lookup_by_phandle(np, + "ti,syscon-unaligned-access"); + if (IS_ERR(regmap)) { + dev_dbg(dev, "can't get ti,syscon-unaligned-access\n"); + return -EINVAL; + } + + ret = of_parse_phandle_with_fixed_args(np, "ti,syscon-unaligned-access", + 2, 0, &args); + if (ret) { + dev_err(dev, "failed to parse ti,syscon-unaligned-access\n"); + return ret; + } + + ret = regmap_update_bits(regmap, args.args[0], args.args[1], + args.args[1]); + if (ret) + dev_err(dev, "failed to enable unaligned access\n"); + + of_node_put(args.np); + + return ret; +} + static int __init dra7xx_pcie_probe(struct platform_device *pdev) { u32 reg; @@ -644,6 +688,11 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev) case DW_PCIE_EP_TYPE: dra7xx_pcie_writel(dra7xx, PCIECTRL_TI_CONF_DEVICE_TYPE, DEVICE_TYPE_EP); + + ret = dra7xx_pcie_ep_unaligned_memaccess(dev); + if (ret) + goto err_gpio; + ret = dra7xx_add_pcie_ep(dra7xx, pdev); if (ret < 0) goto err_gpio; -- cgit v1.2.3-59-g8ed1b From 66abca1d9db24e9e0b0d031d9b246e440e26fdae Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:12 +0530 Subject: dt-bindings: PCI: dra7xx: Add DT bindings to enable unaligned access Update device tree binding documentation of TI's dra7xx PCI controller to include property for enabling unaligned mem access. Signed-off-by: Kishon Vijay Abraham I Acked-by: Rob Herring Signed-off-by: Bjorn Helgaas --- Documentation/devicetree/bindings/pci/ti-pci.txt | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/Documentation/devicetree/bindings/pci/ti-pci.txt b/Documentation/devicetree/bindings/pci/ti-pci.txt index 60c3cccefabc..6a07c96227e0 100644 --- a/Documentation/devicetree/bindings/pci/ti-pci.txt +++ b/Documentation/devicetree/bindings/pci/ti-pci.txt @@ -39,6 +39,11 @@ DEVICE MODE - interrupts : one interrupt entries must be specified for main interrupt. - num-ib-windows : number of inbound address translation windows - num-ob-windows : number of outbound address translation windows + - ti,syscon-unaligned-access: phandle to the syscon DT node. The 1st argument + should contain the register offset within syscon + and the 2nd argument should contain the bit field + for setting the bit to enable unaligned + access. Optional Property: - gpios : Should be added if a gpio line is required to drive PERST# line -- cgit v1.2.3-59-g8ed1b From 984c307878f8924d743c419c79fdebbc19f1285e Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:13 +0530 Subject: PCI: Add device IDs for DRA74x and DRA72x Add device IDs for DRA74x and DRA72x devices. These devices have configurable PCI endpoint. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- include/linux/pci_ids.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index a4f77feecbb0..5f6b71d15393 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -862,6 +862,8 @@ #define PCI_DEVICE_ID_TI_X620 0xac8d #define PCI_DEVICE_ID_TI_X420 0xac8e #define PCI_DEVICE_ID_TI_XX20_FM 0xac8f +#define PCI_DEVICE_ID_TI_DRA74x 0xb500 +#define PCI_DEVICE_ID_TI_DRA72x 0xb501 #define PCI_VENDOR_ID_SONY 0x104d -- cgit v1.2.3-59-g8ed1b From 2c156ac71c6b2518f3d589190f2a8872a8764faf Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:14 +0530 Subject: misc: Add host side PCI driver for PCI test function device Add PCI endpoint test driver that can verify base address register, legacy interrupt/MSI interrupt and read/write/copy buffers between host and device. The corresponding pci-epf-test function driver should be used on the EP side. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- drivers/misc/Kconfig | 7 + drivers/misc/Makefile | 1 + drivers/misc/pci_endpoint_test.c | 534 +++++++++++++++++++++++++++++++++++++++ include/uapi/linux/Kbuild | 1 + include/uapi/linux/pcitest.h | 19 ++ 5 files changed, 562 insertions(+) create mode 100644 drivers/misc/pci_endpoint_test.c create mode 100644 include/uapi/linux/pcitest.h diff --git a/drivers/misc/Kconfig b/drivers/misc/Kconfig index c290990d73ed..527b115c4e23 100644 --- a/drivers/misc/Kconfig +++ b/drivers/misc/Kconfig @@ -771,6 +771,13 @@ config PANEL_BOOT_MESSAGE endif # PANEL +config PCI_ENDPOINT_TEST + depends on PCI + tristate "PCI Endpoint Test driver" + ---help--- + Enable this configuration option to enable the host side test driver + for PCI Endpoint. + source "drivers/misc/c2port/Kconfig" source "drivers/misc/eeprom/Kconfig" source "drivers/misc/cb710/Kconfig" diff --git a/drivers/misc/Makefile b/drivers/misc/Makefile index 7a3ea89339b4..6e139cd70421 100644 --- a/drivers/misc/Makefile +++ b/drivers/misc/Makefile @@ -54,6 +54,7 @@ obj-$(CONFIG_ECHO) += echo/ obj-$(CONFIG_VEXPRESS_SYSCFG) += vexpress-syscfg.o obj-$(CONFIG_CXL_BASE) += cxl/ obj-$(CONFIG_PANEL) += panel.o +obj-$(CONFIG_PCI_ENDPOINT_TEST) += pci_endpoint_test.o lkdtm-$(CONFIG_LKDTM) += lkdtm_core.o lkdtm-$(CONFIG_LKDTM) += lkdtm_bugs.o diff --git a/drivers/misc/pci_endpoint_test.c b/drivers/misc/pci_endpoint_test.c new file mode 100644 index 000000000000..09c10f426b64 --- /dev/null +++ b/drivers/misc/pci_endpoint_test.c @@ -0,0 +1,534 @@ +/** + * Host side test driver to test endpoint functionality + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include + +#define DRV_MODULE_NAME "pci-endpoint-test" + +#define