diff options
Diffstat (limited to 'drivers/hv')
| -rw-r--r-- | drivers/hv/Kconfig | 28 | ||||
| -rw-r--r-- | drivers/hv/Makefile | 12 | ||||
| -rw-r--r-- | drivers/hv/channel.c | 900 | ||||
| -rw-r--r-- | drivers/hv/channel_mgmt.c | 1425 | ||||
| -rw-r--r-- | drivers/hv/connection.c | 480 | ||||
| -rw-r--r-- | drivers/hv/hv.c | 442 | ||||
| -rw-r--r-- | drivers/hv/hv_balloon.c | 1777 | ||||
| -rw-r--r-- | drivers/hv/hv_fcopy.c | 363 | ||||
| -rw-r--r-- | drivers/hv/hv_kvp.c | 767 | ||||
| -rw-r--r-- | drivers/hv/hv_snapshot.c | 398 | ||||
| -rw-r--r-- | drivers/hv/hv_trace.c | 6 | ||||
| -rw-r--r-- | drivers/hv/hv_trace.h | 329 | ||||
| -rw-r--r-- | drivers/hv/hv_trace_balloon.h | 48 | ||||
| -rw-r--r-- | drivers/hv/hv_util.c | 584 | ||||
| -rw-r--r-- | drivers/hv/hv_utils_transport.c | 360 | ||||
| -rw-r--r-- | drivers/hv/hv_utils_transport.h | 55 | ||||
| -rw-r--r-- | drivers/hv/hyperv_vmbus.h | 446 | ||||
| -rw-r--r-- | drivers/hv/ring_buffer.c | 543 | ||||
| -rw-r--r-- | drivers/hv/vmbus_drv.c | 2088 |
19 files changed, 11051 insertions, 0 deletions
diff --git a/drivers/hv/Kconfig b/drivers/hv/Kconfig new file mode 100644 index 000000000..1c1a2514d --- /dev/null +++ b/drivers/hv/Kconfig @@ -0,0 +1,28 @@ +# SPDX-License-Identifier: GPL-2.0 + +menu "Microsoft Hyper-V guest support" + +config HYPERV + tristate "Microsoft Hyper-V client drivers" + depends on X86 && ACPI && X86_LOCAL_APIC && HYPERVISOR_GUEST + select PARAVIRT + help + Select this option to run Linux as a Hyper-V client operating + system. + +config HYPERV_TSCPAGE + def_bool HYPERV && X86_64 + +config HYPERV_UTILS + tristate "Microsoft Hyper-V Utilities driver" + depends on HYPERV && CONNECTOR && NLS + help + Select this option to enable the Hyper-V Utilities. + +config HYPERV_BALLOON + tristate "Microsoft Hyper-V Balloon driver" + depends on HYPERV + help + Select this option to enable Hyper-V Balloon driver. + +endmenu diff --git a/drivers/hv/Makefile b/drivers/hv/Makefile new file mode 100644 index 000000000..a1eec7177 --- /dev/null +++ b/drivers/hv/Makefile @@ -0,0 +1,12 @@ +# SPDX-License-Identifier: GPL-2.0 +obj-$(CONFIG_HYPERV) += hv_vmbus.o +obj-$(CONFIG_HYPERV_UTILS) += hv_utils.o +obj-$(CONFIG_HYPERV_BALLOON) += hv_balloon.o + +CFLAGS_hv_trace.o = -I$(src) +CFLAGS_hv_balloon.o = -I$(src) + +hv_vmbus-y := vmbus_drv.o \ + hv.o connection.o channel.o \ + channel_mgmt.o ring_buffer.o hv_trace.o +hv_utils-y := hv_util.o hv_kvp.o hv_snapshot.o hv_fcopy.o hv_utils_transport.o diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c new file mode 100644 index 000000000..5e515533e --- /dev/null +++ b/drivers/hv/channel.c @@ -0,0 +1,900 @@ +/* + * 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> + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/wait.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/module.h> +#include <linux/hyperv.h> +#include <linux/uio.h> +#include <linux/interrupt.h> +#include <asm/page.h> + +#include "hyperv_vmbus.h" + +#define NUM_PAGES_SPANNED(addr, len) \ +((PAGE_ALIGN(addr + len) >> PAGE_SHIFT) - (addr >> PAGE_SHIFT)) + +static unsigned long virt_to_hvpfn(void *addr) +{ + phys_addr_t paddr; + + if (is_vmalloc_addr(addr)) + paddr = page_to_phys(vmalloc_to_page(addr)) + + offset_in_page(addr); + else + paddr = __pa(addr); + + return paddr >> PAGE_SHIFT; +} + +/* + * vmbus_setevent- Trigger an event notification on the specified + * channel. + */ +void vmbus_setevent(struct vmbus_channel *channel) +{ + struct hv_monitor_page *monitorpage; + + trace_vmbus_setevent(channel); + + /* + * For channels marked as in "low latency" mode + * bypass the monitor page mechanism. + */ + if (channel->offermsg.monitor_allocated && !channel->low_latency) { + vmbus_send_interrupt(channel->offermsg.child_relid); + + /* Get the child to parent monitor page */ + monitorpage = vmbus_connection.monitor_pages[1]; + + sync_set_bit(channel->monitor_bit, + (unsigned long *)&monitorpage->trigger_group + [channel->monitor_grp].pending); + + } else { + vmbus_set_event(channel); + } +} +EXPORT_SYMBOL_GPL(vmbus_setevent); + +/* + * vmbus_open - Open the specified channel. + */ +int vmbus_open(struct vmbus_channel *newchannel, u32 send_ringbuffer_size, + u32 recv_ringbuffer_size, void *userdata, u32 userdatalen, + void (*onchannelcallback)(void *context), void *context) +{ + struct vmbus_channel_open_channel *open_msg; + struct vmbus_channel_msginfo *open_info = NULL; + unsigned long flags; + int ret, err = 0; + struct page *page; + unsigned int order; + + if (send_ringbuffer_size % PAGE_SIZE || + recv_ringbuffer_size % PAGE_SIZE) + return -EINVAL; + + order = get_order(send_ringbuffer_size + recv_ringbuffer_size); + + spin_lock_irqsave(&newchannel->lock, flags); + if (newchannel->state == CHANNEL_OPEN_STATE) { + newchannel->state = CHANNEL_OPENING_STATE; + } else { + spin_unlock_irqrestore(&newchannel->lock, flags); + return -EINVAL; + } + spin_unlock_irqrestore(&newchannel->lock, flags); + + newchannel->onchannel_callback = onchannelcallback; + newchannel->channel_callback_context = context; + + /* Allocate the ring buffer */ + page = alloc_pages_node(cpu_to_node(newchannel->target_cpu), + GFP_KERNEL|__GFP_ZERO, order); + + if (!page) + page = alloc_pages(GFP_KERNEL|__GFP_ZERO, order); + + if (!page) { + err = -ENOMEM; + goto error_set_chnstate; + } + + newchannel->ringbuffer_page = page; + newchannel->ringbuffer_pagecount = (send_ringbuffer_size + + recv_ringbuffer_size) >> PAGE_SHIFT; + + ret = hv_ringbuffer_init(&newchannel->outbound, page, + send_ringbuffer_size >> PAGE_SHIFT); + + if (ret != 0) { + err = ret; + goto error_free_pages; + } + + ret = hv_ringbuffer_init(&newchannel->inbound, + &page[send_ringbuffer_size >> PAGE_SHIFT], + recv_ringbuffer_size >> PAGE_SHIFT); + if (ret != 0) { + err = ret; + goto error_free_pages; + } + + + /* Establish the gpadl for the ring buffer */ + newchannel->ringbuffer_gpadlhandle = 0; + + ret = vmbus_establish_gpadl(newchannel, + page_address(page), + send_ringbuffer_size + + recv_ringbuffer_size, + &newchannel->ringbuffer_gpadlhandle); + + if (ret != 0) { + err = ret; + goto error_free_pages; + } + + /* Create and init the channel open message */ + open_info = kmalloc(sizeof(*open_info) + + sizeof(struct vmbus_channel_open_channel), + GFP_KERNEL); + if (!open_info) { + err = -ENOMEM; + goto error_free_gpadl; + } + + init_completion(&open_info->waitevent); + open_info->waiting_channel = newchannel; + + open_msg = (struct vmbus_channel_open_channel *)open_info->msg; + open_msg->header.msgtype = CHANNELMSG_OPENCHANNEL; + open_msg->openid = newchannel->offermsg.child_relid; + open_msg->child_relid = newchannel->offermsg.child_relid; + open_msg->ringbuffer_gpadlhandle = newchannel->ringbuffer_gpadlhandle; + open_msg->downstream_ringbuffer_pageoffset = send_ringbuffer_size >> + PAGE_SHIFT; + open_msg->target_vp = newchannel->target_vp; + + if (userdatalen > MAX_USER_DEFINED_BYTES) { + err = -EINVAL; + goto error_free_gpadl; + } + + if (userdatalen) + memcpy(open_msg->userdata, userdata, userdatalen); + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_add_tail(&open_info->msglistentry, + &vmbus_connection.chn_msg_list); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + if (newchannel->rescind) { + err = -ENODEV; + goto error_free_gpadl; + } + + ret = vmbus_post_msg(open_msg, + sizeof(struct vmbus_channel_open_channel), true); + + trace_vmbus_open(open_msg, ret); + + if (ret != 0) { + err = ret; + goto error_clean_msglist; + } + + wait_for_completion(&open_info->waitevent); + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&open_info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + if (newchannel->rescind) { + err = -ENODEV; + goto error_free_gpadl; + } + + if (open_info->response.open_result.status) { + err = -EAGAIN; + goto error_free_gpadl; + } + + newchannel->state = CHANNEL_OPENED_STATE; + kfree(open_info); + return 0; + +error_clean_msglist: + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&open_info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + +error_free_gpadl: + vmbus_teardown_gpadl(newchannel, newchannel->ringbuffer_gpadlhandle); + kfree(open_info); +error_free_pages: + hv_ringbuffer_cleanup(&newchannel->outbound); + hv_ringbuffer_cleanup(&newchannel->inbound); + __free_pages(page, order); +error_set_chnstate: + newchannel->state = CHANNEL_OPEN_STATE; + return err; +} +EXPORT_SYMBOL_GPL(vmbus_open); + +/* Used for Hyper-V Socket: a guest client's connect() to the host */ +int vmbus_send_tl_connect_request(const uuid_le *shv_guest_servie_id, + const uuid_le *shv_host_servie_id) +{ + struct vmbus_channel_tl_connect_request conn_msg; + int ret; + + memset(&conn_msg, 0, sizeof(conn_msg)); + conn_msg.header.msgtype = CHANNELMSG_TL_CONNECT_REQUEST; + conn_msg.guest_endpoint_id = *shv_guest_servie_id; + conn_msg.host_service_id = *shv_host_servie_id; + + ret = vmbus_post_msg(&conn_msg, sizeof(conn_msg), true); + + trace_vmbus_send_tl_connect_request(&conn_msg, ret); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_send_tl_connect_request); + +/* + * create_gpadl_header - Creates a gpadl for the specified buffer + */ +static int create_gpadl_header(void *kbuffer, u32 size, + struct vmbus_channel_msginfo **msginfo) +{ + int i; + int pagecount; + struct vmbus_channel_gpadl_header *gpadl_header; + struct vmbus_channel_gpadl_body *gpadl_body; + struct vmbus_channel_msginfo *msgheader; + struct vmbus_channel_msginfo *msgbody = NULL; + u32 msgsize; + + int pfnsum, pfncount, pfnleft, pfncurr, pfnsize; + + pagecount = size >> PAGE_SHIFT; + + /* do we need a gpadl body msg */ + pfnsize = MAX_SIZE_CHANNEL_MESSAGE - + sizeof(struct vmbus_channel_gpadl_header) - + sizeof(struct gpa_range); + pfncount = pfnsize / sizeof(u64); + + if (pagecount > pfncount) { + /* we need a gpadl body */ + /* fill in the header */ + msgsize = sizeof(struct vmbus_channel_msginfo) + + sizeof(struct vmbus_channel_gpadl_header) + + sizeof(struct gpa_range) + pfncount * sizeof(u64); + msgheader = kzalloc(msgsize, GFP_KERNEL); + if (!msgheader) + goto nomem; + + INIT_LIST_HEAD(&msgheader->submsglist); + msgheader->msgsize = msgsize; + + gpadl_header = (struct vmbus_channel_gpadl_header *) + msgheader->msg; + gpadl_header->rangecount = 1; + gpadl_header->range_buflen = sizeof(struct gpa_range) + + pagecount * sizeof(u64); + gpadl_header->range[0].byte_offset = 0; + gpadl_header->range[0].byte_count = size; + for (i = 0; i < pfncount; i++) + gpadl_header->range[0].pfn_array[i] = virt_to_hvpfn( + kbuffer + PAGE_SIZE * i); + *msginfo = msgheader; + + pfnsum = pfncount; + pfnleft = pagecount - pfncount; + + /* how many pfns can we fit */ + pfnsize = MAX_SIZE_CHANNEL_MESSAGE - + sizeof(struct vmbus_channel_gpadl_body); + pfncount = pfnsize / sizeof(u64); + + /* fill in the body */ + while (pfnleft) { + if (pfnleft > pfncount) + pfncurr = pfncount; + else + pfncurr = pfnleft; + + msgsize = sizeof(struct vmbus_channel_msginfo) + + sizeof(struct vmbus_channel_gpadl_body) + + pfncurr * sizeof(u64); + msgbody = kzalloc(msgsize, GFP_KERNEL); + + if (!msgbody) { + struct vmbus_channel_msginfo *pos = NULL; + struct vmbus_channel_msginfo *tmp = NULL; + /* + * Free up all the allocated messages. + */ + list_for_each_entry_safe(pos, tmp, + &msgheader->submsglist, + msglistentry) { + + list_del(&pos->msglistentry); + kfree(pos); + } + + goto nomem; + } + + msgbody->msgsize = msgsize; + gpadl_body = + (struct vmbus_channel_gpadl_body *)msgbody->msg; + + /* + * Gpadl is u32 and we are using a pointer which could + * be 64-bit + * This is governed by the guest/host protocol and + * so the hypervisor guarantees that this is ok. + */ + for (i = 0; i < pfncurr; i++) + gpadl_body->pfn[i] = virt_to_hvpfn( + kbuffer + PAGE_SIZE * (pfnsum + i)); + + /* add to msg header */ + list_add_tail(&msgbody->msglistentry, + &msgheader->submsglist); + pfnsum += pfncurr; + pfnleft -= pfncurr; + } + } else { + /* everything fits in a header */ + msgsize = sizeof(struct vmbus_channel_msginfo) + + sizeof(struct vmbus_channel_gpadl_header) + + sizeof(struct gpa_range) + pagecount * sizeof(u64); + msgheader = kzalloc(msgsize, GFP_KERNEL); + if (msgheader == NULL) + goto nomem; + + INIT_LIST_HEAD(&msgheader->submsglist); + msgheader->msgsize = msgsize; + + gpadl_header = (struct vmbus_channel_gpadl_header *) + msgheader->msg; + gpadl_header->rangecount = 1; + gpadl_header->range_buflen = sizeof(struct gpa_range) + + pagecount * sizeof(u64); + gpadl_header->range[0].byte_offset = 0; + gpadl_header->range[0].byte_count = size; + for (i = 0; i < pagecount; i++) + gpadl_header->range[0].pfn_array[i] = virt_to_hvpfn( + kbuffer + PAGE_SIZE * i); + + *msginfo = msgheader; + } + + return 0; +nomem: + kfree(msgheader); + kfree(msgbody); + return -ENOMEM; +} + +/* + * vmbus_establish_gpadl - Establish a GPADL for the specified buffer + * + * @channel: a channel + * @kbuffer: from kmalloc or vmalloc + * @size: page-size multiple + * @gpadl_handle: some funky thing + */ +int vmbus_establish_gpadl(struct vmbus_channel *channel, void *kbuffer, + u32 size, u32 *gpadl_handle) +{ + struct vmbus_channel_gpadl_header *gpadlmsg; + struct vmbus_channel_gpadl_body *gpadl_body; + struct vmbus_channel_msginfo *msginfo = NULL; + struct vmbus_channel_msginfo *submsginfo, *tmp; + struct list_head *curr; + u32 next_gpadl_handle; + unsigned long flags; + int ret = 0; + + next_gpadl_handle = + (atomic_inc_return(&vmbus_connection.next_gpadl_handle) - 1); + + ret = create_gpadl_header(kbuffer, size, &msginfo); + if (ret) + return ret; + + init_completion(&msginfo->waitevent); + msginfo->waiting_channel = channel; + + gpadlmsg = (struct vmbus_channel_gpadl_header *)msginfo->msg; + gpadlmsg->header.msgtype = CHANNELMSG_GPADL_HEADER; + gpadlmsg->child_relid = channel->offermsg.child_relid; + gpadlmsg->gpadl = next_gpadl_handle; + + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_add_tail(&msginfo->msglistentry, + &vmbus_connection.chn_msg_list); + + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + if (channel->rescind) { + ret = -ENODEV; + goto cleanup; + } + + ret = vmbus_post_msg(gpadlmsg, msginfo->msgsize - + sizeof(*msginfo), true); + + trace_vmbus_establish_gpadl_header(gpadlmsg, ret); + + if (ret != 0) + goto cleanup; + + list_for_each(curr, &msginfo->submsglist) { + submsginfo = (struct vmbus_channel_msginfo *)curr; + gpadl_body = + (struct vmbus_channel_gpadl_body *)submsginfo->msg; + + gpadl_body->header.msgtype = + CHANNELMSG_GPADL_BODY; + gpadl_body->gpadl = next_gpadl_handle; + + ret = vmbus_post_msg(gpadl_body, + submsginfo->msgsize - sizeof(*submsginfo), + true); + + trace_vmbus_establish_gpadl_body(gpadl_body, ret); + + if (ret != 0) + goto cleanup; + + } + wait_for_completion(&msginfo->waitevent); + + if (msginfo->response.gpadl_created.creation_status != 0) { + pr_err("Failed to establish GPADL: err = 0x%x\n", + msginfo->response.gpadl_created.creation_status); + + ret = -EDQUOT; + goto cleanup; + } + + if (channel->rescind) { + ret = -ENODEV; + goto cleanup; + } + + /* At this point, we received the gpadl created msg */ + *gpadl_handle = gpadlmsg->gpadl; + +cleanup: + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&msginfo->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + list_for_each_entry_safe(submsginfo, tmp, &msginfo->submsglist, + msglistentry) { + kfree(submsginfo); + } + + kfree(msginfo); + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_establish_gpadl); + +/* + * vmbus_teardown_gpadl -Teardown the specified GPADL handle + */ +int vmbus_teardown_gpadl(struct vmbus_channel *channel, u32 gpadl_handle) +{ + struct vmbus_channel_gpadl_teardown *msg; + struct vmbus_channel_msginfo *info; + unsigned long flags; + int ret; + + info = kmalloc(sizeof(*info) + + sizeof(struct vmbus_channel_gpadl_teardown), GFP_KERNEL); + if (!info) + return -ENOMEM; + + init_completion(&info->waitevent); + info->waiting_channel = channel; + + msg = (struct vmbus_channel_gpadl_teardown *)info->msg; + + msg->header.msgtype = CHANNELMSG_GPADL_TEARDOWN; + msg->child_relid = channel->offermsg.child_relid; + msg->gpadl = gpadl_handle; + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_add_tail(&info->msglistentry, + &vmbus_connection.chn_msg_list); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + if (channel->rescind) + goto post_msg_err; + + ret = vmbus_post_msg(msg, sizeof(struct vmbus_channel_gpadl_teardown), + true); + + trace_vmbus_teardown_gpadl(msg, ret); + + if (ret) + goto post_msg_err; + + wait_for_completion(&info->waitevent); + +post_msg_err: + /* + * If the channel has been rescinded; + * we will be awakened by the rescind + * handler; set the error code to zero so we don't leak memory. + */ + if (channel->rescind) + ret = 0; + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + kfree(info); + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_teardown_gpadl); + +static void reset_channel_cb(void *arg) +{ + struct vmbus_channel *channel = arg; + + channel->onchannel_callback = NULL; +} + +void vmbus_reset_channel_cb(struct vmbus_channel *channel) +{ + /* + * vmbus_on_event(), running in the per-channel tasklet, can race + * with vmbus_close_internal() in the case of SMP guest, e.g., when + * the former is accessing channel->inbound.ring_buffer, the latter + * could be freeing the ring_buffer pages, so here we must stop it + * first. + */ + tasklet_disable(&channel->callback_event); + + channel->sc_creation_callback = NULL; + + /* Stop the callback asap */ + if (channel->target_cpu != get_cpu()) { + put_cpu(); + smp_call_function_single(channel->target_cpu, reset_channel_cb, + channel, true); + } else { + reset_channel_cb(channel); + put_cpu(); + } + + /* Re-enable tasklet for use on re-open */ + tasklet_enable(&channel->callback_event); +} + +static int vmbus_close_internal(struct vmbus_channel *channel) +{ + struct vmbus_channel_close_channel *msg; + int ret; + + vmbus_reset_channel_cb(channel); + + /* + * In case a device driver's probe() fails (e.g., + * util_probe() -> vmbus_open() returns -ENOMEM) and the device is + * rescinded later (e.g., we dynamically disable an Integrated Service + * in Hyper-V Manager), the driver's remove() invokes vmbus_close(): + * here we should skip most of the below cleanup work. + */ + if (channel->state != CHANNEL_OPENED_STATE) { + ret = -EINVAL; + goto out; + } + + channel->state = CHANNEL_OPEN_STATE; + + /* Send a closing message */ + + msg = &channel->close_msg.msg; + + msg->header.msgtype = CHANNELMSG_CLOSECHANNEL; + msg->child_relid = channel->offermsg.child_relid; + + ret = vmbus_post_msg(msg, sizeof(struct vmbus_channel_close_channel), + true); + + trace_vmbus_close_internal(msg, ret); + + if (ret) { + pr_err("Close failed: close post msg return is %d\n", ret); + /* + * If we failed to post the close msg, + * it is perhaps better to leak memory. + */ + goto out; + } + + /* Tear down the gpadl for the channel's ring buffer */ + if (channel->ringbuffer_gpadlhandle) { + ret = vmbus_teardown_gpadl(channel, + channel->ringbuffer_gpadlhandle); + if (ret) { + pr_err("Close failed: teardown gpadl return %d\n", ret); + /* + * If we failed to teardown gpadl, + * it is perhaps better to leak memory. + */ + goto out; + } + } + + /* Cleanup the ring buffers for this channel */ + hv_ringbuffer_cleanup(&channel->outbound); + hv_ringbuffer_cleanup(&channel->inbound); + + __free_pages(channel->ringbuffer_page, + get_order(channel->ringbuffer_pagecount << PAGE_SHIFT)); + +out: + return ret; +} + +/* + * vmbus_close - Close the specified channel + */ +void vmbus_close(struct vmbus_channel *channel) +{ + struct list_head *cur, *tmp; + struct vmbus_channel *cur_channel; + + if (channel->primary_channel != NULL) { + /* + * We will only close sub-channels when + * the primary is closed. + */ + return; + } + /* + * Close all the sub-channels first and then close the + * primary channel. + */ + list_for_each_safe(cur, tmp, &channel->sc_list) { + cur_channel = list_entry(cur, struct vmbus_channel, sc_list); + if (cur_channel->rescind) { + wait_for_completion(&cur_channel->rescind_event); + mutex_lock(&vmbus_connection.channel_mutex); + vmbus_close_internal(cur_channel); + hv_process_channel_removal( + cur_channel->offermsg.child_relid); + } else { + mutex_lock(&vmbus_connection.channel_mutex); + vmbus_close_internal(cur_channel); + } + mutex_unlock(&vmbus_connection.channel_mutex); + } + /* + * Now close the primary. + */ + mutex_lock(&vmbus_connection.channel_mutex); + vmbus_close_internal(channel); + mutex_unlock(&vmbus_connection.channel_mutex); +} +EXPORT_SYMBOL_GPL(vmbus_close); + +/** + * vmbus_sendpacket() - Send the specified buffer on the given channel + * @channel: Pointer to vmbus_channel structure. + * @buffer: Pointer to the buffer you want to receive the data into. + * @bufferlen: Maximum size of what the the buffer will hold + * @requestid: Identifier of the request + * @type: Type of packet that is being send e.g. negotiate, time + * packet etc. + * + * Sends data in @buffer directly to hyper-v via the vmbus + * This will send the data unparsed to hyper-v. + * + * Mainly used by Hyper-V drivers. + */ +int vmbus_sendpacket(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u64 requestid, + enum vmbus_packet_type type, u32 flags) +{ + struct vmpacket_descriptor desc; + u32 packetlen = sizeof(struct vmpacket_descriptor) + bufferlen; + u32 packetlen_aligned = ALIGN(packetlen, sizeof(u64)); + struct kvec bufferlist[3]; + u64 aligned_data = 0; + int num_vecs = ((bufferlen != 0) ? 3 : 1); + + + /* Setup the descriptor */ + desc.type = type; /* VmbusPacketTypeDataInBand; */ + desc.flags = flags; /* VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED; */ + /* in 8-bytes granularity */ + desc.offset8 = sizeof(struct vmpacket_descriptor) >> 3; + desc.len8 = (u16)(packetlen_aligned >> 3); + desc.trans_id = requestid; + + bufferlist[0].iov_base = &desc; + bufferlist[0].iov_len = sizeof(struct vmpacket_descriptor); + bufferlist[1].iov_base = buffer; + bufferlist[1].iov_len = bufferlen; + bufferlist[2].iov_base = &aligned_data; + bufferlist[2].iov_len = (packetlen_aligned - packetlen); + + return hv_ringbuffer_write(channel, bufferlist, num_vecs); +} +EXPORT_SYMBOL(vmbus_sendpacket); + +/* + * vmbus_sendpacket_pagebuffer - Send a range of single-page buffer + * packets using a GPADL Direct packet type. This interface allows you + * to control notifying the host. This will be useful for sending + * batched data. Also the sender can control the send flags + * explicitly. + */ +int vmbus_sendpacket_pagebuffer(struct vmbus_channel *channel, + struct hv_page_buffer pagebuffers[], + u32 pagecount, void *buffer, u32 bufferlen, + u64 requestid) +{ + int i; + struct vmbus_channel_packet_page_buffer desc; + u32 descsize; + u32 packetlen; + u32 packetlen_aligned; + struct kvec bufferlist[3]; + u64 aligned_data = 0; + + if (pagecount > MAX_PAGE_BUFFER_COUNT) + return -EINVAL; + + /* + * Adjust the size down since vmbus_channel_packet_page_buffer is the + * largest size we support + */ + descsize = sizeof(struct vmbus_channel_packet_page_buffer) - + ((MAX_PAGE_BUFFER_COUNT - pagecount) * + sizeof(struct hv_page_buffer)); + packetlen = descsize + bufferlen; + packetlen_aligned = ALIGN(packetlen, sizeof(u64)); + + /* Setup the descriptor */ + desc.type = VM_PKT_DATA_USING_GPA_DIRECT; + desc.flags = VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED; + desc.dataoffset8 = descsize >> 3; /* in 8-bytes granularity */ + desc.length8 = (u16)(packetlen_aligned >> 3); + desc.transactionid = requestid; + desc.reserved = 0; + desc.rangecount = pagecount; + + for (i = 0; i < pagecount; i++) { + desc.range[i].len = pagebuffers[i].len; + desc.range[i].offset = pagebuffers[i].offset; + desc.range[i].pfn = pagebuffers[i].pfn; + } + + bufferlist[0].iov_base = &desc; + bufferlist[0].iov_len = descsize; + bufferlist[1].iov_base = buffer; + bufferlist[1].iov_len = bufferlen; + bufferlist[2].iov_base = &aligned_data; + bufferlist[2].iov_len = (packetlen_aligned - packetlen); + + return hv_ringbuffer_write(channel, bufferlist, 3); +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_pagebuffer); + +/* + * vmbus_sendpacket_multipagebuffer - Send a multi-page buffer packet + * using a GPADL Direct packet type. + * The buffer includes the vmbus descriptor. + */ +int vmbus_sendpacket_mpb_desc(struct vmbus_channel *channel, + struct vmbus_packet_mpb_array *desc, + u32 desc_size, + void *buffer, u32 bufferlen, u64 requestid) +{ + u32 packetlen; + u32 packetlen_aligned; + struct kvec bufferlist[3]; + u64 aligned_data = 0; + + packetlen = desc_size + bufferlen; + packetlen_aligned = ALIGN(packetlen, sizeof(u64)); + + /* Setup the descriptor */ + desc->type = VM_PKT_DATA_USING_GPA_DIRECT; + desc->flags = VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED; + desc->dataoffset8 = desc_size >> 3; /* in 8-bytes granularity */ + desc->length8 = (u16)(packetlen_aligned >> 3); + desc->transactionid = requestid; + desc->reserved = 0; + desc->rangecount = 1; + + bufferlist[0].iov_base = desc; + bufferlist[0].iov_len = desc_size; + bufferlist[1].iov_base = buffer; + bufferlist[1].iov_len = bufferlen; + bufferlist[2].iov_base = &aligned_data; + bufferlist[2].iov_len = (packetlen_aligned - packetlen); + + return hv_ringbuffer_write(channel, bufferlist, 3); +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_mpb_desc); + +/** + * vmbus_recvpacket() - Retrieve the user packet on the specified channel + * @channel: Pointer to vmbus_channel structure. + * @buffer: Pointer to the buffer you want to receive the data into. + * @bufferlen: Maximum size of what the the buffer will hold + * @buffer_actual_len: The actual size of the data after it was received + * @requestid: Identifier of the request + * + * Receives directly from the hyper-v vmbus and puts the data it received + * into Buffer. This will receive the data unparsed from hyper-v. + * + * Mainly used by Hyper-V drivers. + */ +static inline int +__vmbus_recvpacket(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u32 *buffer_actual_len, u64 *requestid, + bool raw) +{ + return hv_ringbuffer_read(channel, buffer, bufferlen, + buffer_actual_len, requestid, raw); + +} + +int vmbus_recvpacket(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u32 *buffer_actual_len, + u64 *requestid) +{ + return __vmbus_recvpacket(channel, buffer, bufferlen, + buffer_actual_len, requestid, false); +} +EXPORT_SYMBOL(vmbus_recvpacket); + +/* + * vmbus_recvpacket_raw - Retrieve the raw packet on the specified channel + */ +int vmbus_recvpacket_raw(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u32 *buffer_actual_len, + u64 *requestid) +{ + return __vmbus_recvpacket(channel, buffer, bufferlen, + buffer_actual_len, requestid, true); +} +EXPORT_SYMBOL_GPL(vmbus_recvpacket_raw); diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c new file mode 100644 index 000000000..a3f6933f9 --- /dev/null +++ b/drivers/hv/channel_mgmt.c @@ -0,0 +1,1425 @@ +/* + * 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> + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/interrupt.h> +#include <linux/sched.h> +#include <linux/wait.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/list.h> +#include <linux/module.h> +#include <linux/completion.h> +#include <linux/delay.h> +#include <linux/hyperv.h> +#include <asm/mshyperv.h> + +#include "hyperv_vmbus.h" + +static void init_vp_index(struct vmbus_channel *channel, u16 dev_type); + +static const struct vmbus_device vmbus_devs[] = { + /* IDE */ + { .dev_type = HV_IDE, + HV_IDE_GUID, + .perf_device = true, + }, + + /* SCSI */ + { .dev_type = HV_SCSI, + HV_SCSI_GUID, + .perf_device = true, + }, + + /* Fibre Channel */ + { .dev_type = HV_FC, + HV_SYNTHFC_GUID, + .perf_device = true, + }, + + /* Synthetic NIC */ + { .dev_type = HV_NIC, + HV_NIC_GUID, + .perf_device = true, + }, + + /* Network Direct */ + { .dev_type = HV_ND, + HV_ND_GUID, + .perf_device = true, + }, + + /* PCIE */ + { .dev_type = HV_PCIE, + HV_PCIE_GUID, + .perf_device = false, + }, + + /* Synthetic Frame Buffer */ + { .dev_type = HV_FB, + HV_SYNTHVID_GUID, + .perf_device = false, + }, + + /* Synthetic Keyboard */ + { .dev_type = HV_KBD, + HV_KBD_GUID, + .perf_device = false, + }, + + /* Synthetic MOUSE */ + { .dev_type = HV_MOUSE, + HV_MOUSE_GUID, + .perf_device = false, + }, + + /* KVP */ + { .dev_type = HV_KVP, + HV_KVP_GUID, + .perf_device = false, + }, + + /* Time Synch */ + { .dev_type = HV_TS, + HV_TS_GUID, + .perf_device = false, + }, + + /* Heartbeat */ + { .dev_type = HV_HB, + HV_HEART_BEAT_GUID, + .perf_device = false, + }, + + /* Shutdown */ + { .dev_type = HV_SHUTDOWN, + HV_SHUTDOWN_GUID, + .perf_device = false, + }, + + /* File copy */ + { .dev_type = HV_FCOPY, + HV_FCOPY_GUID, + .perf_device = false, + }, + + /* Backup */ + { .dev_type = HV_BACKUP, + HV_VSS_GUID, + .perf_device = false, + }, + + /* Dynamic Memory */ + { .dev_type = HV_DM, + HV_DM_GUID, + .perf_device = false, + }, + + /* Unknown GUID */ + { .dev_type = HV_UNKNOWN, + .perf_device = false, + }, +}; + +static const struct { + uuid_le guid; +} vmbus_unsupported_devs[] = { + { HV_AVMA1_GUID }, + { HV_AVMA2_GUID }, + { HV_RDV_GUID }, +}; + +/* + * The rescinded channel may be blocked waiting for a response from the host; + * take care of that. + */ +static void vmbus_rescind_cleanup(struct vmbus_channel *channel) +{ + struct vmbus_channel_msginfo *msginfo; + unsigned long flags; + + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + channel->rescind = true; + list_for_each_entry(msginfo, &vmbus_connection.chn_msg_list, + msglistentry) { + + if (msginfo->waiting_channel == channel) { + complete(&msginfo->waitevent); + break; + } + } + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); +} + +static bool is_unsupported_vmbus_devs(const uuid_le *guid) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(vmbus_unsupported_devs); i++) + if (!uuid_le_cmp(*guid, vmbus_unsupported_devs[i].guid)) + return true; + return false; +} + +static u16 hv_get_dev_type(const struct vmbus_channel *channel) +{ + const uuid_le *guid = &channel->offermsg.offer.if_type; + u16 i; + + if (is_hvsock_channel(channel) || is_unsupported_vmbus_devs(guid)) + return HV_UNKNOWN; + + for (i = HV_IDE; i < HV_UNKNOWN; i++) { + if (!uuid_le_cmp(*guid, vmbus_devs[i].guid)) + return i; + } + pr_info("Unknown GUID: %pUl\n", guid); + return i; +} + +/** + * vmbus_prep_negotiate_resp() - Create default response for Hyper-V Negotiate message + * @icmsghdrp: Pointer to msg header structure + * @icmsg_negotiate: Pointer to negotiate message structure + * @buf: Raw buffer channel data + * + * @icmsghdrp is of type &struct icmsg_hdr. + * Set up and fill in default negotiate response message. + * + * The fw_version and fw_vercnt specifies the framework version that + * we can support. + * + * The srv_version and srv_vercnt specifies the service + * versions we can support. + * + * Versions are given in decreasing order. + * + * nego_fw_version and nego_srv_version store the selected protocol versions. + * + * Mainly used by Hyper-V drivers. + */ +bool vmbus_prep_negotiate_resp(struct icmsg_hdr *icmsghdrp, + u8 *buf, const int *fw_version, int fw_vercnt, + const int *srv_version, int srv_vercnt, + int *nego_fw_version, int *nego_srv_version) +{ + int icframe_major, icframe_minor; + int icmsg_major, icmsg_minor; + int fw_major, fw_minor; + int srv_major, srv_minor; + int i, j; + bool found_match = false; + struct icmsg_negotiate *negop; + + icmsghdrp->icmsgsize = 0x10; + negop = (struct icmsg_negotiate *)&buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + icframe_major = negop->icframe_vercnt; + icframe_minor = 0; + + icmsg_major = negop->icmsg_vercnt; + icmsg_minor = 0; + + /* + * Select the framework version number we will + * support. + */ + + for (i = 0; i < fw_vercnt; i++) { + fw_major = (fw_version[i] >> 16); + fw_minor = (fw_version[i] & 0xFFFF); + + for (j = 0; j < negop->icframe_vercnt; j++) { + if ((negop->icversion_data[j].major == fw_major) && + (negop->icversion_data[j].minor == fw_minor)) { + icframe_major = negop->icversion_data[j].major; + icframe_minor = negop->icversion_data[j].minor; + found_match = true; + break; + } + } + + if (found_match) + break; + } + + if (!found_match) + goto fw_error; + + found_match = false; + + for (i = 0; i < srv_vercnt; i++) { + srv_major = (srv_version[i] >> 16); + srv_minor = (srv_version[i] & 0xFFFF); + + for (j = negop->icframe_vercnt; + (j < negop->icframe_vercnt + negop->icmsg_vercnt); + j++) { + + if ((negop->icversion_data[j].major == srv_major) && + (negop->icversion_data[j].minor == srv_minor)) { + + icmsg_major = negop->icversion_data[j].major; + icmsg_minor = negop->icversion_data[j].minor; + found_match = true; + break; + } + } + + if (found_match) + break; + } + + /* + * Respond with the framework and service + * version numbers we can support. + */ + +fw_error: + if (!found_match) { + negop->icframe_vercnt = 0; + negop->icmsg_vercnt = 0; + } else { + negop->icframe_vercnt = 1; + negop->icmsg_vercnt = 1; + } + + if (nego_fw_version) + *nego_fw_version = (icframe_major << 16) | icframe_minor; + + if (nego_srv_version) + *nego_srv_version = (icmsg_major << 16) | icmsg_minor; + + negop->icversion_data[0].major = icframe_major; + negop->icversion_data[0].minor = icframe_minor; + negop->icversion_data[1].major = icmsg_major; + negop->icversion_data[1].minor = icmsg_minor; + return found_match; +} + +EXPORT_SYMBOL_GPL(vmbus_prep_negotiate_resp); + +/* + * alloc_channel - Allocate and initialize a vmbus channel object + */ +static struct vmbus_channel *alloc_channel(void) +{ + struct vmbus_channel *channel; + + channel = kzalloc(sizeof(*channel), GFP_ATOMIC); + if (!channel) + return NULL; + + spin_lock_init(&channel->lock); + init_completion(&channel->rescind_event); + + INIT_LIST_HEAD(&channel->sc_list); + INIT_LIST_HEAD(&channel->percpu_list); + + tasklet_init(&channel->callback_event, + vmbus_on_event, (unsigned long)channel); + + return channel; +} + +/* + * free_channel - Release the resources used by the vmbus channel object + */ +static void free_channel(struct vmbus_channel *channel) +{ + tasklet_kill(&channel->callback_event); + vmbus_remove_channel_attr_group(channel); + + kobject_put(&channel->kobj); +} + +static void percpu_channel_enq(void *arg) +{ + struct vmbus_channel *channel = arg; + struct hv_per_cpu_context *hv_cpu + = this_cpu_ptr(hv_context.cpu_context); + + list_add_tail_rcu(&channel->percpu_list, &hv_cpu->chan_list); +} + +static void percpu_channel_deq(void *arg) +{ + struct vmbus_channel *channel = arg; + + list_del_rcu(&channel->percpu_list); +} + + +static void vmbus_release_relid(u32 relid) +{ + struct vmbus_channel_relid_released msg; + int ret; + + memset(&msg, 0, sizeof(struct vmbus_channel_relid_released)); + msg.child_relid = relid; + msg.header.msgtype = CHANNELMSG_RELID_RELEASED; + ret = vmbus_post_msg(&msg, sizeof(struct vmbus_channel_relid_released), + true); + + trace_vmbus_release_relid(&msg, ret); +} + +void hv_process_channel_removal(u32 relid) +{ + unsigned long flags; + struct vmbus_channel *primary_channel, *channel; + + BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex)); + + /* + * Make sure channel is valid as we may have raced. + */ + channel = relid2channel(relid); + if (!channel) + return; + + BUG_ON(!channel->rescind); + if (channel->target_cpu != get_cpu()) { + put_cpu(); + smp_call_function_single(channel->target_cpu, + percpu_channel_deq, channel, true); + } else { + percpu_channel_deq(channel); + put_cpu(); + } + + if (channel->primary_channel == NULL) { + list_del(&channel->listentry); + + primary_channel = channel; + } else { + primary_channel = channel->primary_channel; + spin_lock_irqsave(&primary_channel->lock, flags); + list_del(&channel->sc_list); + primary_channel->num_sc--; + spin_unlock_irqrestore(&primary_channel->lock, flags); + } + + /* + * We need to free the bit for init_vp_index() to work in the case + * of sub-channel, when we reload drivers like hv_netvsc. + */ + if (channel->affinity_policy == HV_LOCALIZED) + cpumask_clear_cpu(channel->target_cpu, + &primary_channel->alloced_cpus_in_node); + + vmbus_release_relid(relid); + + free_channel(channel); +} + +void vmbus_free_channels(void) +{ + struct vmbus_channel *channel, *tmp; + + list_for_each_entry_safe(channel, tmp, &vmbus_connection.chn_list, + listentry) { + /* hv_process_channel_removal() needs this */ + channel->rescind = true; + + vmbus_device_unregister(channel->device_obj); + } +} + +/* Note: the function can run concurrently for primary/sub channels. */ +static void vmbus_add_channel_work(struct work_struct *work) +{ + struct vmbus_channel *newchannel = + container_of(work, struct vmbus_channel, add_channel_work); + struct vmbus_channel *primary_channel = newchannel->primary_channel; + unsigned long flags; + u16 dev_type; + int ret; + + dev_type = hv_get_dev_type(newchannel); + + init_vp_index(newchannel, dev_type); + + if (newchannel->target_cpu != get_cpu()) { + put_cpu(); + smp_call_function_single(newchannel->target_cpu, + percpu_channel_enq, + newchannel, true); + } else { + percpu_channel_enq(newchannel); + put_cpu(); + } + + /* + * This state is used to indicate a successful open + * so that when we do close the channel normally, we + * can cleanup properly. + */ + newchannel->state = CHANNEL_OPEN_STATE; + + if (primary_channel != NULL) { + /* newchannel is a sub-channel. */ + struct hv_device *dev = primary_channel->device_obj; + + if (vmbus_add_channel_kobj(dev, newchannel)) + goto err_deq_chan; + + if (primary_channel->sc_creation_callback != NULL) + primary_channel->sc_creation_callback(newchannel); + + newchannel->probe_done = true; + return; + } + + /* + * Start the process of binding the primary channel to the driver + */ + newchannel->device_obj = vmbus_device_create( + &newchannel->offermsg.offer.if_type, + &newchannel->offermsg.offer.if_instance, + newchannel); + if (!newchannel->device_obj) + goto err_deq_chan; + + newchannel->device_obj->device_id = dev_type; + /* + * Add the new device to the bus. This will kick off device-driver + * binding which eventually invokes the device driver's AddDevice() + * method. + */ + ret = vmbus_device_register(newchannel->device_obj); + + if (ret != 0) { + pr_err("unable to add child device object (relid %d)\n", + newchannel->offermsg.child_relid); + kfree(newchannel->device_obj); + goto err_deq_chan; + } + + newchannel->probe_done = true; + return; + +err_deq_chan: + mutex_lock(&vmbus_connection.channel_mutex); + + /* + * We need to set the flag, otherwise + * vmbus_onoffer_rescind() can be blocked. + */ + newchannel->probe_done = true; + + if (primary_channel == NULL) { + list_del(&newchannel->listentry); + } else { + spin_lock_irqsave(&primary_channel->lock, flags); + list_del(&newchannel->sc_list); + spin_unlock_irqrestore(&primary_channel->lock, flags); + } + + mutex_unlock(&vmbus_connection.channel_mutex); + + if (newchannel->target_cpu != get_cpu()) { + put_cpu(); + smp_call_function_single(newchannel->target_cpu, + percpu_channel_deq, + newchannel, true); + } else { + percpu_channel_deq(newchannel); + put_cpu(); + } + + vmbus_release_relid(newchannel->offermsg.child_relid); + + free_channel(newchannel); +} + +/* + * vmbus_process_offer - Process the offer by creating a channel/device + * associated with this offer + */ +static void vmbus_process_offer(struct vmbus_channel *newchannel) +{ + struct vmbus_channel *channel; + struct workqueue_struct *wq; + unsigned long flags; + bool fnew = true; + + mutex_lock(&vmbus_connection.channel_mutex); + + /* + * Now that we have acquired the channel_mutex, + * we can release the potentially racing rescind thread. + */ + atomic_dec(&vmbus_connection.offer_in_progress); + + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + if (!uuid_le_cmp(channel->offermsg.offer.if_type, + newchannel->offermsg.offer.if_type) && + !uuid_le_cmp(channel->offermsg.offer.if_instance, + newchannel->offermsg.offer.if_instance)) { + fnew = false; + break; + } + } + + if (fnew) + list_add_tail(&newchannel->listentry, + &vmbus_connection.chn_list); + else { + /* + * Check to see if this is a valid sub-channel. + */ + if (newchannel->offermsg.offer.sub_channel_index == 0) { + mutex_unlock(&vmbus_connection.channel_mutex); + /* + * Don't call free_channel(), because newchannel->kobj + * is not initialized yet. + */ + kfree(newchannel); + WARN_ON_ONCE(1); + return; + } + /* + * Process the sub-channel. + */ + newchannel->primary_channel = channel; + spin_lock_irqsave(&channel->lock, flags); + list_add_tail(&newchannel->sc_list, &channel->sc_list); + spin_unlock_irqrestore(&channel->lock, flags); + } + + mutex_unlock(&vmbus_connection.channel_mutex); + + /* + * vmbus_process_offer() mustn't call channel->sc_creation_callback() + * directly for sub-channels, because sc_creation_callback() -> + * vmbus_open() may never get the host's response to the + * OPEN_CHANNEL message (the host may rescind a channel at any time, + * e.g. in the case of hot removing a NIC), and vmbus_onoffer_rescind() + * may not wake up the vmbus_open() as it's blocked due to a non-zero + * vmbus_connection.offer_in_progress, and finally we have a deadlock. + * + * The above is also true for primary channels, if the related device + * drivers use sync probing mode by default. + * + * And, usually the handling of primary channels and sub-channels can + * depend on each other, so we should offload them to different + * workqueues to avoid possible deadlock, e.g. in sync-probing mode, + * NIC1's netvsc_subchan_work() can race with NIC2's netvsc_probe() -> + * rtnl_lock(), and causes deadlock: the former gets the rtnl_lock + * and waits for all the sub-channels to appear, but the latter + * can't get the rtnl_lock and this blocks the handling of + * sub-channels. + */ + INIT_WORK(&newchannel->add_channel_work, vmbus_add_channel_work); + wq = fnew ? vmbus_connection.handle_primary_chan_wq : + vmbus_connection.handle_sub_chan_wq; + queue_work(wq, &newchannel->add_channel_work); +} + +/* + * We use this state to statically distribute the channel interrupt load. + */ +static int next_numa_node_id; +/* + * init_vp_index() accesses global variables like next_numa_node_id, and + * it can run concurrently for primary channels and sub-channels: see + * vmbus_process_offer(), so we need the lock to protect the global + * variables. + */ +static DEFINE_SPINLOCK(bind_channel_to_cpu_lock); + +/* + * Starting with Win8, we can statically distribute the incoming + * channel interrupt load by binding a channel to VCPU. + * We distribute the interrupt loads to one or more NUMA nodes based on + * the channel's affinity_policy. + * + * For pre-win8 hosts or non-performance critical channels we assign the + * first CPU in the first NUMA node. + */ +static void init_vp_index(struct vmbus_channel *channel, u16 dev_type) +{ + u32 cur_cpu; + bool perf_chn = vmbus_devs[dev_type].perf_device; + struct vmbus_channel *primary = channel->primary_channel; + int next_node; + cpumask_var_t available_mask; + struct cpumask *alloced_mask; + + if ((vmbus_proto_version == VERSION_WS2008) || + (vmbus_proto_version == VERSION_WIN7) || (!perf_chn) || + !alloc_cpumask_var(&available_mask, GFP_KERNEL)) { + /* + * Prior to win8, all channel interrupts are + * delivered on cpu 0. + * Also if the channel is not a performance critical + * channel, bind it to cpu 0. + * In case alloc_cpumask_var() fails, bind it to cpu 0. + */ + channel->numa_node = 0; + channel->target_cpu = 0; + channel->target_vp = hv_cpu_number_to_vp_number(0); + return; + } + + spin_lock(&bind_channel_to_cpu_lock); + + /* + * Based on the channel affinity policy, we will assign the NUMA + * nodes. + */ + + if ((channel->affinity_policy == HV_BALANCED) || (!primary)) { + while (true) { + next_node = next_numa_node_id++; + if (next_node == nr_node_ids) { + next_node = next_numa_node_id = 0; + continue; + } + if (cpumask_empty(cpumask_of_node(next_node))) + continue; + break; + } + channel->numa_node = next_node; + primary = channel; + } + alloced_mask = &hv_context.hv_numa_map[primary->numa_node]; + + if (cpumask_weight(alloced_mask) == + cpumask_weight(cpumask_of_node(primary->numa_node))) { + /* + * We have cycled through all the CPUs in the node; + * reset the alloced map. + */ + cpumask_clear(alloced_mask); + } + + cpumask_xor(available_mask, alloced_mask, + cpumask_of_node(primary->numa_node)); + + cur_cpu = -1; + + if (primary->affinity_policy == HV_LOCALIZED) { + /* + * Normally Hyper-V host doesn't create more subchannels + * than there are VCPUs on the node but it is possible when not + * all present VCPUs on the node are initialized by guest. + * Clear the alloced_cpus_in_node to start over. + */ + if (cpumask_equal(&primary->alloced_cpus_in_node, + cpumask_of_node(primary->numa_node))) + cpumask_clear(&primary->alloced_cpus_in_node); + } + + while (true) { + cur_cpu = cpumask_next(cur_cpu, available_mask); + if (cur_cpu >= nr_cpu_ids) { + cur_cpu = -1; + cpumask_copy(available_mask, + cpumask_of_node(primary->numa_node)); + continue; + } + + if (primary->affinity_policy == HV_LOCALIZED) { + /* + * NOTE: in the case of sub-channel, we clear the + * sub-channel related bit(s) in + * primary->alloced_cpus_in_node in + * hv_process_channel_removal(), so when we + * reload drivers like hv_netvsc in SMP guest, here + * we're able to re-allocate + * bit from primary->alloced_cpus_in_node. + */ + if (!cpumask_test_cpu(cur_cpu, + &primary->alloced_cpus_in_node)) { + cpumask_set_cpu(cur_cpu, + &primary->alloced_cpus_in_node); + cpumask_set_cpu(cur_cpu, alloced_mask); + break; + } + } else { + cpumask_set_cpu(cur_cpu, alloced_mask); + break; + } + } + + channel->target_cpu = cur_cpu; + channel->target_vp = hv_cpu_number_to_vp_number(cur_cpu); + + spin_unlock(&bind_channel_to_cpu_lock); + + free_cpumask_var(available_mask); +} + +#define UNLOAD_DELAY_UNIT_MS 10 /* 10 milliseconds */ +#define UNLOAD_WAIT_MS (100*1000) /* 100 seconds */ +#define UNLOAD_WAIT_LOOPS (UNLOAD_WAIT_MS/UNLOAD_DELAY_UNIT_MS) +#define UNLOAD_MSG_MS (5*1000) /* Every 5 seconds */ +#define UNLOAD_MSG_LOOPS (UNLOAD_MSG_MS/UNLOAD_DELAY_UNIT_MS) + +static void vmbus_wait_for_unload(void) +{ + int cpu; + void *page_addr; + struct hv_message *msg; + struct vmbus_channel_message_header *hdr; + u32 message_type, i; + + /* + * CHANNELMSG_UNLOAD_RESPONSE is always delivered to the CPU which was + * used for initial contact or to CPU0 depending on host version. When + * we're crashing on a different CPU let's hope that IRQ handler on + * the cpu which receives CHANNELMSG_UNLOAD_RESPONSE is still + * functional and vmbus_unload_response() will complete + * vmbus_connection.unload_event. If not, the last thing we can do is + * read message pages for all CPUs directly. + * + * Wait up to 100 seconds since an Azure host must writeback any dirty + * data in its disk cache before the VMbus UNLOAD request will + * complete. This flushing has been empirically observed to take up + * to 50 seconds in cases with a lot of dirty data, so allow additional + * leeway and for inaccuracies in mdelay(). But eventually time out so + * that the panic path can't get hung forever in case the response + * message isn't seen. + */ + for (i = 1; i <= UNLOAD_WAIT_LOOPS; i++) { + if (completion_done(&vmbus_connection.unload_event)) + goto completed; + + for_each_online_cpu(cpu) { + struct hv_per_cpu_context *hv_cpu + = per_cpu_ptr(hv_context.cpu_context, cpu); + + page_addr = hv_cpu->synic_message_page; + msg = (struct hv_message *)page_addr + + VMBUS_MESSAGE_SINT; + + message_type = READ_ONCE(msg->header.message_type); + if (message_type == HVMSG_NONE) + continue; + + hdr = (struct vmbus_channel_message_header *) + msg->u.payload; + + if (hdr->msgtype == CHANNELMSG_UNLOAD_RESPONSE) + complete(&vmbus_connection.unload_event); + + vmbus_signal_eom(msg, message_type); + } + + /* + * Give a notice periodically so someone watching the + * serial output won't think it is completely hung. + */ + if (!(i % UNLOAD_MSG_LOOPS)) + pr_notice("Waiting for VMBus UNLOAD to complete\n"); + + mdelay(UNLOAD_DELAY_UNIT_MS); + } + pr_err("Continuing even though VMBus UNLOAD did not complete\n"); + +completed: + /* + * We're crashing and already got the UNLOAD_RESPONSE, cleanup all + * maybe-pending messages on all CPUs to be able to receive new + * messages after we reconnect. + */ + for_each_online_cpu(cpu) { + struct hv_per_cpu_context *hv_cpu + = per_cpu_ptr(hv_context.cpu_context, cpu); + + page_addr = hv_cpu->synic_message_page; + msg = (struct hv_message *)page_addr + VMBUS_MESSAGE_SINT; + msg->header.message_type = HVMSG_NONE; + } +} + +/* + * vmbus_unload_response - Handler for the unload response. + */ +static void vmbus_unload_response(struct vmbus_channel_message_header *hdr) +{ + /* + * This is a global event; just wakeup the waiting thread. + * Once we successfully unload, we can cleanup the monitor state. + */ + complete(&vmbus_connection.unload_event); +} + +void vmbus_initiate_unload(bool crash) +{ + struct vmbus_channel_message_header hdr; + + if (xchg(&vmbus_connection.conn_state, DISCONNECTED) == DISCONNECTED) + return; + + /* Pre-Win2012R2 hosts don't support reconnect */ + if (vmbus_proto_version < VERSION_WIN8_1) + return; + + init_completion(&vmbus_connection.unload_event); + memset(&hdr, 0, sizeof(struct vmbus_channel_message_header)); + hdr.msgtype = CHANNELMSG_UNLOAD; + vmbus_post_msg(&hdr, sizeof(struct vmbus_channel_message_header), + !crash); + + /* + * vmbus_initiate_unload() is also called on crash and the crash can be + * happening in an interrupt context, where scheduling is impossible. + */ + if (!crash) + wait_for_completion(&vmbus_connection.unload_event); + else + vmbus_wait_for_unload(); +} + +/* + * vmbus_onoffer - Handler for channel offers from vmbus in parent partition. + * + */ +static void vmbus_onoffer(struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_offer_channel *offer; + struct vmbus_channel *newchannel; + + offer = (struct vmbus_channel_offer_channel *)hdr; + + trace_vmbus_onoffer(offer); + + /* Allocate the channel object and save this offer. */ + newchannel = alloc_channel(); + if (!newchannel) { + vmbus_release_relid(offer->child_relid); + atomic_dec(&vmbus_connection.offer_in_progress); + pr_err("Unable to allocate channel object\n"); + return; + } + + /* + * Setup state for signalling the host. + */ + newchannel->sig_event = VMBUS_EVENT_CONNECTION_ID; + + if (vmbus_proto_version != VERSION_WS2008) { + newchannel->is_dedicated_interrupt = + (offer->is_dedicated_interrupt != 0); + newchannel->sig_event = offer->connection_id; + } + + memcpy(&newchannel->offermsg, offer, + sizeof(struct vmbus_channel_offer_channel)); + newchannel->monitor_grp = (u8)offer->monitorid / 32; + newchannel->monitor_bit = (u8)offer->monitorid % 32; + + vmbus_process_offer(newchannel); +} + +/* + * vmbus_onoffer_rescind - Rescind offer handler. + * + * We queue a work item to process this offer synchronously + */ +static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_rescind_offer *rescind; + struct vmbus_channel *channel; + struct device *dev; + + rescind = (struct vmbus_channel_rescind_offer *)hdr; + + trace_vmbus_onoffer_rescind(rescind); + + /* + * The offer msg and the corresponding rescind msg + * from the host are guranteed to be ordered - + * offer comes in first and then the rescind. + * Since we process these events in work elements, + * and with preemption, we may end up processing + * the events out of order. Given that we handle these + * work elements on the same CPU, this is possible only + * in the case of preemption. In any case wait here + * until the offer processing has moved beyond the + * point where the channel is discoverable. + */ + + while (atomic_read(&vmbus_connection.offer_in_progress) != 0) { + /* + * We wait here until any channel offer is currently + * being processed. + */ + msleep(1); + } + + mutex_lock(&vmbus_connection.channel_mutex); + channel = relid2channel(rescind->child_relid); + mutex_unlock(&vmbus_connection.channel_mutex); + + if (channel == NULL) { + /* + * We failed in processing the offer message; + * we would have cleaned up the relid in that + * failure path. + */ + return; + } + + /* + * Before setting channel->rescind in vmbus_rescind_cleanup(), we + * should make sure the channel callback is not running any more. + */ + vmbus_reset_channel_cb(channel); + + /* + * Now wait for offer handling to complete. + */ + vmbus_rescind_cleanup(channel); + while (READ_ONCE(channel->probe_done) == false) { + /* + * We wait here until any channel offer is currently + * being processed. + */ + msleep(1); + } + + /* + * At