diff options
Diffstat (limited to 'drivers/hv/vmbus_drv.c')
| -rw-r--r-- | drivers/hv/vmbus_drv.c | 2065 |
1 files changed, 1527 insertions, 538 deletions
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c index d0ff65675292..a53af6fe81a6 100644 --- a/drivers/hv/vmbus_drv.c +++ b/drivers/hv/vmbus_drv.c @@ -1,30 +1,18 @@ +// SPDX-License-Identifier: GPL-2.0-only /* * Copyright (c) 2009, Microsoft 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. - * - * You should have received a copy of the GNU General Public License along with - * this program; if not, write to the Free Software Foundation, Inc., 59 Temple - * Place - Suite 330, Boston, MA 02111-1307 USA. - * * Authors: * Haiyang Zhang <haiyangz@microsoft.com> * Hank Janssen <hjanssen@microsoft.com> * K. Y. Srinivasan <kys@microsoft.com> - * */ #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt #include <linux/init.h> #include <linux/module.h> #include <linux/device.h> +#include <linux/platform_device.h> #include <linux/interrupt.h> #include <linux/sysctl.h> #include <linux/slab.h> @@ -32,17 +20,25 @@ #include <linux/completion.h> #include <linux/hyperv.h> #include <linux/kernel_stat.h> +#include <linux/of_address.h> #include <linux/clockchips.h> #include <linux/cpu.h> +#include <linux/sched/isolation.h> #include <linux/sched/task_stack.h> -#include <asm/mshyperv.h> -#include <linux/notifier.h> +#include <linux/delay.h> +#include <linux/panic_notifier.h> #include <linux/ptrace.h> #include <linux/screen_info.h> -#include <linux/kdebug.h> #include <linux/efi.h> #include <linux/random.h> +#include <linux/kernel.h> +#include <linux/syscore_ops.h> +#include <linux/dma-map-ops.h> +#include <linux/pci.h> +#include <linux/export.h> +#include <clocksource/hyperv_timer.h> +#include <asm/mshyperv.h> #include "hyperv_vmbus.h" struct vmbus_dynid { @@ -50,63 +46,68 @@ struct vmbus_dynid { struct hv_vmbus_device_id id; }; -static struct acpi_device *hv_acpi_dev; - -static struct completion probe_event; +/* VMBus Root Device */ +static struct device *vmbus_root_device; static int hyperv_cpuhp_online; -static void *hv_panic_page; +static long __percpu *vmbus_evt; -static int hyperv_panic_event(struct notifier_block *nb, unsigned long val, - void *args) -{ - struct pt_regs *regs; +/* Values parsed from ACPI DSDT */ +int vmbus_irq; +int vmbus_interrupt; - regs = current_pt_regs(); +/* + * If the Confidential VMBus is used, the data on the "wire" is not + * visible to either the host or the hypervisor. + */ +static bool is_confidential; - hyperv_report_panic(regs, val); - return NOTIFY_DONE; +bool vmbus_is_confidential(void) +{ + return is_confidential; } +EXPORT_SYMBOL_GPL(vmbus_is_confidential); -static int hyperv_die_event(struct notifier_block *nb, unsigned long val, - void *args) +/* + * The panic notifier below is responsible solely for unloading the + * vmbus connection, which is necessary in a panic event. + * + * Notice an intrincate relation of this notifier with Hyper-V + * framebuffer panic notifier exists - we need vmbus connection alive + * there in order to succeed, so we need to order both with each other + * [see hvfb_on_panic()] - this is done using notifiers' priorities. + */ +static int hv_panic_vmbus_unload(struct notifier_block *nb, unsigned long val, + void *args) { - struct die_args *die = (struct die_args *)args; - struct pt_regs *regs = die->regs; - - hyperv_report_panic(regs, val); + vmbus_initiate_unload(true); return NOTIFY_DONE; } - -static struct notifier_block hyperv_die_block = { - .notifier_call = hyperv_die_event, -}; -static struct notifier_block hyperv_panic_block = { - .notifier_call = hyperv_panic_event, +static struct notifier_block hyperv_panic_vmbus_unload_block = { + .notifier_call = hv_panic_vmbus_unload, + .priority = INT_MIN + 1, /* almost the latest one to execute */ }; static const char *fb_mmio_name = "fb_range"; static struct resource *fb_mmio; static struct resource *hyperv_mmio; -static DEFINE_SEMAPHORE(hyperv_mmio_lock); +static DEFINE_MUTEX(hyperv_mmio_lock); + +struct device *hv_get_vmbus_root_device(void) +{ + return vmbus_root_device; +} +EXPORT_SYMBOL_GPL(hv_get_vmbus_root_device); static int vmbus_exists(void) { - if (hv_acpi_dev == NULL) + if (vmbus_root_device == NULL) return -ENODEV; return 0; } -#define VMBUS_ALIAS_LEN ((sizeof((struct hv_vmbus_device_id *)0)->guid) * 2) -static void print_alias_name(struct hv_device *hv_dev, char *alias_name) -{ - int i; - for (i = 0; i < VMBUS_ALIAS_LEN; i += 2) - sprintf(&alias_name[i], "%02x", hv_dev->dev_type.b[i/2]); -} - static u8 channel_monitor_group(const struct vmbus_channel *channel) { return (u8)channel->offermsg.monitorid / 32; @@ -139,6 +140,7 @@ static u32 channel_conn_id(struct vmbus_channel *channel, { u8 monitor_group = channel_monitor_group(channel); u8 monitor_offset = channel_monitor_offset(channel); + return monitor_page->parameter[monitor_group][monitor_offset].connectionid.u.id; } @@ -149,7 +151,7 @@ static ssize_t id_show(struct device *dev, struct device_attribute *dev_attr, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", hv_dev->channel->offermsg.child_relid); + return sysfs_emit(buf, "%d\n", hv_dev->channel->offermsg.child_relid); } static DEVICE_ATTR_RO(id); @@ -160,7 +162,7 @@ static ssize_t state_show(struct device *dev, struct device_attribute *dev_attr, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", hv_dev->channel->state); + return sysfs_emit(buf, "%d\n", hv_dev->channel->state); } static DEVICE_ATTR_RO(state); @@ -171,7 +173,7 @@ static ssize_t monitor_id_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", hv_dev->channel->offermsg.monitorid); + return sysfs_emit(buf, "%d\n", hv_dev->channel->offermsg.monitorid); } static DEVICE_ATTR_RO(monitor_id); @@ -182,8 +184,8 @@ static ssize_t class_id_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "{%pUl}\n", - hv_dev->channel->offermsg.offer.if_type.b); + return sysfs_emit(buf, "{%pUl}\n", + &hv_dev->channel->offermsg.offer.if_type); } static DEVICE_ATTR_RO(class_id); @@ -194,8 +196,8 @@ static ssize_t device_id_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "{%pUl}\n", - hv_dev->channel->offermsg.offer.if_instance.b); + return sysfs_emit(buf, "{%pUl}\n", + &hv_dev->channel->offermsg.offer.if_instance); } static DEVICE_ATTR_RO(device_id); @@ -203,10 +205,8 @@ static ssize_t modalias_show(struct device *dev, struct device_attribute *dev_attr, char *buf) { struct hv_device *hv_dev = device_to_hv_device(dev); - char alias_name[VMBUS_ALIAS_LEN + 1]; - print_alias_name(hv_dev, alias_name); - return sprintf(buf, "vmbus:%s\n", alias_name); + return sysfs_emit(buf, "vmbus:%*phN\n", UUID_SIZE, &hv_dev->dev_type); } static DEVICE_ATTR_RO(modalias); @@ -219,7 +219,7 @@ static ssize_t numa_node_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", hv_dev->channel->numa_node); + return sysfs_emit(buf, "%d\n", cpu_to_node(hv_dev->channel->target_cpu)); } static DEVICE_ATTR_RO(numa_node); #endif @@ -232,9 +232,8 @@ static ssize_t server_monitor_pending_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_pending(hv_dev->channel, - vmbus_connection.monitor_pages[1])); + return sysfs_emit(buf, "%d\n", channel_pending(hv_dev->channel, + vmbus_connection.monitor_pages[0])); } static DEVICE_ATTR_RO(server_monitor_pending); @@ -246,9 +245,8 @@ static ssize_t client_monitor_pending_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_pending(hv_dev->channel, - vmbus_connection.monitor_pages[1])); + return sysfs_emit(buf, "%d\n", channel_pending(hv_dev->channel, + vmbus_connection.monitor_pages[1])); } static DEVICE_ATTR_RO(client_monitor_pending); @@ -260,9 +258,8 @@ static ssize_t server_monitor_latency_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_latency(hv_dev->channel, - vmbus_connection.monitor_pages[0])); + return sysfs_emit(buf, "%d\n", channel_latency(hv_dev->channel, + vmbus_connection.monitor_pages[0])); } static DEVICE_ATTR_RO(server_monitor_latency); @@ -274,9 +271,8 @@ static ssize_t client_monitor_latency_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_latency(hv_dev->channel, - vmbus_connection.monitor_pages[1])); + return sysfs_emit(buf, "%d\n", channel_latency(hv_dev->channel, + vmbus_connection.monitor_pages[1])); } static DEVICE_ATTR_RO(client_monitor_latency); @@ -288,9 +284,8 @@ static ssize_t server_monitor_conn_id_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_conn_id(hv_dev->channel, - vmbus_connection.monitor_pages[0])); + return sysfs_emit(buf, "%d\n", channel_conn_id(hv_dev->channel, + vmbus_connection.monitor_pages[0])); } static DEVICE_ATTR_RO(server_monitor_conn_id); @@ -302,9 +297,8 @@ static ssize_t client_monitor_conn_id_show(struct device *dev, if (!hv_dev->channel) return -ENODEV; - return sprintf(buf, "%d\n", - channel_conn_id(hv_dev->channel, - vmbus_connection.monitor_pages[1])); + return sysfs_emit(buf, "%d\n", channel_conn_id(hv_dev->channel, + vmbus_connection.monitor_pages[1])); } static DEVICE_ATTR_RO(client_monitor_conn_id); @@ -313,13 +307,17 @@ static ssize_t out_intr_mask_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info outbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); - return sprintf(buf, "%d\n", outbound.current_interrupt_mask); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", outbound.current_interrupt_mask); } static DEVICE_ATTR_RO(out_intr_mask); @@ -328,13 +326,16 @@ static ssize_t out_read_index_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info outbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); - return sprintf(buf, "%d\n", outbound.current_read_index); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sysfs_emit(buf, "%u\n", outbound.current_read_index); } static DEVICE_ATTR_RO(out_read_index); @@ -344,13 +345,16 @@ static ssize_t out_write_index_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info outbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); - return sprintf(buf, "%d\n", outbound.current_write_index); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sysfs_emit(buf, "%u\n", outbound.current_write_index); } static DEVICE_ATTR_RO(out_write_index); @@ -360,13 +364,16 @@ static ssize_t out_read_bytes_avail_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info outbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); - return sprintf(buf, "%d\n", outbound.bytes_avail_toread); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sysfs_emit(buf, "%d\n", outbound.bytes_avail_toread); } static DEVICE_ATTR_RO(out_read_bytes_avail); @@ -376,13 +383,16 @@ static ssize_t out_write_bytes_avail_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info outbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); - return sprintf(buf, "%d\n", outbound.bytes_avail_towrite); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sysfs_emit(buf, "%d\n", outbound.bytes_avail_towrite); } static