PCI_ENDPOINT_TEST_MAGIC 0x0 + +#define PCI_ENDPOINT_TEST_COMMAND 0x4 +#define COMMAND_RAISE_LEGACY_IRQ BIT(0) +#define COMMAND_RAISE_MSI_IRQ BIT(1) +#define MSI_NUMBER_SHIFT 2 +/* 6 bits for MSI number */ +#define COMMAND_READ BIT(8) +#define COMMAND_WRITE BIT(9) +#define COMMAND_COPY BIT(10) + +#define PCI_ENDPOINT_TEST_STATUS 0x8 +#define STATUS_READ_SUCCESS BIT(0) +#define STATUS_READ_FAIL BIT(1) +#define STATUS_WRITE_SUCCESS BIT(2) +#define STATUS_WRITE_FAIL BIT(3) +#define STATUS_COPY_SUCCESS BIT(4) +#define STATUS_COPY_FAIL BIT(5) +#define STATUS_IRQ_RAISED BIT(6) +#define STATUS_SRC_ADDR_INVALID BIT(7) +#define STATUS_DST_ADDR_INVALID BIT(8) + +#define PCI_ENDPOINT_TEST_LOWER_SRC_ADDR 0xc +#define PCI_ENDPOINT_TEST_UPPER_SRC_ADDR 0x10 + +#define PCI_ENDPOINT_TEST_LOWER_DST_ADDR 0x14 +#define PCI_ENDPOINT_TEST_UPPER_DST_ADDR 0x18 + +#define PCI_ENDPOINT_TEST_SIZE 0x1c +#define PCI_ENDPOINT_TEST_CHECKSUM 0x20 + +static DEFINE_IDA(pci_endpoint_test_ida); + +#define to_endpoint_test(priv) container_of((priv), struct pci_endpoint_test, \ + miscdev) +enum pci_barno { + BAR_0, + BAR_1, + BAR_2, + BAR_3, + BAR_4, + BAR_5, +}; + +struct pci_endpoint_test { + struct pci_dev *pdev; + void __iomem *base; + void __iomem *bar[6]; + struct completion irq_raised; + int last_irq; + /* mutex to protect the ioctls */ + struct mutex mutex; + struct miscdevice miscdev; +}; + +static int bar_size[] = { 4, 512, 1024, 16384, 131072, 1048576 }; + +static inline u32 pci_endpoint_test_readl(struct pci_endpoint_test *test, + u32 offset) +{ + return readl(test->base + offset); +} + +static inline void pci_endpoint_test_writel(struct pci_endpoint_test *test, + u32 offset, u32 value) +{ + writel(value, test->base + offset); +} + +static inline u32 pci_endpoint_test_bar_readl(struct pci_endpoint_test *test, + int bar, int offset) +{ + return readl(test->bar[bar] + offset); +} + +static inline void pci_endpoint_test_bar_writel(struct pci_endpoint_test *test, + int bar, u32 offset, u32 value) +{ + writel(value, test->bar[bar] + offset); +} + +static irqreturn_t pci_endpoint_test_irqhandler(int irq, void *dev_id) +{ + struct pci_endpoint_test *test = dev_id; + u32 reg; + + reg = pci_endpoint_test_readl(test, PCI_ENDPOINT_TEST_STATUS); + if (reg & STATUS_IRQ_RAISED) { + test->last_irq = irq; + complete(&test->irq_raised); + reg &= ~STATUS_IRQ_RAISED; + } + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_STATUS, + reg); + + return IRQ_HANDLED; +} + +static bool pci_endpoint_test_bar(struct pci_endpoint_test *test, + enum pci_barno barno) +{ + int j; + u32 val; + int size; + + if (!test->bar[barno]) + return false; + + size = bar_size[barno]; + + for (j = 0; j < size; j += 4) + pci_endpoint_test_bar_writel(test, barno, j, 0xA0A0A0A0); + + for (j = 0; j < size; j += 4) { + val = pci_endpoint_test_bar_readl(test, barno, j); + if (val != 0xA0A0A0A0) + return false; + } + + return true; +} + +static bool pci_endpoint_test_legacy_irq(struct pci_endpoint_test *test) +{ + u32 val; + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_COMMAND, + COMMAND_RAISE_LEGACY_IRQ); + val = wait_for_completion_timeout(&test->irq_raised, + msecs_to_jiffies(1000)); + if (!val) + return false; + + return true; +} + +static bool pci_endpoint_test_msi_irq(struct pci_endpoint_test *test, + u8 msi_num) +{ + u32 val; + struct pci_dev *pdev = test->pdev; + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_COMMAND, + msi_num << MSI_NUMBER_SHIFT | + COMMAND_RAISE_MSI_IRQ); + val = wait_for_completion_timeout(&test->irq_raised, + msecs_to_jiffies(1000)); + if (!val) + return false; + + if (test->last_irq - pdev->irq == msi_num - 1) + return true; + + return false; +} + +static bool pci_endpoint_test_copy(struct pci_endpoint_test *test, size_t size) +{ + bool ret = false; + void *src_addr; + void *dst_addr; + dma_addr_t src_phys_addr; + dma_addr_t dst_phys_addr; + struct pci_dev *pdev = test->pdev; + struct device *dev = &pdev->dev; + u32 src_crc32; + u32 dst_crc32; + + src_addr = dma_alloc_coherent(dev, size, &src_phys_addr, GFP_KERNEL); + if (!src_addr) { + dev_err(dev, "failed to allocate source buffer\n"); + ret = false; + goto err; + } + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_LOWER_SRC_ADDR, + lower_32_bits(src_phys_addr)); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_UPPER_SRC_ADDR, + upper_32_bits(src_phys_addr)); + + get_random_bytes(src_addr, size); + src_crc32 = crc32_le(~0, src_addr, size); + + dst_addr = dma_alloc_coherent(dev, size, &dst_phys_addr, GFP_KERNEL); + if (!dst_addr) { + dev_err(dev, "failed to allocate destination address\n"); + ret = false; + goto err_src_addr; + } + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_LOWER_DST_ADDR, + lower_32_bits(dst_phys_addr)); + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_UPPER_DST_ADDR, + upper_32_bits(dst_phys_addr)); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_SIZE, + size); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_COMMAND, + 1 << MSI_NUMBER_SHIFT | COMMAND_COPY); + + wait_for_completion(&test->irq_raised); + + dst_crc32 = crc32_le(~0, dst_addr, size); + if (dst_crc32 == src_crc32) + ret = true; + + dma_free_coherent(dev, size, dst_addr, dst_phys_addr); + +err_src_addr: + dma_free_coherent(dev, size, src_addr, src_phys_addr); + +err: + return ret; +} + +static bool pci_endpoint_test_write(struct pci_endpoint_test *test, size_t size) +{ + bool ret = false; + u32 reg; + void *addr; + dma_addr_t phys_addr; + struct pci_dev *pdev = test->pdev; + struct device *dev = &pdev->dev; + u32 crc32; + + addr = dma_alloc_coherent(dev, size, &phys_addr, GFP_KERNEL); + if (!addr) { + dev_err(dev, "failed to allocate address\n"); + ret = false; + goto err; + } + + get_random_bytes(addr, size); + + crc32 = crc32_le(~0, addr, size); + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_CHECKSUM, + crc32); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_LOWER_SRC_ADDR, + lower_32_bits(phys_addr)); + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_UPPER_SRC_ADDR, + upper_32_bits(phys_addr)); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_SIZE, size); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_COMMAND, + 1 << MSI_NUMBER_SHIFT | COMMAND_READ); + + wait_for_completion(&test->irq_raised); + + reg = pci_endpoint_test_readl(test, PCI_ENDPOINT_TEST_STATUS); + if (reg & STATUS_READ_SUCCESS) + ret = true; + + dma_free_coherent(dev, size, addr, phys_addr); + +err: + return ret; +} + +static bool pci_endpoint_test_read(struct pci_endpoint_test *test, size_t size) +{ + bool ret = false; + void *addr; + dma_addr_t phys_addr; + struct pci_dev *pdev = test->pdev; + struct device *dev = &pdev->dev; + u32 crc32; + + addr = dma_alloc_coherent(dev, size, &phys_addr, GFP_KERNEL); + if (!addr) { + dev_err(dev, "failed to allocate destination address\n"); + ret = false; + goto err; + } + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_LOWER_DST_ADDR, + lower_32_bits(phys_addr)); + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_UPPER_DST_ADDR, + upper_32_bits(phys_addr)); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_SIZE, size); + + pci_endpoint_test_writel(test, PCI_ENDPOINT_TEST_COMMAND, + 1 << MSI_NUMBER_SHIFT | COMMAND_WRITE); + + wait_for_completion(&test->irq_raised); + + crc32 = crc32_le(~0, addr, size); + if (crc32 == pci_endpoint_test_readl(test, PCI_ENDPOINT_TEST_CHECKSUM)) + ret = true; + + dma_free_coherent(dev, size, addr, phys_addr); +err: + return ret; +} + +static long pci_endpoint_test_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + int ret = -EINVAL; + enum pci_barno bar; + struct pci_endpoint_test *test = to_endpoint_test(file->private_data); + + mutex_lock(&test->mutex); + switch (cmd) { + case PCITEST_BAR: + bar = arg; + if (bar < 0 || bar > 5) + goto ret; + ret = pci_endpoint_test_bar(test, bar); + break; + case PCITEST_LEGACY_IRQ: + ret = pci_endpoint_test_legacy_irq(test); + break; + case PCITEST_MSI: + ret = pci_endpoint_test_msi_irq(test, arg); + break; + case PCITEST_WRITE: + ret = pci_endpoint_test_write(test, arg); + break; + case PCITEST_READ: + ret = pci_endpoint_test_read(test, arg); + break; + case PCITEST_COPY: + ret = pci_endpoint_test_copy(test, arg); + break; + } + +ret: + mutex_unlock(&test->mutex); + return ret; +} + +static const struct file_operations pci_endpoint_test_fops = { + .owner = THIS_MODULE, + .unlocked_ioctl = pci_endpoint_test_ioctl, +}; + +static int pci_endpoint_test_probe(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + int i; + int err; + int irq; + int id; + char name[20]; + enum pci_barno bar; + void __iomem *base; + struct device *dev = &pdev->dev; + struct pci_endpoint_test *test; + struct miscdevice *misc_device; + + if (pci_is_bridge(pdev)) + return -ENODEV; + + test = devm_kzalloc(dev, sizeof(*test), GFP_KERNEL); + if (!test) + return -ENOMEM; + + test->pdev = pdev; + init_completion(&test->irq_raised); + mutex_init(&test->mutex); + + err = pci_enable_device(pdev); + if (err) { + dev_err(dev, "Cannot enable PCI device\n"); + return err; + } + + err = pci_request_regions(pdev, DRV_MODULE_NAME); + if (err) { + dev_err(dev, "Cannot obtain PCI resources\n"); + goto err_disable_pdev; + } + + pci_set_master(pdev); + + irq = pci_alloc_irq_vectors(pdev, 1, 32, PCI_IRQ_MSI); + if (irq < 0) + dev_err(dev, "failed to get MSI interrupts\n"); + + err = devm_request_irq(dev, pdev->irq, pci_endpoint_test_irqhandler, + IRQF_SHARED, DRV_MODULE_NAME, test); + if (err) { + dev_err(dev, "failed to request IRQ %d\n", pdev->irq); + goto err_disable_msi; + } + + for (i = 1; i < irq; i++) { + err = devm_request_irq(dev, pdev->irq + i, + pci_endpoint_test_irqhandler, + IRQF_SHARED, DRV_MODULE_NAME, test); + if (err) + dev_err(dev, "failed to request IRQ %d for MSI %d\n", + pdev->irq + i, i + 1); + } + + for (bar = BAR_0; bar <= BAR_5; bar++) { + base = pci_ioremap_bar(pdev, bar); + if (!base) { + dev_err(dev, "failed to read BAR%d\n", bar); + WARN_ON(bar == BAR_0); + } + test->bar[bar] = base; + } + + test->base = test->bar[0]; + if (!test->base) { + dev_err(dev, "Cannot perform PCI test without BAR0\n"); + goto err_iounmap; + } + + pci_set_drvdata(pdev, test); + + id = ida_simple_get(&pci_endpoint_test_ida, 0, 0, GFP_KERNEL); + if (id < 0) { + dev_err(dev, "unable to get id\n"); + goto err_iounmap; + } + + snprintf(name, sizeof(name), DRV_MODULE_NAME ".