this point, the rescind handling can proceed safely. + */ + + if (channel->device_obj) { + if (channel->chn_rescind_callback) { + channel->chn_rescind_callback(channel); + return; + } + /* + * We will have to unregister this device from the + * driver core. + */ + dev = get_device(&channel->device_obj->device); + if (dev) { + vmbus_device_unregister(channel->device_obj); + put_device(dev); + } + } else if (channel->primary_channel != NULL) { + /* + * Sub-channel is being rescinded. Following is the channel + * close sequence when initiated from the driveri (refer to + * vmbus_close() for details): + * 1. Close all sub-channels first + * 2. Then close the primary channel. + */ + mutex_lock(&vmbus_connection.channel_mutex); + if (channel->state == CHANNEL_OPEN_STATE) { + /* + * The channel is currently not open; + * it is safe for us to cleanup the channel. + */ + hv_process_channel_removal(rescind->child_relid); + } else { + complete(&channel->rescind_event); + } + mutex_unlock(&vmbus_connection.channel_mutex); + } +} + +void vmbus_hvsock_device_unregister(struct vmbus_channel *channel) +{ + BUG_ON(!is_hvsock_channel(channel)); + + /* We always get a rescind msg when a connection is closed. */ + while (!READ_ONCE(channel->probe_done) || !READ_ONCE(channel->rescind)) + msleep(1); + + vmbus_device_unregister(channel->device_obj); +} +EXPORT_SYMBOL_GPL(vmbus_hvsock_device_unregister); + + +/* + * vmbus_onoffers_delivered - + * This is invoked when all offers have been delivered. + * + * Nothing to do here. + */ +static void vmbus_onoffers_delivered( + struct vmbus_channel_message_header *hdr) +{ +} + +/* + * vmbus_onopen_result - Open result handler. + * + * This is invoked when we received a response to our channel open request. + * Find the matching request, copy the response and signal the requesting + * thread. + */ +static void vmbus_onopen_result(struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_open_result *result; + struct vmbus_channel_msginfo *msginfo; + struct vmbus_channel_message_header *requestheader; + struct vmbus_channel_open_channel *openmsg; + unsigned long flags; + + result = (struct vmbus_channel_open_result *)hdr; + + trace_vmbus_onopen_result(result); + + /* + * Find the open msg, copy the result and signal/unblock the wait event + */ + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + + list_for_each_entry(msginfo, &vmbus_connection.chn_msg_list, + msglistentry) { + requestheader = + (struct vmbus_channel_message_header *)msginfo->msg; + + if (requestheader->msgtype == CHANNELMSG_OPENCHANNEL) { + openmsg = + (struct vmbus_channel_open_channel *)msginfo->msg; + if (openmsg->child_relid == result->child_relid && + openmsg->openid == result->openid) { + memcpy(&msginfo->response.open_result, + result, + sizeof( + struct vmbus_channel_open_result)); + complete(&msginfo->waitevent); + break; + } + } + } + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); +} + +/* + * vmbus_ongpadl_created - GPADL created handler. + * + * This is invoked when we received a response to our gpadl create request. + * Find the matching request, copy the response and signal the requesting + * thread. + */ +static void vmbus_ongpadl_created(struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_gpadl_created *gpadlcreated; + struct vmbus_channel_msginfo *msginfo; + struct vmbus_channel_message_header *requestheader; + struct vmbus_channel_gpadl_header *gpadlheader; + unsigned long flags; + + gpadlcreated = (struct vmbus_channel_gpadl_created *)hdr; + + trace_vmbus_ongpadl_created(gpadlcreated); + + /* + * Find the establish msg, copy the result and signal/unblock the wait + * event + */ + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + + list_for_each_entry(msginfo, &vmbus_connection.chn_msg_list, + msglistentry) { + requestheader = + (struct vmbus_channel_message_header *)msginfo->msg; + + if (requestheader->msgtype == CHANNELMSG_GPADL_HEADER) { + gpadlheader = + (struct vmbus_channel_gpadl_header *)requestheader; + + if ((gpadlcreated->child_relid == + gpadlheader->child_relid) && + (gpadlcreated->gpadl == gpadlheader->gpadl)) { + memcpy(&msginfo->response.gpadl_created, + gpadlcreated, + sizeof( + struct vmbus_channel_gpadl_created)); + complete(&msginfo->waitevent); + break; + } + } + } + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); +} + +/* + * vmbus_ongpadl_torndown - GPADL torndown handler. + * + * This is invoked when we received a response to our gpadl teardown request. + * Find the matching request, copy the response and signal the requesting + * thread. + */ +static void vmbus_ongpadl_torndown( + struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_gpadl_torndown *gpadl_torndown; + struct vmbus_channel_msginfo *msginfo; + struct vmbus_channel_message_header *requestheader; + struct vmbus_channel_gpadl_teardown *gpadl_teardown; + unsigned long flags; + + gpadl_torndown = (struct vmbus_channel_gpadl_torndown *)hdr; + + trace_vmbus_ongpadl_torndown(gpadl_torndown); + + /* + * Find the open msg, copy the result and signal/unblock the wait event + */ + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + + list_for_each_entry(msginfo, &vmbus_connection.chn_msg_list, + msglistentry) { + requestheader = + (struct vmbus_channel_message_header *)msginfo->msg; + + if (requestheader->msgtype == CHANNELMSG_GPADL_TEARDOWN) { + gpadl_teardown = + (struct vmbus_channel_gpadl_teardown *)requestheader; + + if (gpadl_torndown->gpadl == gpadl_teardown->gpadl) { + memcpy(&msginfo->response.gpadl_torndown, + gpadl_torndown, + sizeof( + struct vmbus_channel_gpadl_torndown)); + complete(&msginfo->waitevent); + break; + } + } + } + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); +} + +/* + * vmbus_onversion_response - Version response handler + * + * This is invoked when we received a response to our initiate contact request. + * Find the matching request, copy the response and signal the requesting + * thread. + */ +static void vmbus_onversion_response( + struct vmbus_channel_message_header *hdr) +{ + struct vmbus_channel_msginfo *msginfo; + struct vmbus_channel_message_header *requestheader; + struct vmbus_channel_version_response *version_response; + unsigned long flags; + + version_response = (struct vmbus_channel_version_response *)hdr; + + trace_vmbus_onversion_response(version_response); + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + + list_for_each_entry(msginfo, &vmbus_connection.chn_msg_list, + msglistentry) { + requestheader = + (struct vmbus_channel_message_header *)msginfo->msg; + + if (requestheader->msgtype == + CHANNELMSG_INITIATE_CONTACT) { + memcpy(&msginfo->response.version_response, + version_response, + sizeof(struct vmbus_channel_version_response)); + complete(&msginfo->waitevent); + } + } + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); +} + +/* Channel message dispatch table */ +const struct vmbus_channel_message_table_entry +channel_message_table[CHANNELMSG_COUNT] = { + { CHANNELMSG_INVALID, 0, NULL }, + { CHANNELMSG_OFFERCHANNEL, 0, vmbus_onoffer }, + { CHANNELMSG_RESCIND_CHANNELOFFER, 0, vmbus_onoffer_rescind }, + { CHANNELMSG_REQUESTOFFERS, 0, NULL }, + { CHANNELMSG_ALLOFFERS_DELIVERED, 1, vmbus_onoffers_delivered }, + { CHANNELMSG_OPENCHANNEL, 0, NULL }, + { CHANNELMSG_OPENCHANNEL_RESULT, 1, vmbus_onopen_result }, + { CHANNELMSG_CLOSECHANNEL, 0, NULL }, + { CHANNELMSG_GPADL_HEADER, 0, NULL }, + { CHANNELMSG_GPADL_BODY, 0, NULL }, + { CHANNELMSG_GPADL_CREATED, 1, vmbus_ongpadl_created }, + { CHANNELMSG_GPADL_TEARDOWN, 0, NULL }, + { CHANNELMSG_GPADL_TORNDOWN, 1, vmbus_ongpadl_torndown }, + { CHANNELMSG_RELID_RELEASED, 0, NULL }, + { CHANNELMSG_INITIATE_CONTACT, 0, NULL }, + { CHANNELMSG_VERSION_RESPONSE, 1, vmbus_onversion_response }, + { CHANNELMSG_UNLOAD, 0, NULL }, + { CHANNELMSG_UNLOAD_RESPONSE, 1, vmbus_unload_response }, + { CHANNELMSG_18, 0, NULL }, + { CHANNELMSG_19, 0, NULL }, + { CHANNELMSG_20, 0, NULL }, + { CHANNELMSG_TL_CONNECT_REQUEST, 0, NULL }, + { CHANNELMSG_22, 0, NULL }, + { CHANNELMSG_TL_CONNECT_RESULT, 0, NULL }, +}; + +/* + * vmbus_onmessage - Handler for channel protocol messages. + * + * This is invoked in the vmbus worker thread context. + */ +void vmbus_onmessage(void *context) +{ + struct hv_message *msg = context; + struct vmbus_channel_message_header *hdr; + + hdr = (struct vmbus_channel_message_header *)msg->u.payload; + + trace_vmbus_on_message(hdr); + + /* + * vmbus_on_msg_dpc() makes sure the hdr->msgtype here can not go + * out of bound and the message_handler pointer can not be NULL. + */ + channel_message_table[hdr->msgtype].message_handler(hdr); +} + +/* + * vmbus_request_offers - Send a request to get all our pending offers. + */ +int vmbus_request_offers(void) +{ + struct vmbus_channel_message_header *msg; + struct vmbus_channel_msginfo *msginfo; + int ret; + + msginfo = kmalloc(sizeof(*msginfo) + + sizeof(struct vmbus_channel_message_header), + GFP_KERNEL); + if (!msginfo) + return -ENOMEM; + + msg = (struct vmbus_channel_message_header *)msginfo->msg; + + msg->msgtype = CHANNELMSG_REQUESTOFFERS; + + ret = vmbus_post_msg(msg, sizeof(struct vmbus_channel_message_header), + true); + + trace_vmbus_request_offers(ret); + + if (ret != 0) { + pr_err("Unable to request offers - %d\n", ret); + + goto cleanup; + } + +cleanup: + kfree(msginfo); + + return ret; +} + +/* + * Retrieve the (sub) channel on which to send an outgoing request. + * When a primary channel has multiple sub-channels, we try to + * distribute the load equally amongst all available channels. + */ +struct vmbus_channel *vmbus_get_outgoing_channel(struct vmbus_channel *primary) +{ + struct list_head *cur, *tmp; + int cur_cpu; + struct vmbus_channel *cur_channel; + struct vmbus_channel *outgoing_channel = primary; + int next_channel; + int i = 1; + + if (list_empty(&primary->sc_list)) + return outgoing_channel; + + next_channel = primary->next_oc++; + + if (next_channel > (primary->num_sc)) { + primary->next_oc = 0; + return outgoing_channel; + } + + cur_cpu = hv_cpu_number_to_vp_number(smp_processor_id()); + list_for_each_safe(cur, tmp, &primary->sc_list) { + cur_channel = list_entry(cur, struct vmbus_channel, sc_list); + if (cur_channel->state != CHANNEL_OPENED_STATE) + continue; + + if (cur_channel->target_vp == cur_cpu) + return cur_channel; + + if (i == next_channel) + return cur_channel; + + i++; + } + + return outgoing_channel; +} +EXPORT_SYMBOL_GPL(vmbus_get_outgoing_channel); + +static void invoke_sc_cb(struct vmbus_channel *primary_channel) +{ + struct list_head *cur, *tmp; + struct vmbus_channel *cur_channel; + + if (primary_channel->sc_creation_callback == NULL) + return; + + list_for_each_safe(cur, tmp, &primary_channel->sc_list) { + cur_channel = list_entry(cur, struct vmbus_channel, sc_list); + + primary_channel->sc_creation_callback(cur_channel); + } +} + +void vmbus_set_sc_create_callback(struct vmbus_channel *primary_channel, + void (*sc_cr_cb)(struct vmbus_channel *new_sc)) +{ + primary_channel->sc_creation_callback = sc_cr_cb; +} +EXPORT_SYMBOL_GPL(vmbus_set_sc_create_callback); + +bool vmbus_are_subchannels_present(struct vmbus_channel *primary) +{ + bool ret; + + ret = !list_empty(&primary->sc_list); + + if (ret) { + /* + * Invoke the callback on sub-channel creation. + * This will present a uniform interface to the + * clients. + */ + invoke_sc_cb(primary); + } + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_are_subchannels_present); + +void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, + void (*chn_rescind_cb)(struct vmbus_channel *)) +{ + channel->chn_rescind_callback = chn_rescind_cb; +} +EXPORT_SYMBOL_GPL(vmbus_set_chn_rescind_callback); diff --git a/drivers/hv/connection.c b/drivers/hv/connection.c new file mode 100644 index 000000000..4fe117b76 --- /dev/null +++ b/drivers/hv/connection.c @@ -0,0 +1,480 @@ +/* + * + * 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> + * + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/sched.h> +#include <linux/wait.h> +#include <linux/delay.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/vmalloc.h> +#include <linux/hyperv.h> +#include <linux/export.h> +#include <asm/mshyperv.h> + +#include "hyperv_vmbus.h" + + +struct vmbus_connection vmbus_connection = { + .conn_state = DISCONNECTED, + .next_gpadl_handle = ATOMIC_INIT(0xE1E10), +}; +EXPORT_SYMBOL_GPL(vmbus_connection); + +/* + * Negotiated protocol version with the host. + */ +__u32 vmbus_proto_version; +EXPORT_SYMBOL_GPL(vmbus_proto_version); + +static __u32 vmbus_get_next_version(__u32 current_version) +{ + switch (current_version) { + case (VERSION_WIN7): + return VERSION_WS2008; + + case (VERSION_WIN8): + return VERSION_WIN7; + + case (VERSION_WIN8_1): + return VERSION_WIN8; + + case (VERSION_WIN10): + return VERSION_WIN8_1; + + case (VERSION_WIN10_V5): + return VERSION_WIN10; + + case (VERSION_WS2008): + default: + return VERSION_INVAL; + } +} + +static int vmbus_negotiate_version(struct vmbus_channel_msginfo *msginfo, + __u32 version) +{ + int ret = 0; + unsigned int cur_cpu; + struct vmbus_channel_initiate_contact *msg; + unsigned long flags; + + init_completion(&msginfo->waitevent); + + msg = (struct vmbus_channel_initiate_contact *)msginfo->msg; + + memset(msg, 0, sizeof(*msg)); + msg->header.msgtype = CHANNELMSG_INITIATE_CONTACT; + msg->vmbus_version_requested = version; + + /* + * VMBus protocol 5.0 (VERSION_WIN10_V5) requires that we must use + * VMBUS_MESSAGE_CONNECTION_ID_4 for the Initiate Contact Message, + * and for subsequent messages, we must use the Message Connection ID + * field in the host-returned Version Response Message. And, with + * VERSION_WIN10_V5, we don't use msg->interrupt_page, but we tell + * the host explicitly that we still use VMBUS_MESSAGE_SINT(2) for + * compatibility. + * + * On old hosts, we should always use VMBUS_MESSAGE_CONNECTION_ID (1). + */ + if (version >= VERSION_WIN10_V5) { + msg->msg_sint = VMBUS_MESSAGE_SINT; + vmbus_connection.msg_conn_id = VMBUS_MESSAGE_CONNECTION_ID_4; + } else { + msg->interrupt_page = virt_to_phys(vmbus_connection.int_page); + vmbus_connection.msg_conn_id = VMBUS_MESSAGE_CONNECTION_ID; + } + + msg->monitor_page1 = virt_to_phys(vmbus_connection.monitor_pages[0]); + msg->monitor_page2 = virt_to_phys(vmbus_connection.monitor_pages[1]); + /* + * We want all channel messages to be delivered on CPU 0. + * This has been the behavior pre-win8. This is not + * perf issue and having all channel messages delivered on CPU 0 + * would be ok. + * For post win8 hosts, we support receiving channel messagges on + * all the CPUs. This is needed for kexec to work correctly where + * the CPU attempting to connect may not be CPU 0. + */ + if (version >= VERSION_WIN8_1) { + cur_cpu = get_cpu(); + msg->target_vcpu = hv_cpu_number_to_vp_number(cur_cpu); + vmbus_connection.connect_cpu = cur_cpu; + put_cpu(); + } else { + msg->target_vcpu = 0; + vmbus_connection.connect_cpu = 0; + } + + /* + * Add to list before we send the request since we may + * receive the response before returning from this routine + */ + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_add_tail(&msginfo->msglistentry, + &vmbus_connection.chn_msg_list); + + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + ret = vmbus_post_msg(msg, + sizeof(struct vmbus_channel_initiate_contact), + true); + + trace_vmbus_negotiate_version(msg, ret); + + if (ret != 0) { + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&msginfo->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, + flags); + return ret; + } + + /* Wait for the connection response */ + wait_for_completion(&msginfo->waitevent); + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&msginfo->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + /* Check if successful */ + if (msginfo->response.version_response.version_supported) { + vmbus_connection.conn_state = CONNECTED; + + if (version >= VERSION_WIN10_V5) + vmbus_connection.msg_conn_id = + msginfo->response.version_response.msg_conn_id; + } else { + return -ECONNREFUSED; + } + + return ret; +} + +/* + * vmbus_connect - Sends a connect request on the partition service connection + */ +int vmbus_connect(void) +{ + int ret = 0; + struct vmbus_channel_msginfo *msginfo = NULL; + __u32 version; + + /* Initialize the vmbus connection */ + vmbus_connection.conn_state = CONNECTING; + vmbus_connection.work_queue = create_workqueue("hv_vmbus_con"); + if (!vmbus_connection.work_queue) { + ret = -ENOMEM; + goto cleanup; + } + + vmbus_connection.handle_primary_chan_wq = + create_workqueue("hv_pri_chan"); + if (!vmbus_connection.handle_primary_chan_wq) { + ret = -ENOMEM; + goto cleanup; + } + + vmbus_connection.handle_sub_chan_wq = + create_workqueue("hv_sub_chan"); + if (!vmbus_connection.handle_sub_chan_wq) { + ret = -ENOMEM; + goto cleanup; + } + + INIT_LIST_HEAD(&vmbus_connection.chn_msg_list); + spin_lock_init(&vmbus_connection.channelmsg_lock); + + INIT_LIST_HEAD(&vmbus_connection.chn_list); + mutex_init(&vmbus_connection.channel_mutex); + + /* + * Setup the vmbus event connection for channel interrupt + * abstraction stuff + */ + vmbus_connection.int_page = + (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, 0); + if (vmbus_connection.int_page == NULL) { + ret = -ENOMEM; + goto cleanup; + } + + vmbus_connection.recv_int_page = vmbus_connection.int_page; + vmbus_connection.send_int_page = + (void *)((unsigned long)vmbus_connection.int_page + + (PAGE_SIZE >> 1)); + + /* + * Setup the monitor notification facility. The 1st page for + * parent->child and the 2nd page for child->parent + */ + vmbus_connection.monitor_pages[0] = (void *)__get_free_pages((GFP_KERNEL|__GFP_ZERO), 0); + vmbus_connection.monitor_pages[1] = (void *)__get_free_pages((GFP_KERNEL|__GFP_ZERO), 0); + if ((vmbus_connection.monitor_pages[0] == NULL) || + (vmbus_connection.monitor_pages[1] == NULL)) { + ret = -ENOMEM; + goto cleanup; + } + + msginfo = kzalloc(sizeof(*msginfo) + + sizeof(struct vmbus_channel_initiate_contact), + GFP_KERNEL); + if (msginfo == NULL) { + ret = -ENOMEM; + goto cleanup; + } + + /* + * Negotiate a compatible VMBUS version number with the + * host. We start with the highest number we can support + * and work our way down until we negotiate a compatible + * version. + */ + + version = VERSION_CURRENT; + + do { + ret = vmbus_negotiate_version(msginfo, version); + if (ret == -ETIMEDOUT) + goto cleanup; + + if (vmbus_connection.conn_state == CONNECTED) + break; + + version = vmbus_get_next_version(version); + } while (version != VERSION_INVAL); + + if (version == VERSION_INVAL) + goto cleanup; + + vmbus_proto_version = version; + pr_info("Vmbus version:%d.