DEVICE_ATTR_RO(out_write_bytes_avail); @@ -391,13 +401,16 @@ static ssize_t in_intr_mask_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info inbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); - return sprintf(buf, "%d\n", inbound.current_interrupt_mask); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", inbound.current_interrupt_mask); } static DEVICE_ATTR_RO(in_intr_mask); @@ -406,13 +419,16 @@ static ssize_t in_read_index_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info inbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); - return sprintf(buf, "%d\n", inbound.current_read_index); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", inbound.current_read_index); } static DEVICE_ATTR_RO(in_read_index); @@ -421,13 +437,16 @@ static ssize_t in_write_index_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info inbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); - return sprintf(buf, "%d\n", inbound.current_write_index); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", inbound.current_write_index); } static DEVICE_ATTR_RO(in_write_index); @@ -437,13 +456,16 @@ static ssize_t in_read_bytes_avail_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info inbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); - return sprintf(buf, "%d\n", inbound.bytes_avail_toread); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", inbound.bytes_avail_toread); } static DEVICE_ATTR_RO(in_read_bytes_avail); @@ -453,13 +475,16 @@ static ssize_t in_write_bytes_avail_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct hv_ring_buffer_debug_info inbound; + int ret; if (!hv_dev->channel) return -ENODEV; - if (hv_dev->channel->state != CHANNEL_OPENED_STATE) - return -EINVAL; - hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); - return sprintf(buf, "%d\n", inbound.bytes_avail_towrite); + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sysfs_emit(buf, "%d\n", inbound.bytes_avail_towrite); } static DEVICE_ATTR_RO(in_write_bytes_avail); @@ -469,34 +494,29 @@ static ssize_t channel_vp_mapping_show(struct device *dev, { struct hv_device *hv_dev = device_to_hv_device(dev); struct vmbus_channel *channel = hv_dev->channel, *cur_sc; - unsigned long flags; - int buf_size = PAGE_SIZE, n_written, tot_written; + int n_written; struct list_head *cur; if (!channel) return -ENODEV; - tot_written = snprintf(buf, buf_size, "%u:%u\n", - channel->offermsg.child_relid, channel->target_cpu); + mutex_lock(&vmbus_connection.channel_mutex); - spin_lock_irqsave(&channel->lock, flags); + n_written = sysfs_emit(buf, "%u:%u\n", + channel->offermsg.child_relid, + channel->target_cpu); list_for_each(cur, &channel->sc_list) { - if (tot_written >= buf_size - 1) - break; cur_sc = list_entry(cur, struct vmbus_channel, sc_list); - n_written = scnprintf(buf + tot_written, - buf_size - tot_written, - "%u:%u\n", - cur_sc->offermsg.child_relid, - cur_sc->target_cpu); - tot_written += n_written; + n_written += sysfs_emit_at(buf, n_written, "%u:%u\n", + cur_sc->offermsg.child_relid, + cur_sc->target_cpu); } - spin_unlock_irqrestore(&channel->lock, flags); + mutex_unlock(&vmbus_connection.channel_mutex); - return tot_written; + return n_written; } static DEVICE_ATTR_RO(channel_vp_mapping); @@ -505,7 +525,8 @@ static ssize_t vendor_show(struct device *dev, char *buf) { struct hv_device *hv_dev = device_to_hv_device(dev); - return sprintf(buf, "0x%x\n", hv_dev->vendor_id); + + return sysfs_emit(buf, "0x%x\n", hv_dev->vendor_id); } static DEVICE_ATTR_RO(vendor); @@ -514,7 +535,8 @@ static ssize_t device_show(struct device *dev, char *buf) { struct hv_device *hv_dev = device_to_hv_device(dev); - return sprintf(buf, "0x%x\n", hv_dev->device_id); + + return sysfs_emit(buf, "0x%x\n", hv_dev->device_id); } static DEVICE_ATTR_RO(device); @@ -523,31 +545,11 @@ static ssize_t driver_override_store(struct device *dev, const char *buf, size_t count) { struct hv_device *hv_dev = device_to_hv_device(dev); - char *driver_override, *old, *cp; - - /* We need to keep extra room for a newline */ - if (count >= (PAGE_SIZE - 1)) - return -EINVAL; - - driver_override = kstrndup(buf, count, GFP_KERNEL); - if (!driver_override) - return -ENOMEM; - - cp = strchr(driver_override, '\n'); - if (cp) - *cp = '\0'; - - device_lock(dev); - old = hv_dev->driver_override; - if (strlen(driver_override)) { - hv_dev->driver_override = driver_override; - } else { - kfree(driver_override); - hv_dev->driver_override = NULL; - } - device_unlock(dev); + int ret; - kfree(old); + ret = driver_set_override(dev, &hv_dev->driver_override, buf, count); + if (ret) + return ret; return count; } @@ -559,7 +561,7 @@ static ssize_t driver_override_show(struct device *dev, ssize_t len; device_lock(dev); - len = snprintf(buf, PAGE_SIZE, "%s\n", hv_dev->driver_override); + len = sysfs_emit(buf, "%s\n", hv_dev->driver_override); device_unlock(dev); return len; @@ -599,7 +601,53 @@ static struct attribute *vmbus_dev_attrs[] = { &dev_attr_driver_override.attr, NULL, }; -ATTRIBUTE_GROUPS(vmbus_dev); + +/* + * Device-level attribute_group callback function. Returns the permission for + * each attribute, and returns 0 if an attribute is not visible. + */ +static umode_t vmbus_dev_attr_is_visible(struct kobject *kobj, + struct attribute *attr, int idx) +{ + struct device *dev = kobj_to_dev(kobj); + const struct hv_device *hv_dev = device_to_hv_device(dev); + + /* Hide the monitor attributes if the monitor mechanism is not used. */ + if (!hv_dev->channel->offermsg.monitor_allocated && + (attr == &dev_attr_monitor_id.attr || + attr == &dev_attr_server_monitor_pending.attr || + attr == &dev_attr_client_monitor_pending.attr || + attr == &dev_attr_server_monitor_latency.attr || + attr == &dev_attr_client_monitor_latency.attr || + attr == &dev_attr_server_monitor_conn_id.attr || + attr == &dev_attr_client_monitor_conn_id.attr)) + return 0; + + return attr->mode; +} + +static const struct attribute_group vmbus_dev_group = { + .attrs = vmbus_dev_attrs, + .is_visible = vmbus_dev_attr_is_visible +}; +__ATTRIBUTE_GROUPS(vmbus_dev); + +/* Set up the attribute for /sys/bus/vmbus/hibernation */ +static ssize_t hibernation_show(const struct bus_type *bus, char *buf) +{ + return sprintf(buf, "%d\n", !!hv_is_hibernation_supported()); +} + +static BUS_ATTR_RO(hibernation); + +static struct attribute *vmbus_bus_attrs[] = { + &bus_attr_hibernation.attr, + NULL, +}; +static const struct attribute_group vmbus_bus_group = { + .attrs = vmbus_bus_attrs, +}; +__ATTRIBUTE_GROUPS(vmbus_bus); /* * vmbus_uevent - add uevent for our device @@ -612,49 +660,36 @@ ATTRIBUTE_GROUPS(vmbus_dev); * representation of the device guid (each byte of the guid will be * represented with two hex characters. */ -static int vmbus_uevent(struct device *device, struct kobj_uevent_env *env) +static int vmbus_uevent(const struct device *device, struct kobj_uevent_env *env) { - struct hv_device *dev = device_to_hv_device(device); - int ret; - char alias_name[VMBUS_ALIAS_LEN + 1]; - - print_alias_name(dev, alias_name); - ret = add_uevent_var(env, "MODALIAS=vmbus:%s", alias_name); - return ret; -} - -static const uuid_le null_guid; + const struct hv_device *dev = device_to_hv_device(device); + const char *format = "MODALIAS=vmbus:%*phN"; -static inline bool is_null_guid(const uuid_le *guid) -{ - if (uuid_le_cmp(*guid, null_guid)) - return false; - return true; + return add_uevent_var(env, format, UUID_SIZE, &dev->dev_type); } static const struct hv_vmbus_device_id * -hv_vmbus_dev_match(const struct hv_vmbus_device_id *id, const uuid_le *guid) - +hv_vmbus_dev_match(const struct hv_vmbus_device_id *id, const guid_t *guid) { if (id == NULL) return NULL; /* empty device table */ - for (; !is_null_guid(&id->guid); id++) - if (!uuid_le_cmp(id->guid, *guid)) + for (; !guid_is_null(&id->guid); id++) + if (guid_equal(&id->guid, guid)) return id; return NULL; } static const struct hv_vmbus_device_id * -hv_vmbus_dynid_match(struct hv_driver *drv, const uuid_le *guid) +hv_vmbus_dynid_match(struct hv_driver *drv, const guid_t *guid) { const struct hv_vmbus_device_id *id = NULL; struct vmbus_dynid *dynid; spin_lock(&drv->dynids.lock); list_for_each_entry(dynid, &drv->dynids.list, node) { - if (!uuid_le_cmp(dynid->id.guid, *guid)) { + if (guid_equal(&dynid->id.guid, guid)) { id = &dynid->id; break; } @@ -664,18 +699,16 @@ hv_vmbus_dynid_match(struct hv_driver *drv, const uuid_le *guid) return id; } -static const struct hv_vmbus_device_id vmbus_device_null = { - .guid = NULL_UUID_LE, -}; +static const struct hv_vmbus_device_id vmbus_device_null; /* * Return a matching hv_vmbus_device_id pointer. * If there is no match, return NULL. */ -static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv, +static const struct hv_vmbus_device_id *hv_vmbus_get_id(const struct hv_driver *drv, struct hv_device *dev) { - const uuid_le *guid = &dev->dev_type; + const guid_t *guid = &dev->dev_type; const struct hv_vmbus_device_id *id; /* When driver_override is set, only bind to the matching driver */ @@ -683,7 +716,7 @@ static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv, return NULL; /* Look at the dynamic ids first, before the static ones */ - id = hv_vmbus_dynid_match(drv, guid); + id = hv_vmbus_dynid_match((struct hv_driver *)drv, guid); if (!id) id = hv_vmbus_dev_match(drv->id_table, guid); @@ -694,8 +727,31 @@ static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv, return id; } -/* vmbus_add_dynid - add a new device ID to this driver and re-probe devices */ -static int vmbus_add_dynid(struct hv_driver *drv, uuid_le *guid) +/* vmbus_add_dynid - add a new device ID to this driver and re-probe devices + * + * This function can race with vmbus_device_register(). This function is + * typically running on a user thread in response to writing to the "new_id" + * sysfs entry for a driver. vmbus_device_register() is running on a + * workqueue thread in response to the Hyper-V host offering a device to the + * guest. This function calls driver_attach(), which looks for an existing + * device matching the new id, and attaches the driver to which the new id + * has been assigned. vmbus_device_register() calls device_register(), which + * looks for a driver that matches the device being registered. If both + * operations are running simultaneously, the device driver probe function runs + * on whichever thread establishes the linkage between the driver and device. + * + * In most cases, it doesn't matter which thread runs the driver probe + * function. But if vmbus_device_register() does not find a matching driver, + * it proceeds to create the "channels" subdirectory and numbered per-channel + * subdirectory in sysfs. While that multi-step creation is in progress, this + * function could run the driver probe function. If the probe function checks + * for, or operates on, entries in the "channels" subdirectory, including by + * calling hv_create_ring_sysfs(), the operation may or may not succeed + * depending on the race. The race can't create a kernel failure in VMBus + * or device subsystem code, but probe functions in VMBus drivers doing such + * operations must be prepared for the failure case. + */ +static int vmbus_add_dynid(struct hv_driver *drv, guid_t *guid) { struct vmbus_dynid *dynid; @@ -733,10 +789,10 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, size_t count) { struct hv_driver *drv = drv_to_hv_drv(driver); - uuid_le guid; + guid_t guid; ssize_t retval; - retval = uuid_le_to_bin(buf, &guid); + retval = guid_parse(buf, &guid); if (retval) return retval; @@ -760,10 +816,10 @@ static ssize_t remove_id_store(struct device_driver *driver, const char *buf, { struct hv_driver *drv = drv_to_hv_drv(driver); struct vmbus_dynid *dynid, *n; - uuid_le guid; + guid_t guid; ssize_t retval; - retval = uuid_le_to_bin(buf, &guid); + retval = guid_parse(buf, &guid); if (retval) return retval; @@ -772,7 +828,7 @@ static ssize_t remove_id_store(struct device_driver *driver, const char *buf, list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) { struct hv_vmbus_device_id *id = &dynid->id; - if (!uuid_le_cmp(id->guid, guid)) { + if (guid_equal(&id->guid, &guid)) { list_del(&dynid->node); kfree(dynid); retval = count; @@ -796,9 +852,9 @@ ATTRIBUTE_GROUPS(vmbus_drv); /* * vmbus_match - Attempt to match the specified device to the specified driver */ -static int vmbus_match(struct device *device, struct device_driver *driver) +static int vmbus_match(struct device *device, const struct device_driver *driver) { - struct hv_driver *drv = drv_to_hv_drv(driver); + const struct hv_driver *drv = drv_to_hv_drv(driver); struct hv_device *hv_dev = device_to_hv_device(device); /* The hv_sock driver handles all hv_sock offers. */ @@ -838,9 +894,24 @@ static int vmbus_probe(struct device *child_device) } /* + * vmbus_dma_configure -- Configure DMA coherence for VMbus device + */ +static int vmbus_dma_configure(struct device *child_device) +{ + /* + * On ARM64, propagate the DMA coherence setting from the top level + * VMbus ACPI device to the child VMbus device being added here. + * On x86/x64 coherence is assumed and these calls have no effect. + */ + hv_setup_dma_ops(child_device, + device_get_dma_attr(vmbus_root_device) == DEV_DMA_COHERENT); + return 0; +} + +/* * vmbus_remove - Remove a vmbus device */ -static int vmbus_remove(struct device *child_device) +static void vmbus_remove(struct device *child_device) { struct hv_driver *drv; struct hv_device *dev = device_to_hv_device(child_device); @@ -850,11 +921,8 @@ static int vmbus_remove(struct device *child_device) if (drv->remove) drv->remove(dev); } - - return 0; } - /* * vmbus_shutdown - Shutdown a vmbus device */ @@ -874,6 +942,48 @@ static void vmbus_shutdown(struct device *child_device) drv->shutdown(dev); } +#ifdef CONFIG_PM_SLEEP +/* + * vmbus_suspend - Suspend a vmbus device + */ +static int vmbus_suspend(struct device *child_device) +{ + struct hv_driver *drv; + struct hv_device *dev = device_to_hv_device(child_device); + + /* The device may not be attached yet */ + if (!child_device->driver) + return 0; + + drv = drv_to_hv_drv(child_device->driver); + if (!drv->suspend) + return -EOPNOTSUPP; + + return drv->suspend(dev); +} + +/* + * vmbus_resume - Resume a vmbus device + */ +static int vmbus_resume(struct device *child_device) +{ + struct hv_driver *drv; + struct hv_device *dev = device_to_hv_device(child_device); + + /* The device may not be attached yet */ + if (!child_device->driver) + return 0; + + drv = drv_to_hv_drv(child_device->driver); + if (!drv->resume) + return -EOPNOTSUPP; + + return drv->resume(dev); +} +#else +#define vmbus_suspend NULL +#define vmbus_resume NULL +#endif /* CONFIG_PM_SLEEP */ /* * vmbus_device_release - Final callback release of the vmbus child device @@ -883,27 +993,54 @@ static void vmbus_device_release(struct device *device) struct hv_device *hv_dev = device_to_hv_device(device); struct vmbus_channel *channel = hv_dev->channel; + hv_debug_rm_dev_dir(hv_dev); + mutex_lock(&vmbus_connection.channel_mutex); hv_process_channel_removal(channel); mutex_unlock(&vmbus_connection.channel_mutex); kfree(hv_dev); } +/* + * Note: we must use the "noirq" ops: see the comment before vmbus_bus_pm. + * + * suspend_noirq/resume_noirq are set to NULL to support Suspend-to-Idle: we + * shouldn't suspend the vmbus devices upon Suspend-to-Idle, otherwise there + * is no way to wake up a Generation-2 VM. + * + * The other 4 ops are for hibernation. + */ + +static const struct dev_pm_ops vmbus_pm = { + .suspend_noirq = NULL, + .resume_noirq = NULL, + .freeze_noirq = vmbus_suspend, + .thaw_noirq = vmbus_resume, + .poweroff_noirq = vmbus_suspend, + .restore_noirq = vmbus_resume, +}; + /* The one and only one */ -static struct bus_type hv_bus = { +static const struct bus_type hv_bus = { .name = "vmbus", .match = vmbus_match, .shutdown = vmbus_shutdown, .remove = vmbus_remove, .probe = vmbus_probe, .uevent = vmbus_uevent, + .dma_configure = vmbus_dma_configure, .dev_groups = vmbus_dev_groups, .drv_groups = vmbus_drv_groups, + .bus_groups = vmbus_bus_groups, + .pm = &vmbus_pm, }; struct onmessage_work_context { struct work_struct work; - struct hv_message msg; + struct { + struct hv_message_header header; + u8 payload[]; + } msg; }; static void vmbus_onmessage_work(struct work_struct *work) @@ -916,76 +1053,139 @@ static void vmbus_onmessage_work(struct work_struct *work) ctx = container_of(work, struct onmessage_work_context, work); - vmbus_onmessage(&ctx->msg); + vmbus_onmessage((struct vmbus_channel_message_header *) + &ctx->msg.payload); kfree(ctx); } -static void hv_process_timer_expiration(struct hv_message *msg, - struct hv_per_cpu_context *hv_cpu) +static void __vmbus_on_msg_dpc(void *message_page_addr) { - struct clock_event_device *dev = hv_cpu->clk_evt; - - if (dev->event_handler) - dev->event_handler(dev); - - vmbus_signal_eom(msg, HVMSG_TIMER_EXPIRED); -} - -void vmbus_on_msg_dpc(unsigned long data) -{ - struct hv_per_cpu_context *hv_cpu = (void *)data; - void *page_addr = hv_cpu->synic_message_page; - struct hv_message *msg = (struct hv_message *)page_addr + - VMBUS_MESSAGE_SINT; + struct hv_message msg_copy, *msg; struct vmbus_channel_message_header *hdr; + enum vmbus_channel_message_type msgtype; const struct vmbus_channel_message_table_entry *entry; struct onmessage_work_context *ctx; - u32 message_type = msg->header.message_type; + __u8 payload_size; + u32 message_type; + + if (!message_page_addr) + return; + msg = (struct hv_message *)message_page_addr + VMBUS_MESSAGE_SINT; + /* + * 'enum vmbus_channel_message_type' is supposed to always be 'u32' as + * it is being used in 'struct vmbus_channel_message_header' definition + * which is supposed to match hypervisor ABI. + */ + BUILD_BUG_ON(sizeof(enum vmbus_channel_message_type) != sizeof(u32)); + + /* + * Since the message is in memory shared with the host, an erroneous or + * malicious Hyper-V could modify the message while vmbus_on_msg_dpc() + * or individual message handlers are executing; to prevent this, copy + * the message into private memory. + */ + memcpy(&msg_copy, msg, sizeof(struct hv_message)); + + message_type = msg_copy.header.message_type; if (message_type == HVMSG_NONE) /* no msg */ return; - hdr = (struct vmbus_channel_message_header *)msg->u.payload; + hdr = (struct vmbus_channel_message_header *)msg_copy.u.payload; + msgtype = hdr->msgtype; trace_vmbus_on_msg_dpc(hdr); - if (hdr->msgtype >= CHANNELMSG_COUNT) { - WARN_ONCE(1, "unknown msgtype=%d\n", hdr->msgtype); + if (msgtype >= CHANNELMSG_COUNT) { + WARN_ONCE(1, "unknown msgtype=%d\n", msgtype); + goto msg_handled; + } + + payload_size = msg_copy.header.payload_size; + if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT) { + WARN_ONCE(1, "payload size is too large (%d)\n", payload_size); + goto msg_handled; + } + + entry = &channel_message_table[msgtype]; + + if (!entry->message_handler) + goto msg_handled; + + if (payload_size < entry->min_payload_len) { + WARN_ONCE(1, "message too short: msgtype=%d len=%d\n", msgtype, payload_size); goto msg_handled; } - entry = &channel_message_table[hdr->msgtype]; if (entry->handler_type == VMHT_BLOCKING) { - ctx = kmalloc(sizeof(*ctx), GFP_ATOMIC); + ctx = kmalloc(struct_size(ctx, msg.payload, payload_size), GFP_ATOMIC); if (ctx == NULL) return; INIT_WORK(&ctx->work, vmbus_onmessage_work); - memcpy(&ctx->msg, msg, sizeof(*msg)); + ctx->msg.header = msg_copy.header; + memcpy(&ctx->msg.payload, msg_copy.u.payload, payload_size); /* * The host can generate a rescind message while we * may still be handling the original offer. We deal with - * this condition by ensuring the processing is done on the - * same CPU. + * this condition by relying on the synchronization provided + * by offer_in_progress and by channel_mutex. See also the + * inline comments in vmbus_onoffer_rescind(). */ - switch (hdr->msgtype) { + switch (msgtype) { case CHANNELMSG_RESCIND_CHANNELOFFER: /* * If we are handling the rescind message; * schedule the work on the global work queue. + * + * The OFFER message and the RESCIND message should + * not be handled by the same serialized work queue, + * because the OFFER handler may call vmbus_open(), + * which tries to open the channel by sending an + * OPEN_CHANNEL message to the host and waits for + * the host's response; however, if the host has + * rescinded the channel before it receives the + * OPEN_CHANNEL message, the host just silently + * ignores the OPEN_CHANNEL message; as a result, + * the guest's OFFER handler hangs for ever, if we + * handle the RESCIND message in the same serialized + * work queue: the RESCIND handler can not start to + * run before the OFFER handler finishes. */ - schedule_work_on(vmbus_connection.connect_cpu, - &ctx->work); + if (vmbus_connection.ignore_any_offer_msg) + break; + queue_work(vmbus_connection.rescind_work_queue, &ctx->work); break; case CHANNELMSG_OFFERCHANNEL: + /* + * The host sends the offer message of a given channel + * before sending the rescind message of the same + * channel. These messages are sent to the guest's + * connect CPU; the guest then starts processing them + * in the tasklet handler on this CPU: + * + * VMBUS_CONNECT_CPU + * + * [vmbus_on_msg_dpc()] + * atomic_inc() // CHANNELMSG_OFFERCHANNEL + * queue_work() + * ... + * [vmbus_on_msg_dpc()] + * schedule_work() // CHANNELMSG_RESCIND_CHANNELOFFER + * + * We rely on the memory-ordering properties of the + * queue_work() and schedule_work() primitives, which + * guarantee that the atomic increment will be visible + * to the CPUs which will execute the offer & rescind + * works by the time these works will start execution. + */ + if (vmbus_connection.ignore_any_offer_msg) + break; atomic_inc(&vmbus_connection.offer_in_progress); - queue_work_on(vmbus_connection.connect_cpu, - vmbus_connection.work_queue, - &ctx->work); - break; + fallthrough; default: queue_work(vmbus_connection.work_queue, &ctx->work); @@ -997,49 +1197,79 @@ msg_handled: vmbus_signal_eom(msg, message_type); } +void vmbus_on_msg_dpc(unsigned long data) +{ + struct hv_per_cpu_context *hv_cpu = (void *)data; + + __vmbus_on_msg_dpc(hv_cpu->hyp_synic_message_page); + __vmbus_on_msg_dpc(hv_cpu->para_synic_message_page); +} +#ifdef CONFIG_PM_SLEEP /* - * Direct callback for channels using other deferred processing + * Fake RESCIND_CHANNEL messages to clean up hv_sock channels by force for + * hibernation, because hv_sock connections can not persist across hibernation. */ -static void vmbus_channel_isr(struct vmbus_channel *channel) +static void vmbus_force_channel_rescinded(struct vmbus_channel *channel) { - void (*callback_fn)(void *); + struct onmessage_work_context *ctx; + struct vmbus_channel_rescind_offer *rescind; + + WARN_ON(!is_hvsock_channel(channel)); + + /* + * Allocation size is small and the allocation should really not fail, + * otherwise the state of the hv_sock connections ends up in limbo. + */ + ctx = kzalloc(sizeof(*ctx) + sizeof(*rescind), + GFP_KERNEL | __GFP_NOFAIL); + + /* + * So far, these are not really used by Linux. Just set them to the + * reasonable values conforming to the definitions of the fields. + */ + ctx->msg.header.message_type = 1; + ctx->msg.header.payload_size = sizeof(*rescind); + + /* These values are actually used by Linux. */ + rescind = (struct vmbus_channel_rescind_offer *)ctx->msg.payload; + rescind->header.msgtype = CHANNELMSG_RESCIND_CHANNELOFFER; + rescind->child_relid = channel->offermsg.child_relid; + + INIT_WORK(&ctx->work, vmbus_onmessage_work); - callback_fn = READ_ONCE(channel->onchannel_callback); - if (likely(callback_fn != NULL)) - (*callback_fn)(channel->channel_callback_context); + queue_work(vmbus_connection.work_queue, &ctx->work); } +#endif /* CONFIG_PM_SLEEP */ /* - * Schedule all channels with events pending + * Schedule all channels with events pending. + * The event page can be directly checked to get the id of + * the channel that has the interrupt pending. */ -static void vmbus_chan_sched(struct hv_per_cpu_context *hv_cpu) +static void vmbus_chan_sched(void *event_page_addr) { unsigned long *recv_int_page; u32 maxbits, relid; + union hv_synic_event_flags *event; - if (vmbus_proto_version < VERSION_WIN8) { - maxbits = MAX_NUM_CHANNELS_SUPPORTED; - recv_int_page = vmbus_connection.recv_int_page; - } else { - /* - * When the host is win8 and beyond, the event page - * can be directly checked to get the id of the channel - * that has the interrupt pending. - */ - void *page_addr = hv_cpu->synic_event_page; - union hv_synic_event_flags *event - = (union hv_synic_event_flags *)page_addr + - VMBUS_MESSAGE_SINT; + if (!event_page_addr) + return; + event = (union hv_synic_event_flags *)event_page_addr + VMBUS_MESSAGE_SINT; - maxbits = HV_EVENT_FLAGS_COUNT; - recv_int_page = event->flags; - } + maxbits = HV_EVENT_FLAGS_COUNT; + recv_int_page = event->flags; if (unlikely(!recv_int_page)) return; + /* + * Suggested-by: Michael Kelley <mhklinux@outlook.com> + * One possible optimization would be to keep track of the largest relID that's in use, + * and only scan up to that relID. + */ for_each_set_bit(relid, recv_int_page, maxbits) { + void (*callback_fn)(void *context); struct vmbus_channel *channel; if (!sync_test_and_clear_bit(relid, recv_int_page)) @@ -1049,156 +1279,157 @@ static void vmbus_chan_sched(struct hv_per_cpu_context *hv_cpu) if (relid == 0) continue; + /* + * Pairs with the kfree_rcu() in vmbus_chan_release(). + * Guarantees that the channel data structure doesn't + * get freed while the channel pointer below is being + * dereferenced. + */ rcu_read_lock(); /* Find channel based on relid */ - list_for_each_entry_rcu(channel, &hv_cpu->chan_list, percpu_list) { - if (channel->offermsg.child_relid != relid) - continue; + channel = relid2channel(relid); + if (channel == NULL) + goto sched_unlock_rcu; - if (channel->rescind) - continue; + if (channel->rescind) + goto sched_unlock_rcu; - trace_vmbus_chan_sched(channel); + /* + * Make sure that the ring buffer data structure doesn't get + * freed while we dereference the ring buffer pointer. Test + * for the channel's onchannel_callback being NULL within a + * sched_lock critical section. See also the inline comments + * in vmbus_reset_channel_cb(). + */ + spin_lock(&channel->sched_lock); - ++channel->interrupts; + callback_fn = channel->onchannel_callback; + if (unlikely(callback_fn == NULL)) + goto sched_unlock; - switch (channel->callback_mode) { - case HV_CALL_ISR: - vmbus_channel_isr(channel); - break; + trace_vmbus_chan_sched(channel); - case HV_CALL_BATCHED: - hv_begin_read(&channel->inbound); - /* fallthrough */ - case HV_CALL_DIRECT: - tasklet_schedule(&channel->callback_event); - } + ++channel->interrupts; + + switch (channel->callback_mode) { + case HV_CALL_ISR: + (*callback_fn)(channel->channel_callback_context); + break; + + case HV_CALL_BATCHED: + hv_begin_read(&channel->inbound); + fallthrough; + case HV_CALL_DIRECT: + tasklet_schedule(&channel->callback_event); } +sched_unlock: + spin_unlock(&channel->sched_lock); +sched_unlock_rcu: rcu_read_unlock(); } } -static void vmbus_isr(void) +static void vmbus_message_sched(struct hv_per_cpu_context *hv_cpu, void *message_page_addr) { - struct hv_per_cpu_context *hv_cpu - = this_cpu_ptr(hv_context.cpu_context); - void *page_addr = hv_cpu->synic_event_page; struct hv_message *msg; - union hv_synic_event_flags *event; - bool handled = false; - if (unlikely(page_addr == NULL)) + if (!message_page_addr) return; - - event = (union hv_synic_event_flags *)page_addr + - VMBUS_MESSAGE_SINT; - /* - * Check for events before checking for messages. This is the order - * in which events and messages are checked in Windows guests on - * Hyper-V, and the Windows team suggested we do the same. - */ - - if ((vmbus_proto_version == VERSION_WS2008) || - (vmbus_proto_version == VERSION_WIN7)) { - - /* Since we are a child, we only need to check bit 0 */ - if (sync_test_and_clear_bit(0, event->flags)) - handled = true; - } else { - /* - * Our host is win8 or above. The signaling mechanism - * has changed and we can directly look at the event page. - * If bit n is set then we have an interrup on the channel - * whose id is n. - */ - handled = true; - } - - if (handled) - vmbus_chan_sched(hv_cpu); - - page_addr = hv_cpu->synic_message_page; - msg = (struct hv_message *)page_addr + VMBUS_MESSAGE_SINT; + msg = (struct hv_message *)message_page_addr + VMBUS_MESSAGE_SINT; /* Check if there are actual msgs to be processed */ if (msg->header.message_type != HVMSG_NONE) { - if (msg->header.message_type == HVMSG_TIMER_EXPIRED) - hv_process_timer_expiration(msg, hv_cpu); - else + if (msg->header.message_type == HVMSG_TIMER_EXPIRED) { + hv_stimer0_isr(); + vmbus_signal_eom(msg, HVMSG_TIMER_EXPIRED); + } else { tasklet_schedule(&hv_cpu->msg_dpc); + } } +} + +void vmbus_isr(void) +{ + struct hv_per_cpu_context *hv_cpu + = this_cpu_ptr(hv_context.cpu_context); + + vmbus_chan_sched(hv_cpu->hyp_synic_event_page); + vmbus_chan_sched(hv_cpu->para_synic_event_page); - add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR, 0); + vmbus_message_sched(hv_cpu, hv_cpu->hyp_synic_message_page); + vmbus_message_sched(hv_cpu, hv_cpu->para_synic_message_page); + + add_interrupt_randomness(vmbus_interrupt); } +EXPORT_SYMBOL_FOR_MODULES(vmbus_isr, "mshv_vtl"); -/* - * Boolean to control whether to report panic messages over Hyper-V. - * - * It can be set via /proc/sys/kernel/hyperv/record_panic_msg - */ -static int sysctl_record_panic_msg = 1; +static irqreturn_t vmbus_percpu_isr(int irq, void *dev_id) +{ + vmbus_isr(); + return IRQ_HANDLED; +} -/* - * Callback from kmsg_dump. Grab as much as possible from the end of the kmsg - * buffer and call into Hyper-V to transfer the data. - */ -static void hv_kmsg_dump(struct kmsg_dumper *dumper, - enum kmsg_dump_reason reason) +static void vmbus_percpu_work(struct work_struct *work) { - size_t bytes_written; - phys_addr_t panic_pa; + unsigned int cpu = smp_processor_id(); - /* We are only interested in panics. */ - if ((reason != KMSG_DUMP_PANIC) || (!sysctl_record_panic_msg)) - return; + hv_synic_init(cpu); +} - panic_pa = virt_to_phys(hv_panic_page); +static int vmbus_alloc_synic_and_connect(void) +{ + int ret, cpu; + struct work_struct __percpu *works; + int hyperv_cpuhp_online; + + ret = hv_synic_alloc(); + if (ret < 0) + goto err_alloc; + + works = alloc_percpu(struct work_struct); + if (!works) { + ret = -ENOMEM; + goto err_alloc; + } /* - * Write dump contents to the page. No need to synchronize; panic should - * be single-threaded. + * Initialize the per-cpu interrupt state and stimer state. + * Then connect to the host. */ - kmsg_dump_get_buffer(dumper, true, hv_panic_page, PAGE_SIZE, - &bytes_written); - if (bytes_written) - hyperv_report_panic_msg(panic_pa, bytes_written); -} + cpus_read_lock(); + for_each_online_cpu(cpu) { + struct work_struct *work = per_cpu_ptr(works, cpu); -static struct kmsg_dumper hv_kmsg_dumper = { - .dump = hv_kmsg_dump, -}; + INIT_WORK(work, vmbus_percpu_work); + schedule_work_on(cpu, work); + } -static struct ctl_table_header *hv_ctl_table_hdr; -static int zero; -static int one = 1; + for_each_online_cpu(cpu) + flush_work(per_cpu_ptr(works, cpu)); -/* - * sysctl option to allow the user to control whether kmsg data should be - * reported to Hyper-V on panic. - */ -static struct ctl_table hv_ctl_table[] = { - { - .procname = "hyperv_record_panic_msg", - .data = &sysctl_record_panic_msg, - .maxlen = sizeof(int), - .mode = 0644, - .proc_handler = proc_dointvec_minmax, - .extra1 = &zero, - .extra2 = &one - }, - {} -}; + /* Register the callbacks for possible CPU online/offline'ing */ + ret = cpuhp_setup_state_nocalls_cpuslocked(CPUHP_AP_ONLINE_DYN, "hyperv/vmbus:online", + hv_synic_init, hv_synic_cleanup); + cpus_read_unlock(); + free_percpu(works); + if (ret < 0) + goto err_alloc; + hyperv_cpuhp_online = ret; -static struct ctl_table hv_root_table[] = { - { - .procname = "kernel", - .mode = 0555, - .child = hv_ctl_table - }, - {} -}; + ret = vmbus_connect(); + if (ret) + goto err_connect; + return 0; + +err_connect: + cpuhp_remove_state(hyperv_cpuhp_online); + return -ENODEV; +err_alloc: + hv_synic_free(); + return -ENOMEM; +} /* * vmbus_bus_init -Main vmbus driver initialization routine. @@ -1212,7 +1443,6 @@ static int vmbus_bus_init(void) { int ret; - /* Hypervisor initialization...setup hypercall page..etc */ ret = hv_init(); if (ret != 0) { pr_err("Unable to initialize the hypervisor - 0x%x\n", ret); @@ -1223,79 +1453,66 @@ static int vmbus_bus_init(void) if (ret) return ret; - hv_setup_vmbus_irq(vmbus_isr); - - ret = hv_synic_alloc(); - if (ret) - goto err_alloc; /* - * Initialize the per-cpu interrupt state and - * connect to the host. + * VMbus interrupts are best modeled as per-cpu interrupts. If + * on an architecture with support for per-cpu IRQs (e.g. ARM64), + * allocate a per-cpu IRQ using standard Linux kernel functionality. + * If not on such an architecture (e.g., x86/x64), then rely on + * code in the arch-specific portion of the code tree to connect + * the VMbus interrupt handler. */ - ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "hyperv/vmbus:online", - hv_synic_init, hv_synic_cleanup); - if (ret < 0) - goto err_alloc; - hyperv_cpuhp_online = ret; - ret = vmbus_connect(); + if (vmbus_irq == -1) { + hv_setup_vmbus_handler(vmbus_isr); + } else { + vmbus_evt = alloc_percpu(long); + ret = request_percpu_irq(vmbus_irq, vmbus_percpu_isr, + "Hyper-V VMbus", vmbus_evt); + if (ret) { + pr_err("Can't request Hyper-V VMbus IRQ %d, Err %d", + vmbus_irq, ret); + free_percpu(vmbus_evt); + goto err_setup; + } + } + + /* + * Cache the value as getting it involves a VM exit on x86(_64), and + * doing that on each VP while initializing SynIC's wastes time. + */ + is_confidential = ms_hyperv.confidential_vmbus_available; + if (is_confidential) + pr_info("Establishing connection to the confidential VMBus\n"); + hv_para_set_sint_proxy(!is_confidential); + ret = vmbus_alloc_synic_and_connect(); if (ret) goto err_connect; /* - * Only register if the crash MSRs are available + * Always register the vmbus unload panic notifier because we + * need to shut the VMbus channel connection on panic. */ - if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) { - u64 hyperv_crash_ctl; - /* - * Sysctl registration is not fatal, since by default - * reporting is enabled. - */ - hv_ctl_table_hdr = register_sysctl_table(hv_root_table); - if (!hv_ctl_table_hdr) - pr_err("Hyper-V: sysctl table register error"); - - /* - * Register for panic kmsg callback only if the right - * capability is supported by the hypervisor. - */ - hv_get_crash_ctl(hyperv_crash_ctl); - if (hyperv_crash_ctl & HV_CRASH_CTL_CRASH_NOTIFY_MSG) { - hv_panic_page = (void *)get_zeroed_page(GFP_KERNEL); - if (hv_panic_page) { - ret = kmsg_dump_register(&hv_kmsg_dumper); - if (ret) - pr_err("Hyper-V: kmsg dump register " - "error 0x%x\n", ret); - } else - pr_err("Hyper-V: panic message page memory " - "allocation failed"); - } - - register_die_notifier(&hyperv_die_block); - atomic_notifier_chain_register(&panic_notifier_list, - &hyperv_panic_block); - } + atomic_notifier_chain_register(&panic_notifier_list, + &hyperv_panic_vmbus_unload_block); vmbus_request_offers(); return 0; err_connect: - cpuhp_remove_state(hyperv_cpuhp_online); -err_alloc: - hv_synic_free(); - hv_remove_vmbus_irq(); - + if (vmbus_irq == -1) { + hv_remove_vmbus_handler(); + } else { + free_percpu_irq(vmbus_irq, vmbus_evt); + free_percpu(vmbus_evt); + } +err_setup: bus_unregister(&hv_bus); - free_page((unsigned long)hv_panic_page); - unregister_sysctl_table(hv_ctl_table_hdr); - hv_ctl_table_hdr = NULL; return ret; } /** - * __vmbus_child_driver_register() - Register a vmbus's driver + * __vmbus_driver_register() - Register a vmbus's driver * @hv_driver: Pointer to driver structure you want to register * @owner: owner module of the drv * @mod_name: module name string @@ -1362,7 +1579,7 @@ static void vmbus_chan_release(struct kobject *kobj) struct vmbus_chan_attribute { struct attribute attr; - ssize_t (*show)(const struct vmbus_channel *chan, char *buf); + ssize_t (*show)(struct vmbus_channel *chan, char *buf); ssize_t (*store)(struct vmbus_channel *chan, const char *buf, size_t count); }; @@ -1381,98 +1598,291 @@ static ssize_t vmbus_chan_attr_show(struct kobject *kobj, { const struct vmbus_chan_attribute *attribute = container_of(attr, struct vmbus_chan_attribute, attr); - const struct vmbus_channel *chan + struct vmbus_channel *chan = container_of(kobj, struct vmbus_channel, kobj); if (!attribute->show) return -EIO; - if (chan->state != CHANNEL_OPENED_STATE) - return -EINVAL; - return attribute->show(chan, buf); } +static ssize_t vmbus_chan_attr_store(struct kobject *kobj, + struct attribute *attr, const char *buf, + size_t count) +{ + const struct vmbus_chan_attribute *attribute + = container_of(attr, struct vmbus_chan_attribute, attr); + struct vmbus_channel *chan + = container_of(kobj, struct vmbus_channel, kobj); + + if (!attribute->store) + return -EIO; + + return attribute->store(chan, buf, count); +} + static const struct sysfs_ops vmbus_chan_sysfs_ops = { .show = vmbus_chan_attr_show, + .store = vmbus_chan_attr_store, }; -static ssize_t out_mask_show(const struct vmbus_channel *channel, char *buf) +static ssize_t out_mask_show(struct vmbus_channel *channel, char *buf) { - const struct hv_ring_buffer_info *rbi = &channel->outbound; + struct hv_ring_buffer_info *rbi = &channel->outbound; + ssize_t ret; + + mutex_lock(&rbi->ring_buffer_mutex); + if (!rbi->ring_buffer) { + mutex_unlock(&rbi->ring_buffer_mutex); + return -EINVAL; + } - return sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); + ret = sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); + mutex_unlock(&rbi->ring_buffer_mutex); + return ret; } static VMBUS_CHAN_ATTR_RO(out_mask); -static ssize_t in_mask_show(const struct vmbus_channel *channel, char *buf) +static ssize_t in_mask_show(struct vmbus_channel *channel, char *buf) { - const struct hv_ring_buffer_info *rbi = &channel->inbound; + struct hv_ring_buffer_info *rbi = &channel->inbound; + ssize_t ret; + + mutex_lock(&rbi->ring_buffer_mutex); + if (!rbi->ring_buffer) { + mutex_unlock(&rbi->ring_buffer_mutex); + return -EINVAL; + } - return sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); + ret = sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); + mutex_unlock(&rbi->ring_buffer_mutex); + return ret; } static VMBUS_CHAN_ATTR_RO(in_mask); -static ssize_t read_avail_show(const struct vmbus_channel *channel, char *buf) +static ssize_t read_avail_show(struct vmbus_channel *channel, char *buf) { - const struct hv_ring_buffer_info *rbi = &channel->inbound; + struct hv_ring_buffer_info *rbi = &channel->inbound; + ssize_t ret; - return sprintf(buf, "%u\n", hv_get_bytes_to_read(rbi)); + mutex_lock(&rbi->ring_buffer_mutex); + if (!rbi->ring_buffer) { + mutex_unlock(&rbi->ring_buffer_mutex); + return -EINVAL; + } + + ret = sprintf(buf, "%u\n", hv_get_bytes_to_read(rbi)); + mutex_unlock(&rbi->ring_buffer_mutex); + return ret; } static VMBUS_CHAN_ATTR_RO(read_avail); -static ssize_t write_avail_show(const struct vmbus_channel *channel, char *buf) +static ssize_t write_avail_show(struct vmbus_channel *channel, char *buf) { - const struct hv_ring_buffer_info *rbi = &channel->outbound; + struct hv_ring_buffer_info *rbi = &channel->outbound; + ssize_t ret; - return sprintf(buf, "%u\n", hv_get_bytes_to_write(rbi)); + mutex_lock(&rbi->ring_buffer_mutex); + if (!rbi->ring_buffer) { + mutex_unlock(&rbi->ring_buffer_mutex); + return -EINVAL; + } + + ret = sprintf(buf, "%u\n", hv_get_bytes_to_write(rbi)); + mutex_unlock(&rbi->ring_buffer_mutex); + return ret; } static VMBUS_CHAN_ATTR_RO(write_avail); -static ssize_t show_target_cpu(const struct vmbus_channel *channel, char *buf) +static ssize_t target_cpu_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%u\n", channel->target_cpu); } -static VMBUS_CHAN_ATTR(cpu, S_IRUGO, show_target_cpu, NULL); -static ssize_t channel_pending_show(const struct vmbus_channel *channel, +int vmbus_channel_set_cpu(struct vmbus_channel *channel, u32 target_cpu) +{ + u32 origin_cpu; + int ret = 0; + + lockdep_assert_cpus_held(); + lockdep_assert_held(&vmbus_connection.channel_mutex); + + if (vmbus_proto_version < VERSION_WIN10_V4_1) + return -EIO; + + /* Validate target_cpu for the cpumask_test_cpu() operation below. */ + if (target_cpu >= nr_cpumask_bits) + return -EINVAL; + + if (!cpumask_test_cpu(target_cpu, housekeeping_cpumask(HK_TYPE_MANAGED_IRQ))) + return -EINVAL; + + if (!cpu_online(target_cpu)) + return -EINVAL; + + /* + * Synchronizes vmbus_channel_set_cpu() and channel closure: + * + * { Initially: state = CHANNEL_OPENED } + * + * CPU1 CPU2 + * + * [vmbus_channel_set_cpu()] [vmbus_disconnect_ring()] + * + * LOCK channel_mutex LOCK channel_mutex + * LOAD r1 = state LOAD r2 = state + * IF (r1 == CHANNEL_OPENED) IF (r2 == CHANNEL_OPENED) + * SEND MODIFYCHANNEL STORE state = CHANNEL_OPEN + * [...] SEND CLOSECHANNEL + * UNLOCK channel_mutex UNLOCK channel_mutex + * + * Forbids: r1 == r2 == CHANNEL_OPENED (i.e., CPU1's LOCK precedes + * CPU2's LOCK) && CPU2's SEND precedes CPU1's SEND + * + * Note. The host processes the channel messages "sequentially", in + * the order in which they are received on a per-partition basis. + */ + + /* + * Hyper-V will ignore MODIFYCHANNEL messages for "non-open" channels; + * avoid sending the message and fail here for such channels. + */ + if (channel->state != CHANNEL_OPENED_STATE) { + ret = -EIO; + goto end; + } + + origin_cpu = channel->target_cpu; + if (target_cpu == origin_cpu) + goto end; + + if (vmbus_send_modifychannel(channel, + hv_cpu_number_to_vp_number(target_cpu))) { + ret = -EIO; + goto end; + } + + /* + * For version before VERSION_WIN10_V5_3, the following warning holds: + * + * Warning. At this point, there is *no* guarantee that the host will + * have successfully processed the vmbus_send_modifychannel() request. + * See the header comment of vmbus_send_modifychannel() for more info. + * + * Lags in the processing of the above vmbus_send_modifychannel() can + * result in missed interrupts if the "old" target CPU is taken offline + * before Hyper-V starts sending interrupts to the "new" target CPU. + * But apart from this offlining scenario, the code tolerates such + * lags. It will function correctly even if a channel interrupt comes + * in on a CPU that is different from the channel target_cpu value. + */ + + channel->target_cpu = target_cpu; + + /* See init_vp_index(). */ + if (hv_is_perf_channel(channel)) + hv_update_allocated_cpus(origin_cpu, target_cpu); + + /* Currently set only for storvsc channels. */ + if (channel->change_target_cpu_callback) { + (*channel->change_target_cpu_callback)(channel, + origin_cpu, target_cpu); + } + +end: + return ret; +} + +static ssize_t target_cpu_store(struct vmbus_channel *channel, + const char *buf, size_t count) +{ + u32 target_cpu; + ssize_t ret; + + if (sscanf(buf, "%u", &target_cpu) != 1) + return -EIO; + + cpus_read_lock(); + mutex_lock(&vmbus_connection.channel_mutex); + ret = vmbus_channel_set_cpu(channel, target_cpu); + mutex_unlock(&vmbus_connection.channel_mutex); + cpus_read_unlock(); + + return ret ?: count; +} +static VMBUS_CHAN_ATTR(cpu, 0644, target_cpu_show, target_cpu_store); + +static ssize_t channel_pending_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%d\n", channel_pending(channel, vmbus_connection.monitor_pages[1])); } -static