%d", id); + misc_device = &test->miscdev; + misc_device->minor = MISC_DYNAMIC_MINOR; + misc_device->name = name; + misc_device->fops = &pci_endpoint_test_fops, + + err = misc_register(misc_device); + if (err) { + dev_err(dev, "failed to register device\n"); + goto err_ida_remove; + } + + return 0; + +err_ida_remove: + ida_simple_remove(&pci_endpoint_test_ida, id); + +err_iounmap: + for (bar = BAR_0; bar <= BAR_5; bar++) { + if (test->bar[bar]) + pci_iounmap(pdev, test->bar[bar]); + } + +err_disable_msi: + pci_disable_msi(pdev); + pci_release_regions(pdev); + +err_disable_pdev: + pci_disable_device(pdev); + + return err; +} + +static void pci_endpoint_test_remove(struct pci_dev *pdev) +{ + int id; + enum pci_barno bar; + struct pci_endpoint_test *test = pci_get_drvdata(pdev); + struct miscdevice *misc_device = &test->miscdev; + + if (sscanf(misc_device->name, DRV_MODULE_NAME ".%d", &id) != 1) + return; + + misc_deregister(&test->miscdev); + ida_simple_remove(&pci_endpoint_test_ida, id); + for (bar = BAR_0; bar <= BAR_5; bar++) { + if (test->bar[bar]) + pci_iounmap(pdev, test->bar[bar]); + } + pci_disable_msi(pdev); + pci_release_regions(pdev); + pci_disable_device(pdev); +} + +static const struct pci_device_id pci_endpoint_test_tbl[] = { + { PCI_DEVICE(PCI_VENDOR_ID_TI, PCI_DEVICE_ID_TI_DRA74x) }, + { PCI_DEVICE(PCI_VENDOR_ID_TI, PCI_DEVICE_ID_TI_DRA72x) }, + { } +}; +MODULE_DEVICE_TABLE(pci, pci_endpoint_test_tbl); + +static struct pci_driver pci_endpoint_test_driver = { + .name = DRV_MODULE_NAME, + .id_table = pci_endpoint_test_tbl, + .probe = pci_endpoint_test_probe, + .remove = pci_endpoint_test_remove, +}; +module_pci_driver(pci_endpoint_test_driver); + +MODULE_DESCRIPTION("PCI ENDPOINT TEST HOST DRIVER"); +MODULE_AUTHOR("Kishon Vijay Abraham I "); +MODULE_LICENSE("GPL v2"); diff --git a/include/uapi/linux/Kbuild b/include/uapi/linux/Kbuild index dd9820b1c779..baee6db08287 100644 --- a/include/uapi/linux/Kbuild +++ b/include/uapi/linux/Kbuild @@ -333,6 +333,7 @@ header-y += parport.h header-y += patchkey.h header-y += pci.h header-y += pci_regs.h +header-y += pcitest.h header-y += perf_event.h header-y += personality.h header-y += pfkeyv2.h diff --git a/include/uapi/linux/pcitest.h b/include/uapi/linux/pcitest.h new file mode 100644 index 000000000000..a6aa10c45ad1 --- /dev/null +++ b/include/uapi/linux/pcitest.h @@ -0,0 +1,19 @@ +/** + * pcitest.h - PCI test uapi defines + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + */ + +#ifndef __UAPI_LINUX_PCITEST_H +#define __UAPI_LINUX_PCITEST_H + +#define PCITEST_BAR _IO('P', 0x1) +#define PCITEST_LEGACY_IRQ _IO('P', 0x2) +#define PCITEST_MSI _IOW('P', 0x3, int) +#define PCITEST_WRITE _IOW('P', 0x4, unsigned long) +#define PCITEST_READ _IOW('P', 0x5, unsigned long) +#define PCITEST_COPY _IOW('P', 0x6, unsigned long) + +#endif /* __UAPI_LINUX_PCITEST_H */ -- cgit v1.2.3-59-g8ed1b From 0cfb512e4acc7049d36478221f55bf6053bd81a6 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:15 +0530 Subject: Documentation: misc-devices: Add Documentation for pci-endpoint-test driver Add Documentation for pci-endpoint-test driver. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- Documentation/misc-devices/pci-endpoint-test.txt | 35 ++++++++++++++++++++++++ 1 file changed, 35 insertions(+) create mode 100644 Documentation/misc-devices/pci-endpoint-test.txt diff --git a/Documentation/misc-devices/pci-endpoint-test.txt b/Documentation/misc-devices/pci-endpoint-test.txt new file mode 100644 index 000000000000..4ebc3594b32c --- /dev/null +++ b/Documentation/misc-devices/pci-endpoint-test.txt @@ -0,0 +1,35 @@ +Driver for PCI Endpoint Test Function + +This driver should be used as a host side driver if the root complex is +connected to a configurable PCI endpoint running *pci_epf_test* function +driver configured according to [1]. + +The "pci_endpoint_test" driver can be used to perform the following tests. + +The PCI driver for the test device performs the following tests + *) verifying addresses programmed in BAR + *) raise legacy IRQ + *) raise MSI IRQ + *) read data + *) write data + *) copy data + +This misc driver creates /dev/pci-endpoint-test. for every +*pci_epf_test* function connected to the root complex and "ioctls" +should be used to perform the above tests. + +ioctl +----- + PCITEST_BAR: Tests the BAR. The number of the BAR to be tested + should be passed as argument. + PCITEST_LEGACY_IRQ: Tests legacy IRQ + PCITEST_MSI: Tests message signalled interrupts. The MSI number + to be tested should be passed as argument. + PCITEST_WRITE: Perform write tests. The size of the buffer should be passed + as argument. + PCITEST_READ: Perform read tests. The size of the buffer should be passed + as argument. + PCITEST_COPY: Perform read tests. The size of the buffer should be passed + as argument. + +[1] -> Documentation/PCI/endpoint/function/binding/pci-test.txt -- cgit v1.2.3-59-g8ed1b From 3f2ed813483452e3f8de56ec40c97c432c16dbf1 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:16 +0530 Subject: tools: PCI: Add a userspace tool to test PCI endpoint Add a userspace tool to invoke the ioctls exposed by the PCI endpoint test driver to perform various PCI tests. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- tools/pci/pcitest.c | 186 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 186 insertions(+) create mode 100644 tools/pci/pcitest.c diff --git a/tools/pci/pcitest.c b/tools/pci/pcitest.c new file mode 100644 index 000000000000..ad54a58d7dda --- /dev/null +++ b/tools/pci/pcitest.c @@ -0,0 +1,186 @@ +/** + * Userspace PCI Endpoint Test Module + * + * Copyright (C) 2017 Texas Instruments + * Author: Kishon Vijay Abraham I + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 of + * the License as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see . + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#define BILLION 1E9 + +static char *result[] = { "NOT OKAY", "OKAY" }; + +struct pci_test { + char *device; + char barnum; + bool legacyirq; + unsigned int msinum; + bool read; + bool write; + bool copy; + unsigned long size; +}; + +static int run_test(struct pci_test *test) +{ + long ret; + int fd; + struct timespec start, end; + double time; + + fd = open(test->device, O_RDWR); + if (fd < 0) { + perror("can't open PCI Endpoint Test device"); + return fd; + } + + if (test->barnum >= 0 && test->barnum <= 5) { + ret = ioctl(fd, PCITEST_BAR, test->barnum); + fprintf(stdout, "BAR%d:\t\t", test->barnum); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + if (test->legacyirq) { + ret = ioctl(fd, PCITEST_LEGACY_IRQ, 0); + fprintf(stdout, "LEGACY IRQ:\t"); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + if (test->msinum > 0 && test->msinum <= 32) { + ret = ioctl(fd, PCITEST_MSI, test->msinum); + fprintf(stdout, "MSI%d:\t\t", test->msinum); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + if (test->write) { + ret = ioctl(fd, PCITEST_WRITE, test->size); + fprintf(stdout, "WRITE (%7ld bytes):\t\t", test->size); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + if (test->read) { + ret = ioctl(fd, PCITEST_READ, test->size); + fprintf(stdout, "READ (%7ld bytes):\t\t", test->size); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + if (test->copy) { + ret = ioctl(fd, PCITEST_COPY, test->size); + fprintf(stdout, "COPY (%7ld bytes):\t\t", test->size); + if (ret < 0) + fprintf(stdout, "TEST FAILED\n"); + else + fprintf(stdout, "%s\n", result[ret]); + } + + fflush(stdout); +} + +int main(int argc, char **argv) +{ + int c; + struct pci_test *test; + + test = calloc(1, sizeof(*test)); + if (!test) { + perror("Fail to allocate memory for pci_test\n"); + return -ENOMEM; + } + + /* since '0' is a valid BAR number, initialize it to -1 */ + test->barnum = -1; + + /* set default size as 100KB */ + test->size = 0x19000; + + /* set default endpoint device */ + test->device = "/dev/pci-endpoint-test.0"; + + while ((c = getopt(argc, argv, "D:b:m:lrwcs:")) != EOF) + switch (c) { + case 'D': + test->device = optarg; + continue; + case 'b': + test->barnum = atoi(optarg); + if (test->barnum < 0 || test->barnum > 5) + goto usage; + continue; + case 'l': + test->legacyirq = true; + continue; + case 'm': + test->msinum = atoi(optarg); + if (test->msinum < 1 || test->msinum > 32) + goto usage; + continue; + case 'r': + test->read = true; + continue; + case 'w': + test->write = true; + continue; + case 'c': + test->copy = true; + continue; + case 's': + test->size = strtoul(optarg, NULL, 0); + continue; + case '?': + case 'h': + default: +usage: + fprintf(stderr, + "usage: %s [options]\n" + "Options:\n" + "\t-D PCI endpoint test device {default: /dev/pci-endpoint-test.0}\n" + "\t-b BAR test (bar number between 0..5)\n" + "\t-m MSI test (msi number between 1..32)\n" + "\t-r Read buffer test\n" + "\t-w Write buffer test\n" + "\t-c Copy buffer test\n" + "\t-s Size of buffer {default: 100KB}\n", + argv[0]); + return -EINVAL; + } + + run_test(test); + return 0; +} -- cgit v1.2.3-59-g8ed1b From 62cb0eff82d8c64b79d3cf46a995306551beeaf3 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:17 +0530 Subject: tools: PCI: Add sample test script to invoke pcitest Add a simple test script that invokes the pcitest userspace tool to perform all the PCI endpoint tests (BAR tests, interrupt tests, read tests, write tests and copy tests). Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- tools/pci/pcitest.sh | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) create mode 100644 tools/pci/pcitest.sh diff --git a/tools/pci/pcitest.sh b/tools/pci/pcitest.sh new file mode 100644 index 000000000000..5442bbea4c22 --- /dev/null +++ b/tools/pci/pcitest.sh @@ -0,0 +1,56 @@ +#!/bin/sh + +echo "BAR tests" +echo + +bar=0 + +while [ $bar -lt 6 ] +do + pcitest -b $bar + bar=`expr $bar + 1` +done +echo + +echo "Interrupt tests" +echo + +pcitest -l +msi=1 + +while [ $msi -lt 33 ] +do + pcitest -m $msi + msi=`expr $msi + 1` +done +echo + +echo "Read Tests" +echo + +pcitest -r -s 1 +pcitest -r -s 1024 +pcitest -r -s 1025 +pcitest -r -s 1024000 +pcitest -r -s 1024001 +echo + +echo "Write Tests" +echo + +pcitest -w -s 1 +pcitest -w -s 1024 +pcitest -w -s 1025 +pcitest -w -s 1024000 +pcitest -w -s 1024001 +echo + +echo "Copy Tests" +echo + +pcitest -c -s 1 +pcitest -c -s 1024 +pcitest -c -s 1025 +pcitest -c -s 1024000 +pcitest -c -s 1024001 +echo -- cgit v1.2.3-59-g8ed1b From 16263d9e1dedb30ddca51d684f306cdc4294e705 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:18 +0530 Subject: Documentation: PCI: Add userguide for PCI endpoint test function Add documentation to help users use pci-epf-test function driver and pci_endpoint_test host driver for testing PCI. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- Documentation/PCI/00-INDEX | 2 + Documentation/PCI/endpoint/pci-test-howto.txt | 179 ++++++++++++++++++++++++++ 2 files changed, 181 insertions(+) create mode 100644 Documentation/PCI/endpoint/pci-test-howto.txt diff --git a/Documentation/PCI/00-INDEX b/Documentation/PCI/00-INDEX index 2fc901a1c32e..00c9a90b6f38 100644 --- a/Documentation/PCI/00-INDEX +++ b/Documentation/PCI/00-INDEX @@ -18,5 +18,7 @@ endpoint/pci-endpoint-cfs.txt - guide to use configfs to configure the PCI endpoint function. endpoint/pci-test-function.txt - specification of *PCI test* function device. +endpoint/pci-test-howto.txt + - userguide for PCI endpoint test function. endpoint/function/binding/ - binding documentation for PCI endpoint function diff --git a/Documentation/PCI/endpoint/pci-test-howto.txt b/Documentation/PCI/endpoint/pci-test-howto.txt new file mode 100644 index 000000000000..75f48c3bb191 --- /dev/null +++ b/Documentation/PCI/endpoint/pci-test-howto.txt @@ -0,0 +1,179 @@ + PCI TEST USERGUIDE + Kishon Vijay Abraham I + +This document is a guide to help users use pci-epf-test function driver +and pci_endpoint_test host driver for testing PCI. The list of steps to +be followed in the host side and EP side is given below. + +1. Endpoint Device + +1.1 Endpoint Controller Devices + +To find the list of endpoint controller devices in the system: + + # ls /sys/class/pci_epc/ + 51000000.pcie_ep + +If PCI_ENDPOINT_CONFIGFS is enabled + # ls /sys/kernel/config/pci_ep/controllers + 51000000.pcie_ep + +1.2 Endpoint Function Drivers + +To find the list of endpoint function drivers in the system: + + # ls /sys/bus/pci-epf/drivers + pci_epf_test + +If PCI_ENDPOINT_CONFIGFS is enabled + # ls /sys/kernel/config/pci_ep/functions + pci_epf_test + +1.3 Creating pci-epf-test Device + +PCI endpoint function device can be created using the configfs. To create +pci-epf-test device, the following commands can be used + + # mount -t configfs none /sys/kernel/config + # cd /sys/kernel/config/pci_ep/ + # mkdir functions/pci_epf_test/func1 + +The "mkdir func1" above creates the pci-epf-test function device that will +be probed by pci_epf_test driver. + +The PCI endpoint framework populates the directory with the following +configurable fields. + + # ls functions/pci_epf_test/func1 + baseclass_code interrupt_pin revid subsys_vendor_id + cache_line_size msi_interrupts subclass_code vendorid + deviceid progif_code subsys_id + +The PCI endpoint function driver populates these entries with default values +when the device is bound to the driver. The pci-epf-test driver populates +vendorid with 0xffff and interrupt_pin with 0x0001 + + # cat functions/pci_epf_test/func1/vendorid + 0xffff + # cat functions/pci_epf_test/func1/interrupt_pin + 0x0001 + +1.4 Configuring pci-epf-test Device + +The user can configure the pci-epf-test device using configfs entry. In order +to change the vendorid and the number of MSI interrupts used by the function +device, the following commands can be used. + + # echo 0x104c > functions/pci_epf_test/func1/vendorid + # echo 0xb500 > functions/pci_epf_test/func1/deviceid + # echo 16 > functions/pci_epf_test/func1/msi_interrupts + +1.5 Binding pci-epf-test Device to EP Controller + +In order for the endpoint function device to be useful, it has to be bound to +a PCI endpoint controller driver. Use the configfs to bind the function +device to one of the controller driver present in the system. + + # ln -s functions/pci_epf_test/func1 controllers/51000000.pcie_ep/ + +Once the above step is completed, the PCI endpoint is ready to establish a link +with the host. + +1.6 Start the Link + +In order for the endpoint device to establish a link with the host, the _start_ +field should be populated with '1'. + + # echo 1 > controllers/51000000.pcie_ep/start + +2. RootComplex Device + +2.1 lspci Output + +Note that the devices listed here correspond to the value populated in 1.4 above + + 00:00.0 PCI bridge: Texas Instruments Device 8888 (rev 01) + 01:00.0 Unassigned class [ff00]: Texas Instruments Device b500 + +2.2 Using Endpoint Test function Device + +pcitest.sh added in tools/pci/ can be used to run all the default PCI endpoint +tests. Before pcitest.sh can be used pcitest.c should be compiled using the +following commands. + + cd + make headers_install ARCH=arm + arm-linux-gnueabihf-gcc -Iusr/include tools/pci/pcitest.c -o pcitest + cp pcitest /usr/sbin/ + cp tools/pci/pcitest.sh + +2.2.1 pcitest.sh Output + # ./pcitest.sh + BAR tests + + BAR0: OKAY + BAR1: OKAY + BAR2: OKAY + BAR3: OKAY + BAR4: NOT OKAY + BAR5: NOT OKAY + + Interrupt tests + + LEGACY IRQ: NOT OKAY + MSI1: OKAY + MSI2: OKAY + MSI3: OKAY + MSI4: OKAY + MSI5: OKAY + MSI6: OKAY + MSI7: OKAY + MSI8: OKAY + MSI9: OKAY + MSI10: OKAY + MSI11: OKAY + MSI12: OKAY + MSI13: OKAY + MSI14: OKAY + MSI15: OKAY + MSI16: OKAY + MSI17: NOT OKAY + MSI18: NOT OKAY + MSI19: NOT OKAY + MSI20: NOT OKAY + MSI21: NOT OKAY + MSI22: NOT OKAY + MSI23: NOT OKAY + MSI24: NOT OKAY + MSI25: NOT OKAY + MSI26: NOT OKAY + MSI27: NOT OKAY + MSI28: NOT OKAY + MSI29: NOT OKAY + MSI30: NOT OKAY + MSI31: NOT OKAY + MSI32: NOT OKAY + + Read Tests + + READ ( 1 bytes): OKAY + READ ( 1024 bytes): OKAY + READ ( 1025 bytes): OKAY + READ (1024000 bytes): OKAY + READ (1024001 bytes): OKAY + + Write Tests + + WRITE ( 1 bytes): OKAY + WRITE ( 1024 bytes): OKAY + WRITE ( 1025 bytes): OKAY + WRITE (1024000 bytes): OKAY + WRITE (1024001 bytes): OKAY + + Copy Tests + + COPY ( 1 bytes): OKAY + COPY ( 1024 bytes): OKAY + COPY ( 1025 bytes): OKAY + COPY (1024000 bytes): OKAY + COPY (1024001 bytes): OKAY -- cgit v1.2.3-59-g8ed1b From f60b15b831b01973c71b54e82a258376f7593427 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:19 +0530 Subject: MAINTAINERS: Add PCI Endpoint maintainer Add maintainer for the newly introduced PCI Endpoint framework. Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- MAINTAINERS | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index c265a5fe4848..15ed84389092 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -9581,6 +9581,15 @@ F: include/linux/pci* F: arch/x86/pci/ F: arch/x86/kernel/quirks.c +PCI ENDPOINT SUBSYSTEM +M: Kishon Vijay Abraham I +L: linux-pci@vger.kernel.org +T: git git://git.kernel.org/pub/scm/linux/kernel/git/kishon/pci-endpoint.git +S: Supported +F: drivers/pci/endpoint/ +F: drivers/misc/pci_endpoint_test.c +F: tools/pci/ + PCI DRIVER FOR ALTERA PCIE IP M: Ley Foon Tan L: rfi@lists.rocketboards.org (moderated for non-subscribers) -- cgit v1.2.3-59-g8ed1b From 2c949ce38f4e81d7487f165fa3b8f77d74a2a6c4 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Mon, 27 Mar 2017 15:15:20 +0530 Subject: ARM: DRA7: clockdomain: Change the CLKTRCTRL of CM_PCIE_CLKSTCTRL to SW_WKUP The PCIe programming sequence in TRM suggests CLKSTCTRL of PCIe should be set to SW_WKUP. There are no issues when CLKSTCTRL is set to HW_AUTO in RC mode. However in EP mode, the host system is not able to access the MEMSPACE and setting the CLKSTCTRL to SW_WKUP fixes it. Acked-by: Tony Lindgren Signed-off-by: Kishon Vijay Abraham I Signed-off-by: Bjorn Helgaas --- arch/arm/mach-omap2/clockdomains7xx_data.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-omap2/clockdomains7xx_data.c b/arch/arm/mach-omap2/clockdomains7xx_data.c index 6c679659cda5..67ebff829cf2 100644 --- a/arch/arm/mach-omap2/clockdomains7xx_data.c +++ b/arch/arm/mach-omap2/clockdomains7xx_data.c @@ -524,7 +524,7 @@ static struct clockdomain pcie_7xx_clkdm = { .dep_bit = DRA7XX_PCIE_STATDEP_SHIFT, .wkdep_srcs = pcie_wkup_sleep_deps, .sleepdep_srcs = pcie_wkup_sleep_deps, - .flags = CLKDM_CAN_HWSUP_SWSUP, + .flags = CLKDM_CAN_SWSUP, }; static struct clockdomain atl_7xx_clkdm = { -- cgit v1.2.3-59-g8ed1b From a5f40e8098fe6d983fdb3beb7b50a8067c136141 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Thu, 20 Apr 2017 15:36:25 -0500 Subject: PCI: Don't allow unbinding host controllers that aren't prepared Many PCI host controller drivers aren't prepared to have their devices unbound from them forcefully (e.g., through /sys/...//unbind), as they don't provide any driver .remove callback, where they'd detach the root bus, release resources, etc. Keeping the driver built in (i.e., not a loadable module) is not enough; and providing no .remove callback just means we don't do any teardown. To rule out the possibility of unbinding a device via sysfs, we need to set the ".suppress_bind_attrs" field. I found the suspect drivers via the following search: git grep -l platform_driver $(git grep -L -e '\.remove' -e suppress_bind_attrs drivers/pci/) Then I inspected them to ensure that (a) they set up a PCI bus in their probe() and (b) they don't have a remove() callback for undoing the setup Suggested-by: Bjorn Helgaas Signed-off-by: Brian Norris Signed-off-by: Bjorn Helgaas --- drivers/pci/dwc/pci-imx6.c | 1 + drivers/pci/dwc/pci-layerscape.c | 1 + drivers/pci/dwc/pcie-armada8k.c | 1 + drivers/pci/dwc/pcie-artpec6.c | 1 + drivers/pci/dwc/pcie-designware-plat.c | 1 + drivers/pci/dwc/pcie-hisi.c | 2 ++ drivers/pci/dwc/pcie-spear13xx.c | 1 + drivers/pci/host/pci-ftpci100.c | 1 + drivers/pci/host/pci-host-generic.c | 1 + drivers/pci/host/pci-thunder-ecam.c | 1 + drivers/pci/host/pci-thunder-pem.c | 1 + drivers/pci/host/pci-versatile.c | 1 + drivers/pci/host/pci-xgene.c | 1 + 13 files changed, 14 insertions(+) diff --git a/drivers/pci/dwc/pci-imx6.c b/drivers/pci/dwc/pci-imx6.c index 129717ae5022..a98cba55c7f0 100644 --- a/drivers/pci/dwc/pci-imx6.c +++ b/drivers/pci/dwc/pci-imx6.c @@ -804,6 +804,7 @@ static struct platform_driver imx6_pcie_driver = { .driver = { .name = "imx6q-pcie", .of_match_table = imx6_pcie_of_match, + .suppress_bind_attrs = true, }, .probe = imx6_pcie_probe, .shutdown = imx6_pcie_shutdown, diff --git a/drivers/pci/dwc/pci-layerscape.c b/drivers/pci/dwc/pci-layerscape.c index 8f0ee0d8129c..27d638c4e134 100644 --- a/drivers/pci/dwc/pci-layerscape.c +++ b/drivers/pci/dwc/pci-layerscape.c @@ -305,6 +305,7 @@ static struct platform_driver ls_pcie_driver = { .driver = { .name = "layerscape-pcie", .of_match_table = ls_pcie_of_match, + .suppress_bind_attrs = true, }, }; builtin_platform_driver_probe(ls_pcie_driver, ls_pcie_probe); diff --git a/drivers/pci/dwc/pcie-armada8k.c b/drivers/pci/dwc/pcie-armada8k.c index 3ff31301323b..495b023042b3 100644 --- a/drivers/pci/dwc/pcie-armada8k.c +++ b/drivers/pci/dwc/pcie-armada8k.c @@ -262,6 +262,7 @@ static struct platform_driver armada8k_pcie_driver = { .driver = { .name = "armada8k-pcie", .of_match_table = of_match_ptr(armada8k_pcie_of_match), + .suppress_bind_attrs = true, }, }; builtin_platform_driver(armada8k_pcie_driver); diff --git a/drivers/pci/dwc/pcie-artpec6.c b/drivers/pci/dwc/pcie-artpec6.c index 5b3b3afc0edb..82a04acc42fd 100644 --- a/drivers/pci/dwc/pcie-artpec6.c +++ b/drivers/pci/dwc/pcie-artpec6.c @@ -295,6 +295,7 @@ static struct platform_driver artpec6_pcie_driver = { .driver = { .name = "artpec6-pcie", .of_match_table = artpec6_pcie_of_match, + .suppress_bind_attrs = true, }, }; builtin_platform_driver(artpec6_pcie_driver); diff --git a/drivers/pci/dwc/pcie-designware-plat.c b/drivers/pci/dwc/pcie-designware-plat.c index f20d494922ab..32091b32f6e1 100644 --- a/drivers/pci/dwc/pcie-designware-plat.c +++ b/drivers/pci/dwc/pcie-designware-plat.c @@ -133,6 +133,7 @@ static struct platform_driver dw_plat_pcie_driver = { .driver = { .name = "dw-pcie", .of_match_table = dw_plat_pcie_of_match, + .suppress_bind_attrs = true, }, .probe = dw_plat_pcie_probe, }; diff --git a/drivers/pci/dwc/pcie-hisi.c b/drivers/pci/dwc/pcie-hisi.c index 1606de5bf757..18af8057cdc4 100644 --- a/drivers/pci/dwc/pcie-hisi.c +++ b/drivers/pci/dwc/pcie-hisi.c @@ -333,6 +333,7 @@ static struct platform_driver hisi_pcie_driver = { .driver = { .name = "hisi-pcie", .of_match_table = hisi_pcie_of_match, + .suppress_bind_attrs = true, }, }; builtin_platform_driver(hisi_pcie_driver); @@ -390,6 +391,7 @@ static struct platform_driver hisi_pcie_almost_ecam_driver = { .driver = { .name = "hisi-pcie-almost-ecam", .of_match_table = hisi_pcie_almost_ecam_of_match, + .suppress_bind_attrs = true, }, }; builtin_platform_driver(hisi_pcie_almost_ecam_driver); diff --git a/drivers/pci/dwc/pcie-spear13xx.c b/drivers/pci/dwc/pcie-spear13xx.c index 3ae59de5c043..8ff36b3dbbdf 100644 --- a/drivers/pci/dwc/pcie-spear13xx.c +++ b/drivers/pci/dwc/pcie-spear13xx.c @@ -308,6 +308,7 @@ static struct platform_driver spear13xx_pcie_driver = { .driver = { .name = "spear-pcie", .of_match_table = of_match_ptr(spear13xx_pcie_of_match), + .suppress_bind_attrs = true, }, }; diff --git a/drivers/pci/host/pci-ftpci100.c b/drivers/pci/host/pci-ftpci100.c index c0f29d1de341..d26501c4145a 100644 --- a/drivers/pci/host/pci-ftpci100.c +++ b/drivers/pci/host/pci-ftpci100.c @@ -556,6 +556,7 @@ static struct platform_driver faraday_pci_driver = { .driver = { .name = "ftpci100", .of_match_table = of_match_ptr(faraday_pci_of_match), + .suppress_bind_attrs = true, }, .probe = faraday_pci_probe, }; diff --git a/drivers/pci/host/pci-host-generic.c b/drivers/pci/host/pci-host-generic.c index c05ea9d72f69..7d709a7e0aa8 100644 --- a/drivers/pci/host/pci-host-generic.c +++ b/drivers/pci/host/pci-host-generic.c @@ -60,6 +60,7 @@ static struct platform_driver gen_pci_driver = { .driver = { .name = "pci-host-generic", .of_match_table = gen_pci_of_match, + .suppress_bind_attrs = true, }, .probe = gen_pci_probe, }; diff --git a/drivers/pci/host/pci-thunder-ecam.c b/drivers/pci/host/pci-thunder-ecam.c index 3f54a43bbbea..fc0ca03f280e 100644 --- a/drivers/pci/host/pci-thunder-ecam.c +++ b/drivers/pci/host/pci-thunder-ecam.c @@ -373,6 +373,7 @@ static struct platform_driver thunder_ecam_driver = { .driver = { .name = KBUILD_MODNAME, .of_match_table = thunder_ecam_of_match, + .suppress_bind_attrs = true, }, .probe = thunder_ecam_probe, }; diff --git a/drivers/pci/host/pci-thunder-pem.c b/drivers/pci/host/pci-thunder-pem.c index 6e031b522529..6e066f8b74df 100644 --- a/drivers/pci/host/pci-thunder-pem.c +++ b/drivers/pci/host/pci-thunder-pem.c @@ -474,6 +474,7 @@ static struct platform_driver thunder_pem_driver = { .driver = { .name = KBUILD_MODNAME, .of_match_table = thunder_pem_of_match, + .suppress_bind_attrs = true, }, .probe = thunder_pem_probe, }; diff --git a/drivers/pci/host/pci-versatile.c b/drivers/pci/host/pci-versatile.c index 85e773661bc8..9281eee2d000 100644 --- a/drivers/pci/host/pci-versatile.c +++ b/drivers/pci/host/pci-versatile.c @@ -222,6 +222,7 @@ static struct platform_driver versatile_pci_driver = { .driver = { .name = "versatile-pci", .of_match_table = versatile_pci_of_match, + .suppress_bind_attrs = true, }, .probe = versatile_pci_probe, }; diff --git a/drivers/pci/host/pci-xgene.c b/drivers/pci/host/pci-xgene.c index de198980432e..8cae013e7188 100644 --- a/drivers/pci/host/pci-xgene.c +++ b/drivers/pci/host/pci-xgene.c @@ -697,6 +697,7 @@ static struct platform_driver xgene_pcie_driver = { .driver = { .name = "xgene-pcie", .of_match_table = of_match_ptr(xgene_pcie_match_table), + .suppress_bind_attrs = true, }, .probe = xgene_pcie_probe_bridge, }; -- cgit v1.2.3-59-g8ed1b