%d\n", + version >> 16, version & 0xFFFF); + + kfree(msginfo); + return 0; + +cleanup: + pr_err("Unable to connect to host\n"); + + vmbus_connection.conn_state = DISCONNECTED; + vmbus_disconnect(); + + kfree(msginfo); + + return ret; +} + +void vmbus_disconnect(void) +{ + /* + * First send the unload request to the host. + */ + vmbus_initiate_unload(false); + + if (vmbus_connection.handle_sub_chan_wq) + destroy_workqueue(vmbus_connection.handle_sub_chan_wq); + + if (vmbus_connection.handle_primary_chan_wq) + destroy_workqueue(vmbus_connection.handle_primary_chan_wq); + + if (vmbus_connection.work_queue) + destroy_workqueue(vmbus_connection.work_queue); + + if (vmbus_connection.int_page) { + free_pages((unsigned long)vmbus_connection.int_page, 0); + vmbus_connection.int_page = NULL; + } + + free_pages((unsigned long)vmbus_connection.monitor_pages[0], 0); + free_pages((unsigned long)vmbus_connection.monitor_pages[1], 0); + vmbus_connection.monitor_pages[0] = NULL; + vmbus_connection.monitor_pages[1] = NULL; +} + +/* + * relid2channel - Get the channel object given its + * child relative id (ie channel id) + */ +struct vmbus_channel *relid2channel(u32 relid) +{ + struct vmbus_channel *channel; + struct vmbus_channel *found_channel = NULL; + struct list_head *cur, *tmp; + struct vmbus_channel *cur_sc; + + BUG_ON(!mutex_is_locked(&vmbus_connection.channel_mutex)); + + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + if (channel->offermsg.child_relid == relid) { + found_channel = channel; + break; + } else if (!list_empty(&channel->sc_list)) { + /* + * Deal with sub-channels. + */ + list_for_each_safe(cur, tmp, &channel->sc_list) { + cur_sc = list_entry(cur, struct vmbus_channel, + sc_list); + if (cur_sc->offermsg.child_relid == relid) { + found_channel = cur_sc; + break; + } + } + } + } + + return found_channel; +} + +/* + * vmbus_on_event - Process a channel event notification + * + * For batched channels (default) optimize host to guest signaling + * by ensuring: + * 1. While reading the channel, we disable interrupts from host. + * 2. Ensure that we process all posted messages from the host + * before returning from this callback. + * 3. Once we return, enable signaling from the host. Once this + * state is set we check to see if additional packets are + * available to read. In this case we repeat the process. + * If this tasklet has been running for a long time + * then reschedule ourselves. + */ +void vmbus_on_event(unsigned long data) +{ + struct vmbus_channel *channel = (void *) data; + unsigned long time_limit = jiffies + 2; + + trace_vmbus_on_event(channel); + + do { + void (*callback_fn)(void *); + + /* A channel once created is persistent even when + * there is no driver handling the device. An + * unloading driver sets the onchannel_callback to NULL. + */ + callback_fn = READ_ONCE(channel->onchannel_callback); + if (unlikely(callback_fn == NULL)) + return; + + (*callback_fn)(channel->channel_callback_context); + + if (channel->callback_mode != HV_CALL_BATCHED) + return; + + if (likely(hv_end_read(&channel->inbound) == 0)) + return; + + hv_begin_read(&channel->inbound); + } while (likely(time_before(jiffies, time_limit))); + + /* The time limit (2 jiffies) has been reached */ + tasklet_schedule(&channel->callback_event); +} + +/* + * vmbus_post_msg - Send a msg on the vmbus's message connection + */ +int vmbus_post_msg(void *buffer, size_t buflen, bool can_sleep) +{ + struct vmbus_channel_message_header *hdr; + union hv_connection_id conn_id; + int ret = 0; + int retries = 0; + u32 usec = 1; + + conn_id.asu32 = 0; + conn_id.u.id = vmbus_connection.msg_conn_id; + + /* + * hv_post_message() can have transient failures because of + * insufficient resources. Retry the operation a couple of + * times before giving up. + */ + while (retries < 100) { + ret = hv_post_message(conn_id, 1, buffer, buflen); + + switch (ret) { + case HV_STATUS_INVALID_CONNECTION_ID: + /* + * See vmbus_negotiate_version(): VMBus protocol 5.0 + * requires that we must use + * VMBUS_MESSAGE_CONNECTION_ID_4 for the Initiate + * Contact message, but on old hosts that only + * support VMBus protocol 4.0 or lower, here we get + * HV_STATUS_INVALID_CONNECTION_ID and we should + * return an error immediately without retrying. + */ + hdr = buffer; + if (hdr->msgtype == CHANNELMSG_INITIATE_CONTACT) + return -EINVAL; + /* + * We could get this if we send messages too + * frequently. + */ + ret = -EAGAIN; + break; + case HV_STATUS_INSUFFICIENT_MEMORY: + case HV_STATUS_INSUFFICIENT_BUFFERS: + ret = -ENOBUFS; + break; + case HV_STATUS_SUCCESS: + return ret; + default: + pr_err("hv_post_msg() failed; error code:%d\n", ret); + return -EINVAL; + } + + retries++; + if (can_sleep && usec > 1000) + msleep(usec / 1000); + else if (usec < MAX_UDELAY_MS * 1000) + udelay(usec); + else + mdelay(usec / 1000); + + if (retries < 22) + usec *= 2; + } + return ret; +} + +/* + * vmbus_set_event - Send an event notification to the parent + */ +void vmbus_set_event(struct vmbus_channel *channel) +{ + u32 child_relid = channel->offermsg.child_relid; + + if (!channel->is_dedicated_interrupt) + vmbus_send_interrupt(child_relid); + + ++channel->sig_events; + + hv_do_fast_hypercall8(HVCALL_SIGNAL_EVENT, channel->sig_event); +} +EXPORT_SYMBOL_GPL(vmbus_set_event); diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c new file mode 100644 index 000000000..08bcefca3 --- /dev/null +++ b/drivers/hv/hv.c @@ -0,0 +1,442 @@ +/* + * 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> + * + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/mm.h> +#include <linux/slab.h> +#include <linux/vmalloc.h> +#include <linux/hyperv.h> +#include <linux/version.h> +#include <linux/random.h> +#include <linux/clockchips.h> +#include <asm/mshyperv.h> +#include "hyperv_vmbus.h" + +/* The one and only */ +struct hv_context hv_context = { + .synic_initialized = false, +}; + +/* + * If false, we're using the old mechanism for stimer0 interrupts + * where it sends a VMbus message when it expires. The old + * mechanism is used when running on older versions of Hyper-V + * that don't support Direct Mode. While Hyper-V provides + * four stimer's per CPU, Linux uses only stimer0. + */ +static bool direct_mode_enabled; +static int stimer0_irq; +static int stimer0_vector; + +#define HV_TIMER_FREQUENCY (10 * 1000 * 1000) /* 100ns period */ +#define HV_MAX_MAX_DELTA_TICKS 0xffffffff +#define HV_MIN_DELTA_TICKS 1 + +/* + * hv_init - Main initialization routine. + * + * This routine must be called before any other routines in here are called + */ +int hv_init(void) +{ + hv_context.cpu_context = alloc_percpu(struct hv_per_cpu_context); + if (!hv_context.cpu_context) + return -ENOMEM; + + direct_mode_enabled = ms_hyperv.misc_features & + HV_STIMER_DIRECT_MODE_AVAILABLE; + return 0; +} + +/* + * hv_post_message - Post a message using the hypervisor message IPC. + * + * This involves a hypercall. + */ +int hv_post_message(union hv_connection_id connection_id, + enum hv_message_type message_type, + void *payload, size_t payload_size) +{ + struct hv_input_post_message *aligned_msg; + struct hv_per_cpu_context *hv_cpu; + u64 status; + + if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT) + return -EMSGSIZE; + + hv_cpu = get_cpu_ptr(hv_context.cpu_context); + aligned_msg = hv_cpu->post_msg_page; + aligned_msg->connectionid = connection_id; + aligned_msg->reserved = 0; + aligned_msg->message_type = message_type; + aligned_msg->payload_size = payload_size; + memcpy((void *)aligned_msg->payload, payload, payload_size); + + status = hv_do_hypercall(HVCALL_POST_MESSAGE, aligned_msg, NULL); + + /* Preemption must remain disabled until after the hypercall + * so some other thread can't get scheduled onto this cpu and + * corrupt the per-cpu post_msg_page + */ + put_cpu_ptr(hv_cpu); + + return status & 0xFFFF; +} + +/* + * ISR for when stimer0 is operating in Direct Mode. Direct Mode + * does not use VMbus or any VMbus messages, so process here and not + * in the VMbus driver code. + */ + +static void hv_stimer0_isr(void) +{ + struct hv_per_cpu_context *hv_cpu; + + hv_cpu = this_cpu_ptr(hv_context.cpu_context); + hv_cpu->clk_evt->event_handler(hv_cpu->clk_evt); + add_interrupt_randomness(stimer0_vector); +} + +static int hv_ce_set_next_event(unsigned long delta, + struct clock_event_device *evt) +{ + u64 current_tick; + + WARN_ON(!clockevent_state_oneshot(evt)); + + current_tick = hyperv_cs->read(NULL); + current_tick += delta; + hv_init_timer(0, current_tick); + return 0; +} + +static int hv_ce_shutdown(struct clock_event_device *evt) +{ + hv_init_timer(0, 0); + hv_init_timer_config(0, 0); + if (direct_mode_enabled) + hv_disable_stimer0_percpu_irq(stimer0_irq); + + return 0; +} + +static int hv_ce_set_oneshot(struct clock_event_device *evt) +{ + union hv_timer_config timer_cfg; + + timer_cfg.as_uint64 = 0; + timer_cfg.enable = 1; + timer_cfg.auto_enable = 1; + if (direct_mode_enabled) { + /* + * When it expires, the timer will directly interrupt + * on the specified hardware vector/IRQ. + */ + timer_cfg.direct_mode = 1; + timer_cfg.apic_vector = stimer0_vector; + hv_enable_stimer0_percpu_irq(stimer0_irq); + } else { + /* + * When it expires, the timer will generate a VMbus message, + * to be handled by the normal VMbus interrupt handler. + */ + timer_cfg.direct_mode = 0; + timer_cfg.sintx = VMBUS_MESSAGE_SINT; + } + hv_init_timer_config(0, timer_cfg.as_uint64); + return 0; +} + +static void hv_init_clockevent_device(struct clock_event_device *dev, int cpu) +{ + dev->name = "Hyper-V clockevent"; + dev->features = CLOCK_EVT_FEAT_ONESHOT; + dev->cpumask = cpumask_of(cpu); + dev->rating = 1000; + /* + * Avoid settint dev->owner = THIS_MODULE deliberately as doing so will + * result in clockevents_config_and_register() taking additional + * references to the hv_vmbus module making it impossible to unload. + */ + + dev->set_state_shutdown = hv_ce_shutdown; + dev->set_state_oneshot = hv_ce_set_oneshot; + dev->set_next_event = hv_ce_set_next_event; +} + + +int hv_synic_alloc(void) +{ + int cpu; + struct hv_per_cpu_context *hv_cpu; + + /* + * First, zero all per-cpu memory areas so hv_synic_free() can + * detect what memory has been allocated and cleanup properly + * after any failures. + */ + for_each_present_cpu(cpu) { + hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu); + memset(hv_cpu, 0, sizeof(*hv_cpu)); + } + + hv_context.hv_numa_map = kcalloc(nr_node_ids, sizeof(struct cpumask), + GFP_KERNEL); + if (hv_context.hv_numa_map == NULL) { + pr_err("Unable to allocate NUMA map\n"); + goto err; + } + + for_each_present_cpu(cpu) { + hv_cpu = per_cpu_ptr(hv_context.cpu_context, cpu); + + tasklet_init(&hv_cpu->msg_dpc, + vmbus_on_msg_dpc, (unsigned long) hv_cpu); + + hv_cpu->clk_evt = kzalloc(sizeof(struct clock_event_device), + GFP_KERNEL); + if (hv_cpu->clk_evt == NULL) { + pr_err("Unable to allocate clock event device\n"); + goto err; + } + hv_init_clockevent_device(hv_cpu->clk_evt, cpu); + + hv_cpu->synic_message_page = + (void *)get_zeroed_page(GFP_ATOMIC); + if (hv_cpu->synic_message_page == NULL) { + pr_err("Unable to allocate SYNIC message page\n"); + goto err; + } + + hv_cpu->synic_event_page = (void *)get_zeroed_page(GFP_ATOMIC); + if (hv_cpu->synic_event_page == NULL) { + pr_err("Unable to allocate SYNIC event page\n"); + goto err; + } + + hv_cpu->post_msg_page = (void *)get_zeroed_page(GFP_ATOMIC); + if (hv_cpu->post_msg_page == NULL) { + pr_err("Unable to allocate post msg page\n"); + goto err; + } + + INIT_LIST_HEAD(&hv_cpu->chan_list); + } + + if (direct_mode_enabled && + hv_setup_stimer0_irq(&stimer0_irq, &stimer0_vector, + hv_stimer0_isr)) + goto err; + + return 0; +err: + /* + * Any memory allocations that succeeded will be freed when + * the caller cleans up by calling hv_synic_free() + */ + return -ENOMEM; +} + + +void hv_synic_free(void) +{ + int cpu; + + for_each_present_cpu(cpu) { + struct hv_per_cpu_context *hv_cpu + = per_cpu_ptr(hv_context.cpu_context, cpu); + + kfree(hv_cpu->clk_evt); + free_page((unsigned long)hv_cpu->synic_event_page); + free_page((unsigned long)hv_cpu->synic_message_page); + free_page((unsigned long)hv_cpu->post_msg_page); + } + + kfree(hv_context.hv_numa_map); +} + +/* + * hv_synic_init - Initialize the Synthetic Interrupt Controller. + * + * If it is already initialized by another entity (ie x2v shim), we need to + * retrieve the initialized message and event pages. Otherwise, we create and + * initialize the message and event pages. + */ +int hv_synic_init(unsigned int cpu) +{ + struct hv_per_cpu_context *hv_cpu + = per_cpu_ptr(hv_context.cpu_context, cpu); + union hv_synic_simp simp; + union hv_synic_siefp siefp; + union hv_synic_sint shared_sint; + union hv_synic_scontrol sctrl; + + /* Setup the Synic's message page */ + hv_get_simp(simp.as_uint64); + simp.simp_enabled = 1; + simp.base_simp_gpa = virt_to_phys(hv_cpu->synic_message_page) + >> PAGE_SHIFT; + + hv_set_simp(simp.as_uint64); + + /* Setup the Synic's event page */ + hv_get_siefp(siefp.as_uint64); + siefp.siefp_enabled = 1; + siefp.base_siefp_gpa = virt_to_phys(hv_cpu->synic_event_page) + >> PAGE_SHIFT; + + hv_set_siefp(siefp.as_uint64); + + /* Setup the shared SINT. */ + hv_get_synint_state(VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + shared_sint.vector = HYPERVISOR_CALLBACK_VECTOR; + shared_sint.masked = false; + if (ms_hyperv.hints & HV_DEPRECATING_AEOI_RECOMMENDED) + shared_sint.auto_eoi = false; + else + shared_sint.auto_eoi = true; + + hv_set_synint_state(VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + /* Enable the global synic bit */ + hv_get_synic_state(sctrl.as_uint64); + sctrl.enable = 1; + + hv_set_synic_state(sctrl.as_uint64); + + hv_context.synic_initialized = true; + + /* + * Register the per-cpu clockevent source. + */ + if (ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE) + clockevents_config_and_register(hv_cpu->clk_evt, + HV_TIMER_FREQUENCY, + HV_MIN_DELTA_TICKS, + HV_MAX_MAX_DELTA_TICKS); + return 0; +} + +/* + * hv_synic_clockevents_cleanup - Cleanup clockevent devices + */ +void hv_synic_clockevents_cleanup(void) +{ + int cpu; + + if (!(ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE)) + return; + + if (direct_mode_enabled) + hv_remove_stimer0_irq(stimer0_irq); + + for_each_present_cpu(cpu) { + struct hv_per_cpu_context *hv_cpu + = per_cpu_ptr(hv_context.cpu_context, cpu); + + clockevents_unbind_device(hv_cpu->clk_evt, cpu); + } +} + +/* + * hv_synic_cleanup - Cleanup routine for hv_synic_init(). + */ +int hv_synic_cleanup(unsigned int cpu) +{ + union hv_synic_sint shared_sint; + union hv_synic_simp simp; + union hv_synic_siefp siefp; + union hv_synic_scontrol sctrl; + struct vmbus_channel *channel, *sc; + bool channel_found = false; + unsigned long flags; + + if (!hv_context.synic_initialized) + return -EFAULT; + + /* + * Search for channels which are bound to the CPU we're about to + * cleanup. In case we find one and vmbus is still connected we need to + * fail, this will effectively prevent CPU offlining. There is no way + * we can re-bind channels to different CPUs for now. + */ + mutex_lock(&vmbus_connection.channel_mutex); + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + if (channel->target_cpu == cpu) { + channel_found = true; + break; + } + spin_lock_irqsave(&channel->lock, flags); + list_for_each_entry(sc, &channel->sc_list, sc_list) { + if (sc->target_cpu == cpu) { + channel_found = true; + break; + } + } + spin_unlock_irqrestore(&channel->lock, flags); + if (channel_found) + break; + } + mutex_unlock(&vmbus_connection.channel_mutex); + + if (channel_found && vmbus_connection.conn_state == CONNECTED) + return -EBUSY; + + /* Turn off clockevent device */ + if (ms_hyperv.features & HV_MSR_SYNTIMER_AVAILABLE) { + struct hv_per_cpu_context *hv_cpu + = this_cpu_ptr(hv_context.cpu_context); + + clockevents_unbind_device(hv_cpu->clk_evt, cpu); + hv_ce_shutdown(hv_cpu->clk_evt); + } + + hv_get_synint_state(VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + shared_sint.masked = 1; + + /* Need to correctly cleanup in the case of SMP!!! */ + /* Disable the interrupt */ + hv_set_synint_state(VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + hv_get_simp(simp.as_uint64); + simp.simp_enabled = 0; + simp.base_simp_gpa = 0; + + hv_set_simp(simp.as_uint64); + + hv_get_siefp(siefp.as_uint64); + siefp.siefp_enabled = 0; + siefp.base_siefp_gpa = 0; + + hv_set_siefp(siefp.as_uint64); + + /* Disable the global synic bit */ + hv_get_synic_state(sctrl.as_uint64); + sctrl.enable = 0; + hv_set_synic_state(sctrl.as_uint64); + + return 0; +} diff --git a/drivers/hv/hv_balloon.c b/drivers/hv/hv_balloon.c new file mode 100644 index 000000000..d442a8d23 --- /dev/null +++ b/drivers/hv/hv_balloon.c @@ -0,0 +1,1777 @@ +/* + * Copyright (c) 2012, Microsoft Corporation. + * + * Author: + * K. Y. Srinivasan <kys@microsoft.