VMBUS_CHAN_ATTR(pending, S_IRUGO, channel_pending_show, NULL); +static VMBUS_CHAN_ATTR(pending, 0444, channel_pending_show, NULL); -static ssize_t channel_latency_show(const struct vmbus_channel *channel, +static ssize_t channel_latency_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%d\n", channel_latency(channel, vmbus_connection.monitor_pages[1])); } -static VMBUS_CHAN_ATTR(latency, S_IRUGO, channel_latency_show, NULL); +static VMBUS_CHAN_ATTR(latency, 0444, channel_latency_show, NULL); -static ssize_t channel_interrupts_show(const struct vmbus_channel *channel, char *buf) +static ssize_t channel_interrupts_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%llu\n", channel->interrupts); } -static VMBUS_CHAN_ATTR(interrupts, S_IRUGO, channel_interrupts_show, NULL); +static VMBUS_CHAN_ATTR(interrupts, 0444, channel_interrupts_show, NULL); -static ssize_t channel_events_show(const struct vmbus_channel *channel, char *buf) +static ssize_t channel_events_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%llu\n", channel->sig_events); } -static VMBUS_CHAN_ATTR(events, S_IRUGO, channel_events_show, NULL); +static VMBUS_CHAN_ATTR(events, 0444, channel_events_show, NULL); -static ssize_t subchannel_monitor_id_show(const struct vmbus_channel *channel, +static ssize_t channel_intr_in_full_show(struct vmbus_channel *channel, + char *buf) +{ + return sprintf(buf, "%llu\n", + (unsigned long long)channel->intr_in_full); +} +static VMBUS_CHAN_ATTR(intr_in_full, 0444, channel_intr_in_full_show, NULL); + +static ssize_t channel_intr_out_empty_show(struct vmbus_channel *channel, + char *buf) +{ + return sprintf(buf, "%llu\n", + (unsigned long long)channel->intr_out_empty); +} +static VMBUS_CHAN_ATTR(intr_out_empty, 0444, channel_intr_out_empty_show, NULL); + +static ssize_t channel_out_full_first_show(struct vmbus_channel *channel, + char *buf) +{ + return sprintf(buf, "%llu\n", + (unsigned long long)channel->out_full_first); +} +static VMBUS_CHAN_ATTR(out_full_first, 0444, channel_out_full_first_show, NULL); + +static ssize_t channel_out_full_total_show(struct vmbus_channel *channel, + char *buf) +{ + return sprintf(buf, "%llu\n", + (unsigned long long)channel->out_full_total); +} +static VMBUS_CHAN_ATTR(out_full_total, 0444, channel_out_full_total_show, NULL); + +static ssize_t subchannel_monitor_id_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%u\n", channel->offermsg.monitorid); } -static VMBUS_CHAN_ATTR(monitor_id, S_IRUGO, subchannel_monitor_id_show, NULL); +static VMBUS_CHAN_ATTR(monitor_id, 0444, subchannel_monitor_id_show, NULL); -static ssize_t subchannel_id_show(const struct vmbus_channel *channel, +static ssize_t subchannel_id_show(struct vmbus_channel *channel, char *buf) { return sprintf(buf, "%u\n", @@ -1480,6 +1890,26 @@ static ssize_t subchannel_id_show(const struct vmbus_channel *channel, } static VMBUS_CHAN_ATTR_RO(subchannel_id); +static int hv_mmap_ring_buffer_wrapper(struct file *filp, struct kobject *kobj, + const struct bin_attribute *attr, + struct vm_area_struct *vma) +{ + struct vmbus_channel *channel = container_of(kobj, struct vmbus_channel, kobj); + + /* + * hv_(create|remove)_ring_sysfs implementation ensures that mmap_ring_buffer + * is not NULL. + */ + return channel->mmap_ring_buffer(channel, vma); +} + +static struct bin_attribute chan_attr_ring_buffer = { + .attr = { + .name = "ring", + .mode = 0600, + }, + .mmap = hv_mmap_ring_buffer_wrapper, +}; static struct attribute *vmbus_chan_attrs[] = { &chan_attr_out_mask.attr, &chan_attr_in_mask.attr, @@ -1490,22 +1920,139 @@ static struct attribute *vmbus_chan_attrs[] = { &chan_attr_latency.attr, &chan_attr_interrupts.attr, &chan_attr_events.attr, + &chan_attr_intr_in_full.attr, + &chan_attr_intr_out_empty.attr, + &chan_attr_out_full_first.attr, + &chan_attr_out_full_total.attr, &chan_attr_monitor_id.attr, &chan_attr_subchannel_id.attr, NULL }; -static struct kobj_type vmbus_chan_ktype = { +static const struct bin_attribute *vmbus_chan_bin_attrs[] = { + &chan_attr_ring_buffer, + NULL +}; + +/* + * Channel-level attribute_group callback function. Returns the permission for + * each attribute, and returns 0 if an attribute is not visible. + */ +static umode_t vmbus_chan_attr_is_visible(struct kobject *kobj, + struct attribute *attr, int idx) +{ + const struct vmbus_channel *channel = + container_of(kobj, struct vmbus_channel, kobj); + + /* Hide the monitor attributes if the monitor mechanism is not used. */ + if (!channel->offermsg.monitor_allocated && + (attr == &chan_attr_pending.attr || + attr == &chan_attr_latency.attr || + attr == &chan_attr_monitor_id.attr)) + return 0; + + return attr->mode; +} + +static umode_t vmbus_chan_bin_attr_is_visible(struct kobject *kobj, + const struct bin_attribute *attr, int idx) +{ + const struct vmbus_channel *channel = + container_of(kobj, struct vmbus_channel, kobj); + + /* Hide ring attribute if channel's ring_sysfs_visible is set to false */ + if (attr == &chan_attr_ring_buffer && !channel->ring_sysfs_visible) + return 0; + + return attr->attr.mode; +} + +static size_t vmbus_chan_bin_size(struct kobject *kobj, + const struct bin_attribute *bin_attr, int a) +{ + const struct vmbus_channel *channel = + container_of(kobj, struct vmbus_channel, kobj); + + return channel->ringbuffer_pagecount << PAGE_SHIFT; +} + +static const struct attribute_group vmbus_chan_group = { + .attrs = vmbus_chan_attrs, + .bin_attrs = vmbus_chan_bin_attrs, + .is_visible = vmbus_chan_attr_is_visible, + .is_bin_visible = vmbus_chan_bin_attr_is_visible, + .bin_size = vmbus_chan_bin_size, +}; + +static const struct kobj_type vmbus_chan_ktype = { .sysfs_ops = &vmbus_chan_sysfs_ops, .release = vmbus_chan_release, - .default_attrs = vmbus_chan_attrs, }; +/** + * hv_create_ring_sysfs() - create "ring" sysfs entry corresponding to ring buffers for a channel. + * @channel: Pointer to vmbus_channel structure + * @hv_mmap_ring_buffer: function pointer for initializing the function to be called on mmap of + * channel's "ring" sysfs node, which is for the ring buffer of that channel. + * Function pointer is of below type: + * int (*hv_mmap_ring_buffer)(struct vmbus_channel *channel, + * struct vm_area_struct *vma)) + * This has a pointer to the channel and a pointer to vm_area_struct, + * used for mmap, as arguments. + * + * Sysfs node for ring buffer of a channel is created along with other fields, however its + * visibility is disabled by default. Sysfs creation needs to be controlled when the use-case + * is running. + * For example, HV_NIC device is used either by uio_hv_generic or hv_netvsc at any given point of + * time, and "ring" sysfs is needed only when uio_hv_generic is bound to that device. To avoid + * exposing the ring buffer by default, this function is responsible to enable visibility of + * ring for userspace to use. + * Note: Race conditions can happen with userspace and it is not encouraged to create new + * use-cases for this. This was added to maintain backward compatibility, while solving + * one of the race conditions in uio_hv_generic while creating sysfs. See comments with + * vmbus_add_dynid() and vmbus_device_register(). + * + * Returns 0 on success or error code on failure. + */ +int hv_create_ring_sysfs(struct vmbus_channel *channel, + int (*hv_mmap_ring_buffer)(struct vmbus_channel *channel, + struct vm_area_struct *vma)) +{ + struct kobject *kobj = &channel->kobj; + + channel->mmap_ring_buffer = hv_mmap_ring_buffer; + channel->ring_sysfs_visible = true; + + return sysfs_update_group(kobj, &vmbus_chan_group); +} +EXPORT_SYMBOL_GPL(hv_create_ring_sysfs); + +/** + * hv_remove_ring_sysfs() - remove ring sysfs entry corresponding to ring buffers for a channel. + * @channel: Pointer to vmbus_channel structure + * + * Hide "ring" sysfs for a channel by changing its is_visible attribute and updating sysfs group. + * + * Returns 0 on success or error code on failure. + */ +int hv_remove_ring_sysfs(struct vmbus_channel *channel) +{ + struct kobject *kobj = &channel->kobj; + int ret; + + channel->ring_sysfs_visible = false; + ret = sysfs_update_group(kobj, &vmbus_chan_group); + channel->mmap_ring_buffer = NULL; + return ret; +} +EXPORT_SYMBOL_GPL(hv_remove_ring_sysfs); + /* * vmbus_add_channel_kobj - setup a sub-directory under device/channels */ int vmbus_add_channel_kobj(struct hv_device *dev, struct vmbus_channel *channel) { + const struct device *device = &dev->device; struct kobject *kobj = &channel->kobj; u32 relid = channel->offermsg.child_relid; int ret; @@ -1513,8 +2060,22 @@ int vmbus_add_channel_kobj(struct hv_device *dev, struct vmbus_channel *channel) kobj->kset = dev->channels_kset; ret = kobject_init_and_add(kobj, &vmbus_chan_ktype, NULL, "%u", relid); - if (ret) + if (ret) { + kobject_put(kobj); + return ret; + } + + ret = sysfs_create_group(kobj, &vmbus_chan_group); + + if (ret) { + /* + * The calling functions' error handling paths will cleanup the + * empty channel directory. + */ + kobject_put(kobj); + dev_err(device, "Unable to set up channel sysfs files\n"); return ret; + } kobject_uevent(kobj, KOBJ_ADD); @@ -1522,11 +2083,19 @@ int vmbus_add_channel_kobj(struct hv_device *dev, struct vmbus_channel *channel) } /* + * vmbus_remove_channel_attr_group - remove the channel's attribute group + */ +void vmbus_remove_channel_attr_group(struct vmbus_channel *channel) +{ + sysfs_remove_group(&channel->kobj, &vmbus_chan_group); +} + +/* * vmbus_device_create - Creates and registers a new child device * on the vmbus. */ -struct hv_device *vmbus_device_create(const uuid_le *type, - const uuid_le *instance, +struct hv_device *vmbus_device_create(const guid_t *type, + const guid_t *instance, struct vmbus_channel *channel) { struct hv_device *child_device_obj; @@ -1538,11 +2107,9 @@ struct hv_device *vmbus_device_create(const uuid_le *type, } child_device_obj->channel = channel; - memcpy(&child_device_obj->dev_type, type, sizeof(uuid_le)); - memcpy(&child_device_obj->dev_instance, instance, - sizeof(uuid_le)); - child_device_obj->vendor_id = 0x1414; /* MSFT vendor ID */ - + guid_copy(&child_device_obj->dev_type, type); + guid_copy(&child_device_obj->dev_instance, instance); + child_device_obj->vendor_id = PCI_VENDOR_ID_MICROSOFT; return child_device_obj; } @@ -1556,12 +2123,16 @@ int vmbus_device_register(struct hv_device *child_device_obj) int ret; dev_set_name(&child_device_obj->device, "%pUl", - child_device_obj->channel->offermsg.offer.if_instance.b); + &child_device_obj->channel->offermsg.offer.if_instance); child_device_obj->device.bus = &hv_bus; - child_device_obj->device.parent = &hv_acpi_dev->dev; + child_device_obj->device.parent = vmbus_root_device; child_device_obj->device.release = vmbus_device_release; + child_device_obj->device.dma_parms = &child_device_obj->dma_parms; + child_device_obj->device.dma_mask = &child_device_obj->dma_mask; + dma_set_mask(&child_device_obj->device, DMA_BIT_MASK(64)); + /* * Register with the LDM. This will kick off the driver/device * binding...which will eventually call vmbus_match() and vmbus_probe() @@ -1569,9 +2140,24 @@ int vmbus_device_register(struct hv_device *child_device_obj) ret = device_register(&child_device_obj->device); if (ret) { pr_err("Unable to register child device\n"); + put_device(&child_device_obj->device); return ret; } + /* + * If device_register() found a driver to assign to the device, the + * driver's probe function has already run at this point. If that + * probe function accesses or operates on the "channels" subdirectory + * in sysfs, those operations will have failed because the "channels" + * subdirectory doesn't exist until the code below runs. Or if the + * probe function creates a /dev entry, a user space program could + * find and open the /dev entry, and then create a race by accessing + * the "channels" subdirectory while the creation steps are in progress + * here. The race can't result in a kernel failure, but the user space + * program may get an error in accessing "channels" or its + * subdirectories. See also comments with vmbus_add_dynid() about a + * related race condition. + */ child_device_obj->channels_kset = kset_create_and_add("channels", NULL, kobj); if (!child_device_obj->channels_kset) { @@ -1582,9 +2168,10 @@ int vmbus_device_register(struct hv_device *child_device_obj) ret = vmbus_add_channel_kobj(child_device_obj, child_device_obj->channel); if (ret) { - pr_err("Unable to register primary channeln"); + pr_err("Unable to register primary channel\n"); goto err_kset_unregister; } + hv_debug_add_dev_dir(child_device_obj); return 0; @@ -1613,13 +2200,13 @@ void vmbus_device_unregister(struct hv_device *device_obj) */ device_unregister(&device_obj->device); } +EXPORT_SYMBOL_GPL(vmbus_device_unregister); - +#ifdef CONFIG_ACPI /* * VMBUS is an acpi enumerated device. Get the information we * need from DSDT. */ -#define VTPM_BASE_ADDRESS 0xfed40000 static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx) { resource_size_t start = 0; @@ -1627,6 +2214,7 @@ static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx) struct resource *new_res; struct resource **old_res = &hyperv_mmio; struct resource **prev_res = NULL; + struct resource r; switch (res->type) { @@ -1645,6 +2233,23 @@ static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx) end = res->data.address64.address.maximum; break; + /* + * The IRQ information is needed only on ARM64, which Hyper-V + * sets up in the extended format. IRQ information is present + * on x86/x64 in the non-extended format but it is not used by + * Linux. So don't bother checking for the non-extended format. + */ + case ACPI_RESOURCE_TYPE_EXTENDED_IRQ: + if (!acpi_dev_resource_interrupt(res, 0, &r)) { + pr_err("Unable to parse Hyper-V ACPI interrupt\n"); + return AE_ERROR; + } + /* ARM64 INTID for VMbus */ + vmbus_interrupt = res->data.extended_irq.interrupts[0]; + /* Linux IRQ number */ + vmbus_irq = r.start; + return AE_OK; + default: /* Unused resource type */ return AE_OK; @@ -1706,8 +2311,9 @@ static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx) return AE_OK; } +#endif -static int vmbus_acpi_remove(struct acpi_device *device) +static void vmbus_mmio_remove(void) { struct resource *cur_res; struct resource *next_res; @@ -1724,32 +2330,49 @@ static int vmbus_acpi_remove(struct acpi_device *device) kfree(cur_res); } } - - return 0; } -static void vmbus_reserve_fb(void) +static void __maybe_unused vmbus_reserve_fb(void) { - int size; + resource_size_t start = 0, size; + struct pci_dev *pdev; + + if (efi_enabled(EFI_BOOT)) { + /* Gen2 VM: get FB base from EFI framebuffer */ + if (IS_ENABLED(CONFIG_SYSFB)) { + start = screen_info.lfb_base; + size = max_t(__u32, screen_info.lfb_size, 0x800000); + } + } else { + /* Gen1 VM: get FB base from PCI */ + pdev = pci_get_device(PCI_VENDOR_ID_MICROSOFT, + PCI_DEVICE_ID_HYPERV_VIDEO, NULL); + if (!pdev) + return; + + if (pdev->resource[0].flags & IORESOURCE_MEM) { + start = pci_resource_start(pdev, 0); + size = pci_resource_len(pdev, 0); + } + + /* + * Release the PCI device so hyperv_drm or hyperv_fb driver can + * grab it later. + */ + pci_dev_put(pdev); + } + + if (!start) + return; + /* * Make a claim for the frame buffer in the resource tree under the * first node, which will be the one below 4GB. The length seems to * be underreported, particularly in a Generation 1 VM. So start out * reserving a larger area and make it smaller until it succeeds. */ - - if (screen_info.lfb_base) { - if (efi_enabled(EFI_BOOT)) - size = max_t(__u32, screen_info.lfb_size, 0x800000); - else - size = max_t(__u32, screen_info.lfb_size, 0x4000000); - - for (; !fb_mmio && (size >= 0x100000); size >>= 1) { - fb_mmio = __request_region(hyperv_mmio, - screen_info.lfb_base, size, - fb_mmio_name, 0); - } - } + for (; !fb_mmio && (size >= 0x100000); size >>= 1) + fb_mmio = __request_region(hyperv_mmio, start, size, fb_mmio_name, 0); } /** @@ -1781,12 +2404,12 @@ int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj, bool fb_overlap_ok) { struct resource *iter, *shadow; - resource_size_t range_min, range_max, start; + resource_size_t range_min, range_max, start, end; const char *dev_n = dev_name(&device_obj->device); int retval; retval = -ENXIO; - down(&hyperv_mmio_lock); + mutex_lock(&hyperv_mmio_lock); /* * If overlaps with frame buffers are allowed, then first attempt to @@ -1816,6 +2439,14 @@ int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj, range_max = iter->end; start = (range_min + align - 1) & ~(align - 1); for (; start + size - 1 <= range_max; start += align) { + end = start + size - 1; + + /* Skip the whole fb_mmio region if not fb_overlap_ok */ + if (!fb_overlap_ok && fb_mmio && + (((start >= fb_mmio->start) && (start <= fb_mmio->end)) || + ((end >= fb_mmio->start) && (end <= fb_mmio->end)))) + continue; + shadow = __request_region(iter, start, size, NULL, IORESOURCE_BUSY); if (!shadow) @@ -1833,7 +2464,7 @@ int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj, } exit: - up(&hyperv_mmio_lock); + mutex_unlock(&hyperv_mmio_lock); return retval; } EXPORT_SYMBOL_GPL(vmbus_allocate_mmio); @@ -1850,26 +2481,56 @@ void vmbus_free_mmio(resource_size_t start, resource_size_t size) { struct resource *iter; - down(&hyperv_mmio_lock); + mutex_lock(&hyperv_mmio_lock); + + /* + * If all bytes of the MMIO range to be released are within the + * special case fb_mmio shadow region, skip releasing the shadow + * region since no corresponding __request_region() was done + * in vmbus_allocate_mmio(). + */ + if (fb_mmio && start >= fb_mmio->start && + (start + size - 1 <= fb_mmio->end)) + goto skip_shadow_release; + for (iter = hyperv_mmio; iter; iter = iter->sibling) { if ((iter->start >= start + size) || (iter->end <= start)) continue; __release_region(iter, start, size); } + +skip_shadow_release: release_mem_region(start, size); - up(&hyperv_mmio_lock); + mutex_unlock(&hyperv_mmio_lock); } EXPORT_SYMBOL_GPL(vmbus_free_mmio); -static int vmbus_acpi_add(struct acpi_device *device) +#ifdef CONFIG_ACPI +static int vmbus_acpi_add(struct platform_device *pdev) { acpi_status result; int ret_val = -ENODEV; struct acpi_device *ancestor; + struct acpi_device *device = ACPI_COMPANION(&pdev->dev); - hv_acpi_dev = device; + vmbus_root_device = &device->dev; + + /* + * Older versions of Hyper-V for ARM64 fail to include the _CCA + * method on the top level VMbus device in the DSDT. But devices + * are hardware coherent in all current Hyper-V use cases, so fix + * up the ACPI device to behave as if _CCA is present and indicates + * hardware coherence. + */ + ACPI_COMPANION_SET(&device->dev, device); + if (IS_ENABLED(CONFIG_ACPI_CCA_REQUIRED) && + device_get_dma_attr(&device->dev) == DEV_DMA_NOT_SUPPORTED) { + pr_info("No ACPI _CCA found; assuming coherent device I/O\n"); + device->flags.cca_seen = true; + device->flags.coherent_dma = true; + } result = acpi_walk_resources(device->handle, METHOD_NAME__CRS, vmbus_walk_resources, NULL); @@ -1880,7 +2541,9 @@ static int vmbus_acpi_add(struct acpi_device *device) * Some ancestor of the vmbus acpi device (Gen1 or Gen2 * firmware) is the VMOD that has the mmio ranges. Get that. */ - for (ancestor = device->parent; ancestor; ancestor = ancestor->parent) { + for (ancestor = acpi_dev_parent(device); + ancestor && ancestor->handle != ACPI_ROOT_OBJECT; + ancestor = acpi_dev_parent(ancestor)) { result = acpi_walk_resources(ancestor->handle, METHOD_NAME__CRS, vmbus_walk_resources, NULL); @@ -1894,75 +2557,391 @@ static int vmbus_acpi_add(struct acpi_device *device) ret_val = 0; acpi_walk_err: - complete(&probe_event); if (ret_val) - vmbus_acpi_remove(device); + vmbus_mmio_remove(); return ret_val; } +#else +static int vmbus_acpi_add(struct platform_device *pdev) +{ + return 0; +} +#endif +#ifndef HYPERVISOR_CALLBACK_VECTOR +static int vmbus_set_irq(struct platform_device *pdev) +{ + struct irq_data *data; + int irq; + irq_hw_number_t hwirq; + + irq = platform_get_irq(pdev, 0); + /* platform_get_irq() may not return 0. */ + if (irq < 0) + return irq; + + data = irq_get_irq_data(irq); + if (!data) { + pr_err("No interrupt data for VMBus virq %d\n", irq); + return -ENODEV; + } + hwirq = irqd_to_hwirq(data); + + vmbus_irq = irq; + vmbus_interrupt = hwirq; + pr_debug("VMBus virq %d, hwirq %d\n", vmbus_irq, vmbus_interrupt); + + return 0; +} +#endif + +static int vmbus_device_add(struct platform_device *pdev) +{ + struct resource **cur_res = &hyperv_mmio; + struct of_range range; + struct of_range_parser parser; + struct device_node *np = pdev->dev.of_node; + int ret; + + vmbus_root_device = &pdev->dev; + + ret = of_range_parser_init(&parser, np); + if (ret) + return ret; + +#ifndef HYPERVISOR_CALLBACK_VECTOR + ret = vmbus_set_irq(pdev); + if (ret) + return ret; +#endif + for_each_of_range(&parser, &range) { + struct resource *res; + + res = kzalloc(sizeof(*res), GFP_KERNEL); + if (!res) { + vmbus_mmio_remove(); + return -ENOMEM; + } + + res->name = "hyperv mmio"; + res->flags = range.flags; + res->start = range.cpu_addr; + res->end = range.cpu_addr + range.size; + + *cur_res = res; + cur_res = &res->sibling; + } + + return ret; +} + +static int vmbus_platform_driver_probe(struct platform_device *pdev) +{ + if (acpi_disabled) + return vmbus_device_add(pdev); + else + return vmbus_acpi_add(pdev); +} + +static void vmbus_platform_driver_remove(struct platform_device *pdev) +{ + vmbus_mmio_remove(); +} + +#ifdef CONFIG_PM_SLEEP +static int vmbus_bus_suspend(struct device *dev) +{ + struct hv_per_cpu_context *hv_cpu = per_cpu_ptr( + hv_context.cpu_context, VMBUS_CONNECT_CPU); + struct vmbus_channel *channel, *sc; + + tasklet_disable(&hv_cpu->msg_dpc); + vmbus_connection.ignore_any_offer_msg = true; + /* The tasklet_enable() takes care of providing a memory barrier */ + tasklet_enable(&hv_cpu->msg_dpc); + + /* Drain all the workqueues as we are in suspend */ + drain_workqueue(vmbus_connection.rescind_work_queue); + drain_workqueue(vmbus_connection.work_queue); + drain_workqueue(vmbus_connection.handle_primary_chan_wq); + drain_workqueue(vmbus_connection.handle_sub_chan_wq); + + mutex_lock(&vmbus_connection.channel_mutex); + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + if (!is_hvsock_channel(channel)) + continue; + + vmbus_force_channel_rescinded(channel); + } + mutex_unlock(&vmbus_connection.channel_mutex); + + /* + * Wait until all the sub-channels and hv_sock channels have been + * cleaned up. Sub-channels should be destroyed upon suspend, otherwise + * they would conflict with the new sub-channels that will be created + * in the resume path. hv_sock channels should also be destroyed, but + * a hv_sock channel of an established hv_sock connection can not be + * really destroyed since it may still be referenced by the userspace + * application, so we just force the hv_sock channel to be rescinded + * by vmbus_force_channel_rescinded(), and the userspace application + * will thoroughly destroy the channel after hibernation. + * + * Note: the counter nr_chan_close_on_suspend may never go above 0 if + * the VM has no sub-channel and hv_sock channel, e.g. a 1-vCPU VM. + */ + if (atomic_read(&vmbus_connection.nr_chan_close_on_suspend) > 0) + wait_for_completion(&vmbus_connection.ready_for_suspend_event); + + mutex_lock(&vmbus_connection.channel_mutex); + + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + /* + * Remove the channel from the array of channels and invalidate + * the channel's relid. Upon resume, vmbus_onoffer() will fix + * up the relid (and other fields, if necessary) and add the + * channel back to the array. + */ + vmbus_channel_unmap_relid(channel); + channel->offermsg.child_relid = INVALID_RELID; + + if (is_hvsock_channel(channel)) { + if (!channel->rescind) { + pr_err("hv_sock channel not rescinded!\n"); + WARN_ON_ONCE(1); + } + continue; + } + + list_for_each_entry(sc, &channel->sc_list, sc_list) { + pr_err("Sub-channel not deleted!\n"); + WARN_ON_ONCE(1); + } + } + + mutex_unlock(&vmbus_connection.channel_mutex); + + vmbus_initiate_unload(false); + + return 0; +} + +static int vmbus_bus_resume(struct device *dev) +{ + struct vmbus_channel *channel; + struct vmbus_channel_msginfo *msginfo; + size_t msgsize; + int ret; + + vmbus_connection.ignore_any_offer_msg = false; + + /* + * We only use the 'vmbus_proto_version', which was in use before + * hibernation, to re-negotiate with the host. + */ + if (!vmbus_proto_version) { + pr_err("Invalid proto version = 0x%x\n", vmbus_proto_version); + return -EINVAL; + } + + msgsize = sizeof(*msginfo) + + sizeof(struct vmbus_channel_initiate_contact); + + msginfo = kzalloc(msgsize, GFP_KERNEL); + + if (msginfo == NULL) + return -ENOMEM; + + ret = vmbus_negotiate_version(msginfo, vmbus_proto_version); + + kfree(msginfo); + + if (ret != 0) + return ret; + + vmbus_request_offers(); -static const struct acpi_device_id vmbus_acpi_device_ids[] = { + mutex_lock(&vmbus_connection.channel_mutex); + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + if (channel->offermsg.child_relid != INVALID_RELID) + continue; + + /* hvsock channels are not expected to be present. */ + if (is_hvsock_channel(channel)) + continue; + + pr_err("channel %pUl/%pUl not present after resume.\n", + &channel->offermsg.offer.if_type, + &channel->offermsg.offer.if_instance); + /* ToDo: Cleanup these channels here */ + } + mutex_unlock(&vmbus_connection.channel_mutex); + + /* Reset the event for the next suspend. */ + reinit_completion(&vmbus_connection.ready_for_suspend_event); + + return 0; +} +#else +#define vmbus_bus_suspend NULL +#define vmbus_bus_resume NULL +#endif /* CONFIG_PM_SLEEP */ + +static const __maybe_unused struct of_device_id vmbus_of_match[] = { + { + .compatible = "microsoft,vmbus", + }, + { + /* sentinel */ + }, +}; +MODULE_DEVICE_TABLE(of, vmbus_of_match); + +static const __maybe_unused struct acpi_device_id vmbus_acpi_device_ids[] = { {"VMBUS", 0}, {"VMBus", 0}, {"", 0}, }; MODULE_DEVICE_TABLE(acpi, vmbus_acpi_device_ids); -static struct acpi_driver vmbus_acpi_driver = { - .name = "vmbus", - .ids = vmbus_acpi_device_ids, - .ops = { - .add = vmbus_acpi_add, - .remove = vmbus_acpi_remove, - }, +/* + * Note: we must use the "no_irq" ops, otherwise hibernation can not work with + * PCI device assignment, because "pci_dev_pm_ops" uses the "noirq" ops: in + * the resume path, the pci "noirq" restore op runs before "non-noirq" op (see + * resume_target_kernel() -> dpm_resume_start(), and hibernation_restore() -> + * dpm_resume_end()). This means vmbus_bus_resume() and the pci-hyperv's + * resume callback must also run via the "noirq" ops. + * + * Set suspend_noirq/resume_noirq to NULL for Suspend-to-Idle: see the comment + * earlier in this file before vmbus_pm. + */ + +static const struct dev_pm_ops vmbus_bus_pm = { + .suspend_noirq = NULL, + .resume_noirq = NULL, + .freeze_noirq = vmbus_bus_suspend, + .thaw_noirq = vmbus_bus_resume, + .poweroff_noirq = vmbus_bus_suspend, + .restore_noirq = vmbus_bus_resume +}; + +static struct platform_driver vmbus_platform_driver = { + .probe = vmbus_platform_driver_probe, + .remove = vmbus_platform_driver_remove, + .driver = { + .name = "vmbus", + .acpi_match_table = ACPI_PTR(vmbus_acpi_device_ids), + .of_match_table = of_match_ptr(vmbus_of_match), + .pm = &vmbus_bus_pm, + .probe_type = PROBE_FORCE_SYNCHRONOUS, + } }; static void hv_kexec_handler(void) { - hv_synic_clockevents_cleanup(); + hv_stimer_global_cleanup(); vmbus_initiate_unload(false); - vmbus_connection.conn_state = DISCONNECTED; /* Make sure conn_state is set as hv_synic_cleanup checks for it */ mb(); cpuhp_remove_state(hyperv_cpuhp_online); - hyperv_cleanup(); }; static void hv_crash_handler(struct pt_regs *regs) { + int cpu; + vmbus_initiate_unload(true); /* * In crash handler we can't schedule synic cleanup for all CPUs, * doing the cleanup for current CPU only. This should be sufficient * for kdump. */ - vmbus_connection.conn_state = DISCONNECTED; - hv_synic_cleanup(smp_processor_id()); - hyperv_cleanup(); + cpu = smp_processor_id(); + hv_stimer_cleanup(cpu); + hv_hyp_synic_disable_regs(cpu); +}; + +static int hv_synic_suspend(void *data) +{ + /* + * When we reach here, all the non-boot CPUs have been offlined. + * If we're in a legacy configuration where stimer Direct Mode is + * not enabled, the stimers on the non-boot CPUs have been unbound + * in hv_synic_cleanup() -> hv_stimer_legacy_cleanup() -> + * hv_stimer_cleanup() -> clockevents_unbind_device(). + * + * hv_synic_suspend() only runs on CPU0 with interrupts disabled. + * Here we do not call hv_stimer_legacy_cleanup() on CPU0 because: + * 1) it's unnecessary as interrupts remain disabled between + * syscore_suspend() and syscore_resume(): see create_image() and + * resume_target_kernel() + * 2) the stimer on CPU0 is automatically disabled later by + * syscore_suspend() -> timekeeping_suspend() -> tick_suspend() -> ... + * -> clockevents_shutdown() -> ... -> hv_ce_shutdown() + * 3) a warning would be triggered if we call + * clockevents_unbind_device(), which may sleep, in an + * interrupts-disabled context. + */ + + hv_hyp_synic_disable_regs(0); + + return 0; +} + +static void hv_synic_resume(void *data) +{ + hv_hyp_synic_enable_regs(0); + + /* + * Note: we don't need to call hv_stimer_init(0), because the timer + * on CPU0 is not unbound in hv_synic_suspend(), and the timer is + * automatically re-enabled in timekeeping_resume(). + */ +} + +/* The callbacks run only on CPU0, with irqs_disabled. */ +static const struct syscore_ops hv_synic_syscore_ops = { + .suspend = hv_synic_suspend, + .resume = hv_synic_resume, +}; + +static struct syscore hv_synic_syscore = { + .ops = &hv_synic_syscore_ops, }; static int __init hv_acpi_init(void) { - int ret, t; + int ret; if (!hv_is_hyperv_initialized()) return -ENODEV; - init_completion(&probe_event); + if (hv_root_partition() && !hv_nested) + return 0; /* * Get ACPI resources first. */ - ret = acpi_bus_register_driver(&vmbus_acpi_driver); - + ret = platform_driver_register(&vmbus_platform_driver); if (ret) return ret; - t = wait_for_completion_timeout(&probe_event, 5*HZ); - if (t == 0) { - ret = -ETIMEDOUT; + if (!vmbus_root_device) { + ret = -ENODEV; goto cleanup; } + /* + * If we're on an architecture with a hardcoded hypervisor + * vector (i.e. x86/x64), override the VMbus interrupt found + * in the ACPI tables. Ensure vmbus_irq is not set since the + * normal Linux IRQ mechanism is not used in this case. + */ +#ifdef HYPERVISOR_CALLBACK_VECTOR + vmbus_interrupt = HYPERVISOR_CALLBACK_VECTOR; + vmbus_irq = -1; +#endif + + hv_debug_init(); + ret = vmbus_bus_init(); if (ret) goto cleanup; @@ -1970,11 +2949,13 @@ static int __init hv_acpi_init(void) hv_setup_kexec_handler(hv_kexec_handler); hv_setup_crash_handler(hv_crash_handler); + register_syscore(&hv_synic_syscore); + return 0; cleanup: - acpi_bus_unregister_driver(&vmbus_acpi_driver); - hv_acpi_dev = NULL; + platform_driver_unregister(&vmbus_platform_driver); + vmbus_root_device = NULL; return ret; } @@ -1982,39 +2963,47 @@ static void __exit vmbus_exit(void) { int cpu; + unregister_syscore(&hv_synic_syscore); + hv_remove_kexec_handler(); hv_remove_crash_handler(); vmbus_connection.conn_state = DISCONNECTED; - hv_synic_clockevents_cleanup(); + hv_stimer_global_cleanup(); vmbus_disconnect(); - hv_remove_vmbus_irq(); + if (vmbus_irq == -1) { + hv_remove_vmbus_handler(); + } else { + free_percpu_irq(vmbus_irq, vmbus_evt); + free_percpu(vmbus_evt); + } for_each_online_cpu(cpu) { struct hv_per_cpu_context *hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu); tasklet_kill(&hv_cpu->msg_dpc); } + hv_debug_rm_all_dir(); + vmbus_free_channels(); + kfree(vmbus_connection.channels); - if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) { - kmsg_dump_unregister(&hv_kmsg_dumper); - unregister_die_notifier(&hyperv_die_block); - atomic_notifier_chain_unregister(&panic_notifier_list, - &hyperv_panic_block); - } + /* + * The vmbus panic notifier is always registered, hence we should + * also unconditionally unregister it here as well. + */ + atomic_notifier_chain_unregister(&panic_notifier_list, + &hyperv_panic_vmbus_unload_block); - free_page((unsigned long)hv_panic_page); - unregister_sysctl_table(hv_ctl_table_hdr); - hv_ctl_table_hdr = NULL; bus_unregister(&hv_bus); cpuhp_remove_state(hyperv_cpuhp_online); hv_synic_free(); - acpi_bus_unregister_driver(&vmbus_acpi_driver); + platform_driver_unregister(&vmbus_platform_driver); } MODULE_LICENSE("GPL"); +MODULE_DESCRIPTION("Microsoft Hyper-V VMBus Driver"); subsys_initcall(hv_acpi_init); module_exit(vmbus_exit); |