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/jiffies.h> +#include <linux/mman.h> +#include <linux/delay.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/kthread.h> +#include <linux/completion.h> +#include <linux/memory_hotplug.h> +#include <linux/memory.h> +#include <linux/notifier.h> +#include <linux/percpu_counter.h> + +#include <linux/hyperv.h> + +#define CREATE_TRACE_POINTS +#include "hv_trace_balloon.h" + +/* + * We begin with definitions supporting the Dynamic Memory protocol + * with the host. + * + * Begin protocol definitions. + */ + + + +/* + * Protocol versions. The low word is the minor version, the high word the major + * version. + * + * History: + * Initial version 1.0 + * Changed to 0.1 on 2009/03/25 + * Changes to 0.2 on 2009/05/14 + * Changes to 0.3 on 2009/12/03 + * Changed to 1.0 on 2011/04/05 + */ + +#define DYNMEM_MAKE_VERSION(Major, Minor) ((__u32)(((Major) << 16) | (Minor))) +#define DYNMEM_MAJOR_VERSION(Version) ((__u32)(Version) >> 16) +#define DYNMEM_MINOR_VERSION(Version) ((__u32)(Version) & 0xff) + +enum { + DYNMEM_PROTOCOL_VERSION_1 = DYNMEM_MAKE_VERSION(0, 3), + DYNMEM_PROTOCOL_VERSION_2 = DYNMEM_MAKE_VERSION(1, 0), + DYNMEM_PROTOCOL_VERSION_3 = DYNMEM_MAKE_VERSION(2, 0), + + DYNMEM_PROTOCOL_VERSION_WIN7 = DYNMEM_PROTOCOL_VERSION_1, + DYNMEM_PROTOCOL_VERSION_WIN8 = DYNMEM_PROTOCOL_VERSION_2, + DYNMEM_PROTOCOL_VERSION_WIN10 = DYNMEM_PROTOCOL_VERSION_3, + + DYNMEM_PROTOCOL_VERSION_CURRENT = DYNMEM_PROTOCOL_VERSION_WIN10 +}; + + + +/* + * Message Types + */ + +enum dm_message_type { + /* + * Version 0.3 + */ + DM_ERROR = 0, + DM_VERSION_REQUEST = 1, + DM_VERSION_RESPONSE = 2, + DM_CAPABILITIES_REPORT = 3, + DM_CAPABILITIES_RESPONSE = 4, + DM_STATUS_REPORT = 5, + DM_BALLOON_REQUEST = 6, + DM_BALLOON_RESPONSE = 7, + DM_UNBALLOON_REQUEST = 8, + DM_UNBALLOON_RESPONSE = 9, + DM_MEM_HOT_ADD_REQUEST = 10, + DM_MEM_HOT_ADD_RESPONSE = 11, + DM_VERSION_03_MAX = 11, + /* + * Version 1.0. + */ + DM_INFO_MESSAGE = 12, + DM_VERSION_1_MAX = 12 +}; + + +/* + * Structures defining the dynamic memory management + * protocol. + */ + +union dm_version { + struct { + __u16 minor_version; + __u16 major_version; + }; + __u32 version; +} __packed; + + +union dm_caps { + struct { + __u64 balloon:1; + __u64 hot_add:1; + /* + * To support guests that may have alignment + * limitations on hot-add, the guest can specify + * its alignment requirements; a value of n + * represents an alignment of 2^n in mega bytes. + */ + __u64 hot_add_alignment:4; + __u64 reservedz:58; + } cap_bits; + __u64 caps; +} __packed; + +union dm_mem_page_range { + struct { + /* + * The PFN number of the first page in the range. + * 40 bits is the architectural limit of a PFN + * number for AMD64. + */ + __u64 start_page:40; + /* + * The number of pages in the range. + */ + __u64 page_cnt:24; + } finfo; + __u64 page_range; +} __packed; + + + +/* + * The header for all dynamic memory messages: + * + * type: Type of the message. + * size: Size of the message in bytes; including the header. + * trans_id: The guest is responsible for manufacturing this ID. + */ + +struct dm_header { + __u16 type; + __u16 size; + __u32 trans_id; +} __packed; + +/* + * A generic message format for dynamic memory. + * Specific message formats are defined later in the file. + */ + +struct dm_message { + struct dm_header hdr; + __u8 data[]; /* enclosed message */ +} __packed; + + +/* + * Specific message types supporting the dynamic memory protocol. + */ + +/* + * Version negotiation message. Sent from the guest to the host. + * The guest is free to try different versions until the host + * accepts the version. + * + * dm_version: The protocol version requested. + * is_last_attempt: If TRUE, this is the last version guest will request. + * reservedz: Reserved field, set to zero. + */ + +struct dm_version_request { + struct dm_header hdr; + union dm_version version; + __u32 is_last_attempt:1; + __u32 reservedz:31; +} __packed; + +/* + * Version response message; Host to Guest and indicates + * if the host has accepted the version sent by the guest. + * + * is_accepted: If TRUE, host has accepted the version and the guest + * should proceed to the next stage of the protocol. FALSE indicates that + * guest should re-try with a different version. + * + * reservedz: Reserved field, set to zero. + */ + +struct dm_version_response { + struct dm_header hdr; + __u64 is_accepted:1; + __u64 reservedz:63; +} __packed; + +/* + * Message reporting capabilities. This is sent from the guest to the + * host. + */ + +struct dm_capabilities { + struct dm_header hdr; + union dm_caps caps; + __u64 min_page_cnt; + __u64 max_page_number; +} __packed; + +/* + * Response to the capabilities message. This is sent from the host to the + * guest. This message notifies if the host has accepted the guest's + * capabilities. If the host has not accepted, the guest must shutdown + * the service. + * + * is_accepted: Indicates if the host has accepted guest's capabilities. + * reservedz: Must be 0. + */ + +struct dm_capabilities_resp_msg { + struct dm_header hdr; + __u64 is_accepted:1; + __u64 reservedz:63; +} __packed; + +/* + * This message is used to report memory pressure from the guest. + * This message is not part of any transaction and there is no + * response to this message. + * + * num_avail: Available memory in pages. + * num_committed: Committed memory in pages. + * page_file_size: The accumulated size of all page files + * in the system in pages. + * zero_free: The nunber of zero and free pages. + * page_file_writes: The writes to the page file in pages. + * io_diff: An indicator of file cache efficiency or page file activity, + * calculated as File Cache Page Fault Count - Page Read Count. + * This value is in pages. + * + * Some of these metrics are Windows specific and fortunately + * the algorithm on the host side that computes the guest memory + * pressure only uses num_committed value. + */ + +struct dm_status { + struct dm_header hdr; + __u64 num_avail; + __u64 num_committed; + __u64 page_file_size; + __u64 zero_free; + __u32 page_file_writes; + __u32 io_diff; +} __packed; + + +/* + * Message to ask the guest to allocate memory - balloon up message. + * This message is sent from the host to the guest. The guest may not be + * able to allocate as much memory as requested. + * + * num_pages: number of pages to allocate. + */ + +struct dm_balloon { + struct dm_header hdr; + __u32 num_pages; + __u32 reservedz; +} __packed; + + +/* + * Balloon response message; this message is sent from the guest + * to the host in response to the balloon message. + * + * reservedz: Reserved; must be set to zero. + * more_pages: If FALSE, this is the last message of the transaction. + * if TRUE there will atleast one more message from the guest. + * + * range_count: The number of ranges in the range array. + * + * range_array: An array of page ranges returned to the host. + * + */ + +struct dm_balloon_response { + struct dm_header hdr; + __u32 reservedz; + __u32 more_pages:1; + __u32 range_count:31; + union dm_mem_page_range range_array[]; +} __packed; + +/* + * Un-balloon message; this message is sent from the host + * to the guest to give guest more memory. + * + * more_pages: If FALSE, this is the last message of the transaction. + * if TRUE there will atleast one more message from the guest. + * + * reservedz: Reserved; must be set to zero. + * + * range_count: The number of ranges in the range array. + * + * range_array: An array of page ranges returned to the host. + * + */ + +struct dm_unballoon_request { + struct dm_header hdr; + __u32 more_pages:1; + __u32 reservedz:31; + __u32 range_count; + union dm_mem_page_range range_array[]; +} __packed; + +/* + * Un-balloon response message; this message is sent from the guest + * to the host in response to an unballoon request. + * + */ + +struct dm_unballoon_response { + struct dm_header hdr; +} __packed; + + +/* + * Hot add request message. Message sent from the host to the guest. + * + * mem_range: Memory range to hot add. + * + * On Linux we currently don't support this since we cannot hot add + * arbitrary granularity of memory. + */ + +struct dm_hot_add { + struct dm_header hdr; + union dm_mem_page_range range; +} __packed; + +/* + * Hot add response message. + * This message is sent by the guest to report the status of a hot add request. + * If page_count is less than the requested page count, then the host should + * assume all further hot add requests will fail, since this indicates that + * the guest has hit an upper physical memory barrier. + * + * Hot adds may also fail due to low resources; in this case, the guest must + * not complete this message until the hot add can succeed, and the host must + * not send a new hot add request until the response is sent. + * If VSC fails to hot add memory DYNMEM_NUMBER_OF_UNSUCCESSFUL_HOTADD_ATTEMPTS + * times it fails the request. + * + * + * page_count: number of pages that were successfully hot added. + * + * result: result of the operation 1: success, 0: failure. + * + */ + +struct dm_hot_add_response { + struct dm_header hdr; + __u32 page_count; + __u32 result; +} __packed; + +/* + * Types of information sent from host to the guest. + */ + +enum dm_info_type { + INFO_TYPE_MAX_PAGE_CNT = 0, + MAX_INFO_TYPE +}; + + +/* + * Header for the information message. + */ + +struct dm_info_header { + enum dm_info_type type; + __u32 data_size; +} __packed; + +/* + * This message is sent from the host to the guest to pass + * some relevant information (win8 addition). + * + * reserved: no used. + * info_size: size of the information blob. + * info: information blob. + */ + +struct dm_info_msg { + struct dm_header hdr; + __u32 reserved; + __u32 info_size; + __u8 info[]; +}; + +/* + * End protocol definitions. + */ + +/* + * State to manage hot adding memory into the guest. + * The range start_pfn : end_pfn specifies the range + * that the host has asked us to hot add. The range + * start_pfn : ha_end_pfn specifies the range that we have + * currently hot added. We hot add in multiples of 128M + * chunks; it is possible that we may not be able to bring + * online all the pages in the region. The range + * covered_start_pfn:covered_end_pfn defines the pages that can + * be brough online. + */ + +struct hv_hotadd_state { + struct list_head list; + unsigned long start_pfn; + unsigned long covered_start_pfn; + unsigned long covered_end_pfn; + unsigned long ha_end_pfn; + unsigned long end_pfn; + /* + * A list of gaps. + */ + struct list_head gap_list; +}; + +struct hv_hotadd_gap { + struct list_head list; + unsigned long start_pfn; + unsigned long end_pfn; +}; + +struct balloon_state { + __u32 num_pages; + struct work_struct wrk; +}; + +struct hot_add_wrk { + union dm_mem_page_range ha_page_range; + union dm_mem_page_range ha_region_range; + struct work_struct wrk; +}; + +static bool hot_add = true; +static bool do_hot_add; +/* + * Delay reporting memory pressure by + * the specified number of seconds. + */ +static uint pressure_report_delay = 45; + +/* + * The last time we posted a pressure report to host. + */ +static unsigned long last_post_time; + +module_param(hot_add, bool, (S_IRUGO | S_IWUSR)); +MODULE_PARM_DESC(hot_add, "If set attempt memory hot_add"); + +module_param(pressure_report_delay, uint, (S_IRUGO | S_IWUSR)); +MODULE_PARM_DESC(pressure_report_delay, "Delay in secs in reporting pressure"); +static atomic_t trans_id = ATOMIC_INIT(0); + +static int dm_ring_size = (5 * PAGE_SIZE); + +/* + * Driver specific state. + */ + +enum hv_dm_state { + DM_INITIALIZING = 0, + DM_INITIALIZED, + DM_BALLOON_UP, + DM_BALLOON_DOWN, + DM_HOT_ADD, + DM_INIT_ERROR +}; + + +static __u8 recv_buffer[PAGE_SIZE]; +static __u8 *send_buffer; +#define PAGES_IN_2M 512 +#define HA_CHUNK (32 * 1024) + +struct hv_dynmem_device { + struct hv_device *dev; + enum hv_dm_state state; + struct completion host_event; + struct completion config_event; + + /* + * Number of pages we have currently ballooned out. + */ + unsigned int num_pages_ballooned; + unsigned int num_pages_onlined; + unsigned int num_pages_added; + + /* + * State to manage the ballooning (up) operation. + */ + struct balloon_state balloon_wrk; + + /* + * State to execute the "hot-add" operation. + */ + struct hot_add_wrk ha_wrk; + + /* + * This state tracks if the host has specified a hot-add + * region. + */ + bool host_specified_ha_region; + + /* + * State to synchronize hot-add. + */ + struct completion ol_waitevent; + bool ha_waiting; + /* + * This thread handles hot-add + * requests from the host as well as notifying + * the host with regards to memory pressure in + * the guest. + */ + struct task_struct *thread; + + /* + * Protects ha_region_list, num_pages_onlined counter and individual + * regions from ha_region_list. + */ + spinlock_t ha_lock; + + /* + * A list of hot-add regions. + */ + struct list_head ha_region_list; + + /* + * We start with the highest version we can support + * and downgrade based on the host; we save here the + * next version to try. + */ + __u32 next_version; + + /* + * The negotiated version agreed by host. + */ + __u32 version; +}; + +static struct hv_dynmem_device dm_device; + +static void post_status(struct hv_dynmem_device *dm); + +#ifdef CONFIG_MEMORY_HOTPLUG +static inline bool has_pfn_is_backed(struct hv_hotadd_state *has, + unsigned long pfn) +{ + struct hv_hotadd_gap *gap; + + /* The page is not backed. */ + if ((pfn < has->covered_start_pfn) || (pfn >= has->covered_end_pfn)) + return false; + + /* Check for gaps. */ + list_for_each_entry(gap, &has->gap_list, list) { + if ((pfn >= gap->start_pfn) && (pfn < gap->end_pfn)) + return false; + } + + return true; +} + +static unsigned long hv_page_offline_check(unsigned long start_pfn, + unsigned long nr_pages) +{ + unsigned long pfn = start_pfn, count = 0; + struct hv_hotadd_state *has; + bool found; + + while (pfn < start_pfn + nr_pages) { + /* + * Search for HAS which covers the pfn and when we find one + * count how many consequitive PFNs are covered. + */ + found = false; + list_for_each_entry(has, &dm_device.ha_region_list, list) { + while ((pfn >= has->start_pfn) && + (pfn < has->end_pfn) && + (pfn < start_pfn + nr_pages)) { + found = true; + if (has_pfn_is_backed(has, pfn)) + count++; + pfn++; + } + } + + /* + * This PFN is not in any HAS (e.g. we're offlining a region + * which was present at boot), no need to account for it. Go + * to the next one. + */ + if (!found) + pfn++; + } + + return count; +} + +static int hv_memory_notifier(struct notifier_block *nb, unsigned long val, + void *v) +{ + struct memory_notify *mem = (struct memory_notify *)v; + unsigned long flags, pfn_count; + + switch (val) { + case MEM_ONLINE: + case MEM_CANCEL_ONLINE: + if (dm_device.ha_waiting) { + dm_device.ha_waiting = false; + complete(&dm_device.ol_waitevent); + } + break; + + case MEM_OFFLINE: + spin_lock_irqsave(&dm_device.ha_lock, flags); + pfn_count = hv_page_offline_check(mem->start_pfn, + mem->nr_pages); + if (pfn_count <= dm_device.num_pages_onlined) { + dm_device.num_pages_onlined -= pfn_count; + } else { + /* + * We're offlining more pages than we managed to online. + * This is unexpected. In any case don't let + * num_pages_onlined wrap around zero. + */ + WARN_ON_ONCE(1); + dm_device.num_pages_onlined = 0; + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + break; + case MEM_GOING_ONLINE: + case MEM_GOING_OFFLINE: + case MEM_CANCEL_OFFLINE: + break; + } + return NOTIFY_OK; +} + +static struct notifier_block hv_memory_nb = { + .notifier_call = hv_memory_notifier, + .priority = 0 +}; + +/* Check if the particular page is backed and can be onlined and online it. */ +static void hv_page_online_one(struct hv_hotadd_state *has, struct page *pg) +{ + if (!has_pfn_is_backed(has, page_to_pfn(pg))) + return; + + /* This frame is currently backed; online the page. */ + __online_page_set_limits(pg); + __online_page_increment_counters(pg); + __online_page_free(pg); + + WARN_ON_ONCE(!spin_is_locked(&dm_device.ha_lock)); + dm_device.num_pages_onlined++; +} + +static void hv_bring_pgs_online(struct hv_hotadd_state *has, + unsigned long start_pfn, unsigned long size) +{ + int i; + + pr_debug("Online %lu pages starting at pfn 0x%lx\n", size, start_pfn); + for (i = 0; i < size; i++) + hv_page_online_one(has, pfn_to_page(start_pfn + i)); +} + +static void hv_mem_hot_add(unsigned long start, unsigned long size, + unsigned long pfn_count, + struct hv_hotadd_state *has) +{ + int ret = 0; + int i, nid; + unsigned long start_pfn; + unsigned long processed_pfn; + unsigned long total_pfn = pfn_count; + unsigned long flags; + + for (i = 0; i < (size/HA_CHUNK); i++) { + start_pfn = start + (i * HA_CHUNK); + + spin_lock_irqsave(&dm_device.ha_lock, flags); + has->ha_end_pfn += HA_CHUNK; + + if (total_pfn > HA_CHUNK) { + processed_pfn = HA_CHUNK; + total_pfn -= HA_CHUNK; + } else { + processed_pfn = total_pfn; + total_pfn = 0; + } + + has->covered_end_pfn += processed_pfn; + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + + init_completion(&dm_device.ol_waitevent); + dm_device.ha_waiting = !memhp_auto_online; + + nid = memory_add_physaddr_to_nid(PFN_PHYS(start_pfn)); + ret = add_memory(nid, PFN_PHYS((start_pfn)), + (HA_CHUNK << PAGE_SHIFT)); + + if (ret) { + pr_err("hot_add memory failed error is %d\n", ret); + if (ret == -EEXIST) { + /* + * This error indicates that the error + * is not a transient failure. This is the + * case where the guest's physical address map + * precludes hot adding memory. Stop all further + * memory hot-add. + */ + do_hot_add = false; + } + spin_lock_irqsave(&dm_device.ha_lock, flags); + has->ha_end_pfn -= HA_CHUNK; + has->covered_end_pfn -= processed_pfn; + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + break; + } + + /* + * Wait for the memory block to be onlined when memory onlining + * is done outside of kernel (memhp_auto_online). Since the hot + * add has succeeded, it is ok to proceed even if the pages in + * the hot added region have not been "onlined" within the + * allowed time. + */ + if (dm_device.ha_waiting) + wait_for_completion_timeout(&dm_device.ol_waitevent, + 5*HZ); + post_status(&dm_device); + } +} + +static void hv_online_page(struct page *pg) +{ + struct hv_hotadd_state *has; + unsigned long flags; + unsigned long pfn = page_to_pfn(pg); + + spin_lock_irqsave(&dm_device.ha_lock, flags); + list_for_each_entry(has, &dm_device.ha_region_list, list) { + /* The page belongs to a different HAS. */ + if ((pfn < has->start_pfn) || (pfn >= has->end_pfn)) + continue; + + hv_page_online_one(has, pg); + break; + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); +} + +static int pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt) +{ + struct hv_hotadd_state *has; + struct hv_hotadd_gap *gap; + unsigned long residual, new_inc; + int ret = 0; + unsigned long flags; + + spin_lock_irqsave(&dm_device.ha_lock, flags); + list_for_each_entry(has, &dm_device.ha_region_list, list) { + /* + * If the pfn range we are dealing with is not in the current + * "hot add block", move on. + */ + if (start_pfn < has->start_pfn || start_pfn >= has->end_pfn) + continue; + + /* + * If the current start pfn is not where the covered_end + * is, create a gap and update covered_end_pfn. + */ + if (has->covered_end_pfn != start_pfn) { + gap = kzalloc(sizeof(struct hv_hotadd_gap), GFP_ATOMIC); + if (!gap) { + ret = -ENOMEM; + break; + } + + INIT_LIST_HEAD(&gap->list); + gap->start_pfn = has->covered_end_pfn; + gap->end_pfn = start_pfn; + list_add_tail(&gap->list, &has->gap_list); + + has->covered_end_pfn = start_pfn; + } + + /* + * If the current hot add-request extends beyond + * our current limit; extend it. + */ + if ((start_pfn + pfn_cnt) > has->end_pfn) { + residual = (start_pfn + pfn_cnt - has->end_pfn); + /* + * Extend the region by multiples of HA_CHUNK. + */ + new_inc = (residual / HA_CHUNK) * HA_CHUNK; + if (residual % HA_CHUNK) + new_inc += HA_CHUNK; + + has->end_pfn += new_inc; + } + + ret = 1; + break; + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + + return ret; +} + +static unsigned long handle_pg_range(unsigned long pg_start, + unsigned long pg_count) +{ + unsigned long start_pfn = pg_start; + unsigned long pfn_cnt = pg_count; + unsigned long size; + struct hv_hotadd_state *has; + unsigned long pgs_ol = 0; + unsigned long old_covered_state; + unsigned long res = 0, flags; + + pr_debug("Hot adding %lu pages starting at pfn 0x%lx.\n", pg_count, + pg_start); + + spin_lock_irqsave(&dm_device.ha_lock, flags); + list_for_each_entry(has, &dm_device.ha_region_list, list) { + /* + * If the pfn range we are dealing with is not in the current + * "hot add block", move on. + */ + if (start_pfn < has->start_pfn || start_pfn >= has->end_pfn) + continue; + + old_covered_state = has->covered_end_pfn; + + if (start_pfn < has->ha_end_pfn) { + /* + * This is the case where we are backing pages + * in an already hot added region. Bring + * these pages online first. + */ + pgs_ol = has->ha_end_pfn - start_pfn; + if (pgs_ol > pfn_cnt) + pgs_ol = pfn_cnt; + + has->covered_end_pfn += pgs_ol; + pfn_cnt -= pgs_ol; + /* + * Check if the corresponding memory block is already + * online. It is possible to observe struct pages still + * being uninitialized here so check section instead. + * In case the section is online we need to bring the + * rest of pfns (which were not backed previously) + * online too. + */ + if (start_pfn > has->start_pfn && + online_section_nr(pfn_to_section_nr(start_pfn))) + hv_bring_pgs_online(has, start_pfn, pgs_ol); + + } + + if ((has->ha_end_pfn < has->end_pfn) && (pfn_cnt > 0)) { + /* + * We have some residual hot add range + * that needs to be hot added; hot add + * it now. Hot add a multiple of + * of HA_CHUNK that fully covers the pages + * we have. + */ + size = (has->end_pfn - has->ha_end_pfn); + if (pfn_cnt <= size) { + size = ((pfn_cnt / HA_CHUNK) * HA_CHUNK); + if (pfn_cnt % HA_CHUNK) + size += HA_CHUNK; + } else { + pfn_cnt = size; + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + hv_mem_hot_add(has->ha_end_pfn, size, pfn_cnt, has); + spin_lock_irqsave(&dm_device.ha_lock, flags); + } + /* + * If we managed to online any pages that were given to us, + * we declare success. + */ + res = has->covered_end_pfn - old_covered_state; + break; + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + + return res; +} + +static unsigned long process_hot_add(unsigned long pg_start, + unsigned long pfn_cnt, + unsigned long rg_start, + unsigned long rg_size) +{ + struct hv_hotadd_state *ha_region = NULL; + int covered; + unsigned long flags; + + if (pfn_cnt == 0) + return 0; + + if (!dm_device.host_specified_ha_region) { + covered = pfn_covered(pg_start, pfn_cnt); + if (covered < 0) + return 0; + + if (covered) + goto do_pg_range; + } + + /* + * If the host has specified a hot-add range; deal with it first. + */ + + if (rg_size != 0) { + ha_region = kzalloc(sizeof(struct hv_hotadd_state), GFP_KERNEL); + if (!ha_region) + return 0; + + INIT_LIST_HEAD(&ha_region->list); + INIT_LIST_HEAD(&ha_region->gap_list); + + ha_region->start_pfn = rg_start; + ha_region->ha_end_pfn = rg_start; + ha_region->covered_start_pfn = pg_start; + ha_region->covered_end_pfn = pg_start; + ha_region->end_pfn = rg_start + rg_size; + + spin_lock_irqsave(&dm_device.ha_lock, flags); + list_add_tail(&ha_region->list, &dm_device.ha_region_list); + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + } + +do_pg_range: + /* + * Process the page range specified; bringing them + * online if possible. + */ + return handle_pg_range(pg_start, pfn_cnt); +} + +#endif + +static void hot_add_req(struct work_struct *dummy) +{ + struct dm_hot_add_response resp; +#ifdef CONFIG_MEMORY_HOTPLUG + unsigned long pg_start, pfn_cnt; + unsigned long rg_start, rg_sz; +#endif + struct hv_dynmem_device *dm = &dm_device; + + memset(&resp, 0, sizeof(struct dm_hot_add_response)); + resp.hdr.type = DM_MEM_HOT_ADD_RESPONSE; + resp.hdr.size = sizeof(struct dm_hot_add_response); + +#ifdef CONFIG_MEMORY_HOTPLUG + pg_start = dm->ha_wrk.ha_page_range.finfo.start_page; + pfn_cnt = dm->ha_wrk.ha_page_range.finfo.page_cnt; + + rg_start = dm->ha_wrk.ha_region_range.finfo.start_page; + rg_sz = dm->ha_wrk.ha_region_range.finfo.page_cnt; + + if ((rg_start == 0) && (!dm->host_specified_ha_region)) { + unsigned long region_size; + unsigned long region_start; + + /* + * The host has not specified the hot-add region. + * Based on the hot-add page range being specified, + * compute a hot-add region that can cover the pages + * that need to be hot-added while ensuring the alignment + * and size requirements of Linux as it relates to hot-add. + */ + region_start = pg_start; + region_size = (pfn_cnt / HA_CHUNK) * HA_CHUNK; + if (pfn_cnt % HA_CHUNK) + region_size += HA_CHUNK; + + region_start = (pg_start / HA_CHUNK) * HA_CHUNK; + + rg_start = region_start; + rg_sz = region_size; + } + + if (do_hot_add) + resp.page_count = process_hot_add(pg_start, pfn_cnt, + rg_start, rg_sz); + + dm->num_pages_added += resp.page_count; +#endif + /* + * The result field of the response structure has the + * following semantics: + * + * 1. If all or some pages hot-added: Guest should return success. + * + * 2. If no pages could be hot-added: + * + * If the guest returns success, then the host + * will not attempt any further hot-add operations. This + * signifies a permanent failure. + * + * If the guest returns failure, then this failure will be + * treated as a transient failure and the host may retry the + * hot-add operation after some delay. + */ + if (resp.page_count > 0) + resp.result = 1; + else if (!do_hot_add) + resp.result = 1; + else + resp.result = 0; + + if (!do_hot_add || (resp.page_count == 0)) + pr_err("Memory hot add failed\n"); + + dm->state = DM_INITIALIZED; + resp.hdr.trans_id = atomic_inc_return(&trans_id); + vmbus_sendpacket(dm->dev->channel, &resp, + sizeof(struct dm_hot_add_response), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); +} + +static void process_info(struct hv_dynmem_device *dm, struct dm_info_msg *msg) +{ + struct dm_info_header *info_hdr; + + info_hdr = (struct dm_info_header *)msg->info; + + switch (info_hdr->type) { + case INFO_TYPE_MAX_PAGE_CNT: + if (info_hdr->data_size == sizeof(__u64)) { + __u64 *max_page_count = (__u64 *)&info_hdr[1]; + + pr_info("Max. dynamic memory size: %llu MB\n", + (*max_page_count) >> (20 - PAGE_SHIFT)); + } + + break; + default: + pr_warn("Received Unknown type: %d\n", info_hdr->type); + } +} + +static unsigned long compute_balloon_floor(void) +{ + unsigned long min_pages; +#define MB2PAGES(mb) ((mb) << (20 - PAGE_SHIFT)) + /* Simple continuous piecewiese linear function: + * max MiB -> min MiB gradient + * 0 0 + * 16 16 + * 32 24 + * 128 72 (1/2) + * 512 168 (1/4) + * 2048 360 (1/8) + * 8192 744 (1/16) + * 32768 1512 (1/32) + */ + if (totalram_pages < MB2PAGES(128)) + min_pages = MB2PAGES(8) + (totalram_pages >> 1); + else if (totalram_pages < MB2PAGES(512)) + min_pages = MB2PAGES(40) + (totalram_pages >> 2); + else if (totalram_pages < MB2PAGES(2048)) + min_pages = MB2PAGES(104) + (totalram_pages >> 3); + else if (totalram_pages < MB2PAGES(8192)) + min_pages = MB2PAGES(232) + (totalram_pages >> 4); + else + min_pages = MB2PAGES(488) + (totalram_pages >> 5); +#undef MB2PAGES + return min_pages; +} + +/* + * Post our status as it relates memory pressure to the + * host. Host expects the guests to post this status + * periodically at 1 second intervals. + * + * The metrics specified in this protocol are very Windows + * specific and so we cook up numbers here to convey our memory + * pressure. + */ + +static void post_status(struct hv_dynmem_device *dm) +{ + struct dm_status status; + unsigned long now = jiffies; + unsigned long last_post = last_post_time; + + if (pressure_report_delay > 0) { + --pressure_report_delay; + return; + } + + if (!time_after(now, (last_post_time + HZ))) + return; + + memset(&status, 0, sizeof(struct dm_status)); + status.hdr.type = DM_STATUS_REPORT; + status.hdr.size = sizeof(struct dm_status); + status.hdr.trans_id = atomic_inc_return(&trans_id); + + /* + * The host expects the guest to report free and committed memory. + * Furthermore, the host expects the pressure information to include + * the ballooned out pages. For a given amount of memory that we are + * managing we need to compute a floor below which we should not + * balloon. Compute this and add it to the pressure report. + * We also need to report all offline pages (num_pages_added - + * num_pages_onlined) as committed to the host, otherwise it can try + * asking us to balloon them out. + */ + status.num_avail = si_mem_available(); + status.num_committed = vm_memory_committed() + + dm->num_pages_ballooned + + (dm->num_pages_added > dm->num_pages_onlined ? + dm->num_pages_added - dm->num_pages_onlined : 0) + + compute_balloon_floor(); + + trace_balloon_status(status.num_avail, status.num_committed, + vm_memory_committed(), dm->num_pages_ballooned, + dm->num_pages_added, dm->num_pages_onlined); + /* + * If our transaction ID is no longer current, just don't + * send the status. This can happen if we were interrupted + * after we picked our transaction ID. + */ + if (status.hdr.trans_id != atomic_read(&trans_id)) + return; + + /* + * If the last post time that we sampled has changed, + * we have raced, don't post the status. + */ + if (last_post != last_post_time) + return; + + last_post_time = jiffies; + vmbus_sendpacket(dm->dev->channel, &status, + sizeof(struct dm_status), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + +} + +static void free_balloon_pages(struct hv_dynmem_device *dm, + union dm_mem_page_range *range_array) +{ + int num_pages = range_array->finfo.page_cnt; + __u64 start_frame = range_array->finfo.start_page; + struct page *pg; + int i; + + for (i = 0; i < num_pages; i++) { + pg = pfn_to_page(i + start_frame); + __free_page(pg); + dm->num_pages_ballooned--; + } +} + + + +static unsigned int alloc_balloon_pages(struct hv_dynmem_device *dm, + unsigned int num_pages, + struct dm_balloon_response *bl_resp, + int alloc_unit) +{ + unsigned int i = 0; + struct page *pg; + + for (i = 0; i < num_pages / alloc_unit; i++) { + if (bl_resp->hdr.size + sizeof(union dm_mem_page_range) > + PAGE_SIZE) + return i * alloc_unit; + + /* + * We execute this code in a thread context. Furthermore, + * we don't want the kernel to try too hard. + */ + pg = alloc_pages(GFP_HIGHUSER | __GFP_NORETRY | + __GFP_NOMEMALLOC | __GFP_NOWARN, + get_order(alloc_unit << PAGE_SHIFT)); + + if (!pg) + return i * alloc_unit; + + dm->num_pages_ballooned += alloc_unit; + + /* + * If we allocatted 2M pages; split them so we + * can free them in any order we get. + */ + + if (alloc_unit != 1) + split_page(pg, get_order(alloc_unit << PAGE_SHIFT)); + + bl_resp->range_count++; + bl_resp->range_array[i].finfo.start_page = + page_to_pfn(pg); + bl_resp->range_array[i].finfo.page_cnt = alloc_unit; + bl_resp->hdr.size += sizeof(union dm_mem_page_range); + + } + + return i * alloc_unit; +} + +static void balloon_up(struct work_struct *dummy) +{ + unsigned int num_pages = dm_device.balloon_wrk.num_pages; + unsigned int num_ballooned = 0; + struct dm_balloon_response *bl_resp; + int alloc_unit; + int ret; + bool done = false; + int i; + long avail_pages; + unsigned long floor; + + /* + * We will attempt 2M allocations. However, if we fail to + * allocate 2M chunks, we will go back to 4k allocations. + */ + alloc_unit = 512; + + avail_pages = si_mem_available(); + floor = compute_balloon_floor(); + + /* Refuse to balloon below the floor. */ + if (avail_pages < num_pages || avail_pages - num_pages < floor) { + pr_info("Balloon request will be partially fulfilled. %s\n", + avail_pages < num_pages ? "Not enough memory." : + "Balloon floor reached."); + + num_pages = avail_pages > floor ? (avail_pages - floor) : 0; + } + + while (!done) { + bl_resp = (struct dm_balloon_response *)send_buffer; + memset(send_buffer, 0, PAGE_SIZE); + bl_resp->hdr.type = DM_BALLOON_RESPONSE; + bl_resp->hdr.size = sizeof(struct dm_balloon_response); + bl_resp->more_pages = 1; + + num_pages -= num_ballooned; + num_ballooned = alloc_balloon_pages(&dm_device, num_pages, + bl_resp, alloc_unit); + + if (alloc_unit != 1 && num_ballooned == 0) { + alloc_unit = 1; + continue; + } + + if (num_ballooned == 0 || num_ballooned == num_pages) { + pr_debug("Ballooned %u out of %u requested pages.\n", + num_pages, dm_device.balloon_wrk.num_pages); + + bl_resp->more_pages = 0; + done = true; + dm_device.state = DM_INITIALIZED; + } + + /* + * We are pushing a lot of data through the channel; + * deal with transient failures caused because of the + * lack of space in the ring buffer. + */ + + do { + bl_resp->hdr.trans_id = atomic_inc_return(&trans_id); + ret = vmbus_sendpacket(dm_device.dev->channel, + bl_resp, + bl_resp->hdr.size, + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + + if (ret == -EAGAIN) + msleep(20); + post_status(&dm_device); + } while (ret == -EAGAIN); + + if (ret) { + /* + * Free up the memory we allocatted. + */ + pr_err("Balloon response failed\n"); + + for (i = 0; i < bl_resp->range_count; i++) + free_balloon_pages(&dm_device, + &bl_resp->range_array[i]); + + done = true; + } + } + +} + +static void balloon_down(struct hv_dynmem_device *dm, + struct dm_unballoon_request *req) +{ + union dm_mem_page_range *range_array = req->range_array; + int range_count = req->range_count; + struct dm_unballoon_response resp; + int i; + unsigned int prev_pages_ballooned = dm->num_pages_ballooned; + + for (i = 0; i < range_count; i++) { + free_balloon_pages(dm, &range_array[i]); + complete(&dm_device.config_event); + } + + pr_debug("Freed %u ballooned pages.\n", + prev_pages_ballooned - dm->num_pages_ballooned); + + if (req->more_pages == 1) + return; + + memset(&resp, 0, sizeof(struct dm_unballoon_response)); + resp.hdr.type = DM_UNBALLOON_RESPONSE; + resp.hdr.trans_id = atomic_inc_return(&trans_id); + resp.hdr.size = sizeof(struct dm_unballoon_response); + + vmbus_sendpacket(dm_device.dev->channel, &resp, + sizeof(struct dm_unballoon_response), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + + dm->state = DM_INITIALIZED; +} + +static void balloon_onchannelcallback(void *context); + +static int dm_thread_func(void *dm_dev) +{ + struct hv_dynmem_device *dm = dm_dev; + + while (!kthread_should_stop()) { + wait_for_completion_interruptible_timeout( + &dm_device.config_event, 1*HZ); + /* + * The host expects us to post information on the memory + * pressure every second. + */ + reinit_completion(&dm_device.config_event); + post_status(dm); + } + + return 0; +} + + +static void version_resp(struct hv_dynmem_device *dm, + struct dm_version_response *vresp) +{ + struct dm_version_request version_req; + int ret; + + if (vresp->is_accepted) { + /* + * We are done; wakeup the + * context waiting for version + * negotiation. + */ + complete(&dm->host_event); + return; + } + /* + * If there are more versions to try, continue + * with negotiations; if not + * shutdown the service since we are not able + * to negotiate a suitable version number + * with the host. + */ + if (dm->next_version == 0) + goto version_error; + + memset(&version_req, 0, sizeof(struct dm_version_request)); + version_req.hdr.type = DM_VERSION_REQUEST; + version_req.hdr.size = sizeof(struct dm_version_request); + version_req.hdr.trans_id = atomic_inc_return(&trans_id); + version_req.version.version = dm->next_version; + dm->version = version_req.version.version; + + /* + * Set the next version to try in case current version fails. + * Win7 protocol ought to be the last one to try. + */ + switch (version_req.version.version) { + case DYNMEM_PROTOCOL_VERSION_WIN8: + dm->next_version = DYNMEM_PROTOCOL_VERSION_WIN7; + version_req.is_last_attempt = 0; + break; + default: + dm->next_version = 0; + version_req.is_last_attempt = 1; + } + + ret = vmbus_sendpacket(dm->dev->channel, &version_req, + sizeof(struct dm_version_request), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + + if (ret) + goto version_error; + + return; + +version_error: + dm->state = DM_INIT_ERROR; + complete(&dm->host_event); +} + +static void cap_resp(struct hv_dynmem_device *dm, + struct dm_capabilities_resp_msg *cap_resp) +{ + if (!cap_resp->is_accepted) { + pr_err("Capabilities not accepted by host\n"); + dm->state = DM_INIT_ERROR; + } + complete(&dm->host_event); +} + +static void balloon_onchannelcallback(void *context) +{ + struct hv_device *dev = context; + u32 recvlen; + u64 requestid; + struct dm_message *dm_msg; + struct dm_header *dm_hdr; + struct hv_dynmem_device *dm = hv_get_drvdata(dev); + struct dm_balloon *bal_msg; + struct dm_hot_add *ha_msg; + union dm_mem_page_range *ha_pg_range; + union dm_mem_page_range *ha_region; + + memset(recv_buffer, 0, sizeof(recv_buffer)); + vmbus_recvpacket(dev->channel, recv_buffer, + PAGE_SIZE, &recvlen, &requestid); + + if (recvlen > 0) { + dm_msg = (struct dm_message *)recv_buffer; + dm_hdr = &dm_msg->hdr; + + switch (dm_hdr->type) { + case DM_VERSION_RESPONSE: + version_resp(dm, + (struct dm_version_response *)dm_msg); + break; + + case DM_CAPABILITIES_RESPONSE: + cap_resp(dm, + (struct dm_capabilities_resp_msg *)dm_msg); + break; + + case DM_BALLOON_REQUEST: + if (dm->state == DM_BALLOON_UP) + pr_warn("Currently ballooning\n"); + bal_msg = (struct dm_balloon *)recv_buffer; + dm->state = DM_BALLOON_UP; + dm_device.balloon_wrk.num_pages = bal_msg->num_pages; + schedule_work(&dm_device.balloon_wrk.wrk); + break; + + case DM_UNBALLOON_REQUEST: + dm->state = DM_BALLOON_DOWN; + balloon_down(dm, + (struct dm_unballoon_request *)recv_buffer); + break; + + case DM_MEM_HOT_ADD_REQUEST: + if (dm->state == DM_HOT_ADD) + pr_warn("Currently hot-adding\n"); + dm->state = DM_HOT_ADD; + ha_msg = (struct dm_hot_add *)recv_buffer; + if (ha_msg->hdr.size == sizeof(struct dm_hot_add)) { + /* + * This is a normal hot-add request specifying + * hot-add memory. + */ + dm->host_specified_ha_region = false; + ha_pg_range = &ha_msg->range; + dm->ha_wrk.ha_page_range = *ha_pg_range; + dm->ha_wrk.ha_region_range.page_range = 0; + } else { + /* + * Host is specifying that we first hot-add + * a region and then partially populate this + * region. + */ + dm->host_specified_ha_region = true; + ha_pg_range = &ha_msg->range; + ha_region = &ha_pg_range[1]; + dm->ha_wrk.ha_page_range = *ha_pg_range; + dm->ha_wrk.ha_region_range = *ha_region; + } + schedule_work(&dm_device.ha_wrk.wrk); + break; + + case DM_INFO_MESSAGE: + process_info(dm, (struct dm_info_msg *)dm_msg); + break; + + default: + pr_warn_ratelimited("Unhandled message: type: %d\n", dm_hdr->type); + + } + } + +} + +static int balloon_probe(struct hv_device *dev, + const struct hv_vmbus_device_id *dev_id) +{ + int ret; + unsigned long t; + struct dm_version_request version_req; + struct dm_capabilities cap_msg; + +#ifdef CONFIG_MEMORY_HOTPLUG + do_hot_add = hot_add; +#else + do_hot_add = false; +#endif + + /* + * First allocate a send buffer. + */ + + send_buffer = kmalloc(PAGE_SIZE, GFP_KERNEL); + if (!send_buffer) + return -ENOMEM; + + ret = vmbus_open(dev->channel, dm_ring_size, dm_ring_size, NULL, 0, + balloon_onchannelcallback, dev); + + if (ret) + goto probe_error0; + + dm_device.dev = dev; + dm_device.state = DM_INITIALIZING; + dm_device.next_version = DYNMEM_PROTOCOL_VERSION_WIN8; + init_completion(&dm_device.host_event); + init_completion(&dm_device.config_event); + INIT_LIST_HEAD(&dm_device.ha_region_list); + spin_lock_init(&dm_device.ha_lock); + INIT_WORK(&dm_device.balloon_wrk.wrk, balloon_up); + INIT_WORK(&dm_device.ha_wrk.wrk, hot_add_req); + dm_device.host_specified_ha_region = false; + + dm_device.thread = + kthread_run(dm_thread_func, &dm_device, "hv_balloon"); + if (IS_ERR(dm_device.thread)) { + ret = PTR_ERR(dm_device.thread); + goto probe_error1; + } + +#ifdef CONFIG_MEMORY_HOTPLUG + set_online_page_callback(&hv_online_page); + register_memory_notifier(&hv_memory_nb); +#endif + + hv_set_drvdata(dev, &dm_device); + /* + * Initiate the hand shake with the host and negotiate + * a version that the host can support. We start with the + * highest version number and go down if the host cannot + * support it. + */ + memset(&version_req, 0, sizeof(struct dm_version_request)); + version_req.hdr.type = DM_VERSION_REQUEST; + version_req.hdr.size = sizeof(struct dm_version_request); + version_req.hdr.trans_id = atomic_inc_return(&trans_id); + version_req.version.version = DYNMEM_PROTOCOL_VERSION_WIN10; + version_req.is_last_attempt = 0; + dm_device.version = version_req.version.version; + + ret = vmbus_sendpacket(dev->channel, &version_req, + sizeof(struct dm_version_request), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + if (ret) + goto probe_error2; + + t = wait_for_completion_timeout(&dm_device.host_event, 5*HZ); + if (t == 0) { + ret = -ETIMEDOUT; + goto probe_error2; + } + + /* + * If we could not negotiate a compatible version with the host + * fail the probe function. + */ + if (dm_device.state == DM_INIT_ERROR) { + ret = -ETIMEDOUT; + goto probe_error2; + } + + pr_info("Using Dynamic Memory protocol version %u.%u\n", + DYNMEM_MAJOR_VERSION(dm_device.version), + DYNMEM_MINOR_VERSION(dm_device.version)); + + /* + * Now submit our capabilities to the host. + */ + memset(&cap_msg, 0, sizeof(struct dm_capabilities)); + cap_msg.hdr.type = DM_CAPABILITIES_REPORT; + cap_msg.hdr.size = sizeof(struct dm_capabilities); + cap_msg.hdr.trans_id = atomic_inc_return(&trans_id); + + cap_msg.caps.cap_bits.balloon = 1; + cap_msg.caps.cap_bits.hot_add = 1; + + /* + * Specify our alignment requirements as it relates + * memory hot-add. Specify 128MB alignment. + */ + cap_msg.caps.cap_bits.hot_add_alignment = 7; + + /* + * Currently the host does not use these + * values and we set them to what is done in the + * Windows driver. + */ + cap_msg.min_page_cnt = 0; + cap_msg.max_page_number = -1; + + ret = vmbus_sendpacket(dev->channel, &cap_msg, + sizeof(struct dm_capabilities), + (unsigned long)NULL, + VM_PKT_DATA_INBAND, 0); + if (ret) + goto probe_error2; + + t = wait_for_completion_timeout(&dm_device.host_event, 5*HZ); + if (t == 0) { + ret = -ETIMEDOUT; + goto probe_error2; + } + + /* + * If the host does not like our capabilities, + * fail the probe function. + */ + if (dm_device.state == DM_INIT_ERROR) { + ret = -ETIMEDOUT; + goto probe_error2; + } + + dm_device.state = DM_INITIALIZED; + last_post_time = jiffies; + + return 0; + +probe_error2: +#ifdef CONFIG_MEMORY_HOTPLUG + restore_online_page_callback(&hv_online_page); +#endif + kthread_stop(dm_device.thread); + +probe_error1: + vmbus_close(dev->channel); +probe_error0: + kfree(send_buffer); + return ret; +} + +static int balloon_remove(struct hv_device *dev) +{ + struct hv_dynmem_device *dm = hv_get_drvdata(dev); + struct hv_hotadd_state *has, *tmp; + struct hv_hotadd_gap *gap, *tmp_gap; + unsigned long flags; + + if (dm->num_pages_ballooned != 0) + pr_warn("Ballooned pages: %d\n", dm->num_pages_ballooned); + + cancel_work_sync(&dm->balloon_wrk.wrk); + cancel_work_sync(&dm->ha_wrk.wrk); + + vmbus_close(dev->channel); + kthread_stop(dm->thread); + kfree(send_buffer); +#ifdef CONFIG_MEMORY_HOTPLUG + restore_online_page_callback(&hv_online_page); + unregister_memory_notifier(&hv_memory_nb); +#endif + spin_lock_irqsave(&dm_device.ha_lock, flags); + list_for_each_entry_safe(has, tmp, &dm->ha_region_list, list) { + list_for_each_entry_safe(gap, tmp_gap, &has->gap_list, list) { + list_del(&gap->list); + kfree(gap); + } + list_del(&has->list); + kfree(has); + } + spin_unlock_irqrestore(&dm_device.ha_lock, flags); + + return 0; +} + +static const struct hv_vmbus_device_id id_table[] = { + /* Dynamic Memory Class ID */ + /* 525074DC-8985-46e2-8057-A307DC18A502 */ + { HV_DM_GUID, }, + { }, +}; + +MODULE_DEVICE_TABLE(vmbus, id_table); + +static struct hv_driver balloon_drv = { + .name = "hv_balloon", + .id_table = id_table, + .probe = balloon_probe, + .remove = balloon_remove, + .driver = { + .probe_type = PROBE_PREFER_ASYNCHRONOUS, + }, +}; + +static int __init init_balloon_drv(void) +{ + + return vmbus_driver_register(&balloon_drv); +} + +module_init(init_balloon_drv); + +MODULE_DESCRIPTION("Hyper-V Balloon"); +MODULE_LICENSE("GPL"); diff --git a/drivers/hv/hv_fcopy.c b/drivers/hv/hv_fcopy.c new file mode 100644 index 000000000..2364281d8 --- /dev/null +++ b/drivers/hv/hv_fcopy.c @@ -0,0 +1,363 @@ +/* + * An implementation of file copy service. + * + * Copyright (C) 2014, Microsoft, Inc. + * + * Author : K. Y. Srinivasan <ksrinivasan@novell.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/nls.h> +#include <linux/workqueue.h> +#include <linux/hyperv.h> +#include <linux/sched.h> + +#include "hyperv_vmbus.h" +#include "hv_utils_transport.h" + +#define WIN8_SRV_MAJOR 1 +#define WIN8_SRV_MINOR 1 +#define WIN8_SRV_VERSION (WIN8_SRV_MAJOR << 16 | WIN8_SRV_MINOR) + +#define FCOPY_VER_COUNT 1 +static const int fcopy_versions[] = { + WIN8_SRV_VERSION +}; + +#define FW_VER_COUNT 1 +static const int fw_versions[] = { + UTIL_FW_VERSION +}; + +/* + * Global state maintained for transaction that is being processed. + * For a class of integration services, including the "file copy service", + * the specified protocol is a "request/response" protocol which means that + * there can only be single outstanding transaction from the host at any + * given point in time. We use this to simplify memory management in this + * driver - we cache and process only one message at a time. + * + * While the request/response protocol is guaranteed by the host, we further + * ensure this by serializing packet processing in this driver - we do not + * read additional packets from the VMBUs until the current packet is fully + * handled. + */ + +static struct { + int state; /* hvutil_device_state */ + int recv_len; /* number of bytes received. */ + struct hv_fcopy_hdr *fcopy_msg; /* current message */ + struct vmbus_channel *recv_channel; /* chn we got the request */ + u64 recv_req_id; /* request ID. */ +} fcopy_transaction; + +static void fcopy_respond_to_host(int error); +static void fcopy_send_data(struct work_struct *dummy); +static void fcopy_timeout_func(struct work_struct *dummy); +static DECLARE_DELAYED_WORK(fcopy_timeout_work, fcopy_timeout_func); +static DECLARE_WORK(fcopy_send_work, fcopy_send_data); +static const char fcopy_devname[] = "vmbus/hv_fcopy"; +static u8 *recv_buffer; +static struct hvutil_transport *hvt; +/* + * This state maintains the version number registered by the daemon. + */ +static int dm_reg_value; + +static void fcopy_poll_wrapper(void *channel) +{ + /* Transaction is finished, reset the state here to avoid races. */ + fcopy_transaction.state = HVUTIL_READY; + hv_fcopy_onchannelcallback(channel); +} + +static void fcopy_timeout_func(struct work_struct *dummy) +{ + /* + * If the timer fires, the user-mode component has not responded; + * process the pending transaction. + */ + fcopy_respond_to_host(HV_E_FAIL); + hv_poll_channel(fcopy_transaction.recv_channel, fcopy_poll_wrapper); +} + +static void fcopy_register_done(void) +{ + pr_debug("FCP: userspace daemon registered\n"); + hv_poll_channel(fcopy_transaction.recv_channel, fcopy_poll_wrapper); +} + +static int fcopy_handle_handshake(u32 version) +{ + u32 our_ver = FCOPY_CURRENT_VERSION; + + switch (version) { + case FCOPY_VERSION_0: + /* Daemon doesn't expect us to reply */ + dm_reg_value = version; + break; + case FCOPY_VERSION_1: + /* Daemon expects us to reply with our own version */ + if (hvutil_transport_send(hvt, &our_ver, sizeof(our_ver), + fcopy_register_done)) + return -EFAULT; + dm_reg_value = version; + break; + default: + /* + * For now we will fail the registration. + * If and when we have multiple versions to + * deal with, we will be backward compatible. + * We will add this code when needed. + */ + return -EINVAL; + } + pr_debug("FCP: userspace daemon ver. %d connected\n", version); + return 0; +} + +static void fcopy_send_data(struct work_struct *dummy) +{ + struct hv_start_fcopy *smsg_out = NULL; + int operation = fcopy_transaction.fcopy_msg->operation; + struct hv_start_fcopy *smsg_in; + void *out_src; + int rc, out_len; + + /* + * The strings sent from the host are encoded in + * in utf16; convert it to utf8 strings. + * The host assures us that the utf16 strings will not exceed + * the max lengths specified. We will however, reserve room + * for the string terminating character - in the utf16s_utf8s() + * function we limit the size of the buffer where the converted + * string is placed to W_MAX_PATH -1 to guarantee + * that the strings can be properly terminated! + */ + + switch (operation) { + case START_FILE_COPY: + out_len = sizeof(struct hv_start_fcopy); + smsg_out = kzalloc(sizeof(*smsg_out), GFP_KERNEL); + if (!smsg_out) + return; + + smsg_out->hdr.operation = operation; + smsg_in = (struct hv_start_fcopy *)fcopy_transaction.fcopy_msg; + + utf16s_to_utf8s((wchar_t *)smsg_in->file_name, W_MAX_PATH, + UTF16_LITTLE_ENDIAN, + (__u8 *)&smsg_out->file_name, W_MAX_PATH - 1); + + utf16s_to_utf8s((wchar_t *)smsg_in->path_name, W_MAX_PATH, + UTF16_LITTLE_ENDIAN, + (__u8 *)&smsg_out->path_name, W_MAX_PATH - 1); + + smsg_out->copy_flags = smsg_in->copy_flags; + smsg_out->file_size = smsg_in->file_size; + out_src = smsg_out; + break; + + case WRITE_TO_FILE: + out_src = fcopy_transaction.fcopy_msg; + out_len = sizeof(struct hv_do_fcopy); + break; + default: + out_src = fcopy_transaction.fcopy_msg; + out_len = fcopy_transaction.recv_len; + break; + } + + fcopy_transaction.state = HVUTIL_USERSPACE_REQ; + rc = hvutil_transport_send(hvt, out_src, out_len, NULL); + if (rc) { + pr_debug("FCP: failed to communicate to the daemon: %d\n", rc); + if (cancel_delayed_work_sync(&fcopy_timeout_work)) { + fcopy_respond_to_host(HV_E_FAIL); + fcopy_transaction.state = HVUTIL_READY; + } + } + kfree(smsg_out); +} + +/* + * Send a response back to the host. + */ + +static void +fcopy_respond_to_host(int error) +{ + struct icmsg_hdr *icmsghdr; + u32 buf_len; + struct vmbus_channel *channel; + u64 req_id; + + /* + * Copy the global state for completing the transaction. Note that + * only one transaction can be active at a time. This is guaranteed + * by the file copy protocol implemented by the host. Furthermore, + * the "transaction active" state we maintain ensures that there can + * only be one active transaction at a time. + */ + + buf_len = fcopy_transaction.recv_len; + channel = fcopy_transaction.recv_channel; + req_id = fcopy_transaction.recv_req_id; + + icmsghdr = (struct icmsg_hdr *) + &recv_buffer[sizeof(struct vmbuspipe_hdr)]; + + if (channel->onchannel_callback == NULL) + /* + * We have raced with util driver being unloaded; + * silently return. + */ + return; + + icmsghdr->status = error; + icmsghdr->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + vmbus_sendpacket(channel, recv_buffer, buf_len, req_id, + VM_PKT_DATA_INBAND, 0); +} + +void hv_fcopy_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + struct hv_fcopy_hdr *fcopy_msg; + struct icmsg_hdr *icmsghdr; + int fcopy_srv_version; + + if (fcopy_transaction.state > HVUTIL_READY) + return; + + vmbus_recvpacket(channel, recv_buffer, PAGE_SIZE * 2, &recvlen, + &requestid); + if (recvlen <= 0) + return; + + icmsghdr = (struct icmsg_hdr *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr)]; + if (icmsghdr->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdr, recv_buffer, + fw_versions, FW_VER_COUNT, + fcopy_versions, FCOPY_VER_COUNT, + NULL, &fcopy_srv_version)) { + + pr_info("FCopy IC version %d.%d\n", + fcopy_srv_version >> 16, + fcopy_srv_version & 0xFFFF); + } + } else { + fcopy_msg = (struct hv_fcopy_hdr *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + /* + * Stash away this global state for completing the + * transaction; note transactions are serialized. + */ + + fcopy_transaction.recv_len = recvlen; + fcopy_transaction.recv_req_id = requestid; + fcopy_transaction.fcopy_msg = fcopy_msg; + + if (fcopy_transaction.state < HVUTIL_READY) { + /* Userspace is not registered yet */ + fcopy_respond_to_host(HV_E_FAIL); + return; + } + fcopy_transaction.state = HVUTIL_HOSTMSG_RECEIVED; + + /* + * Send the information to the user-level daemon. + */ + schedule_work(&fcopy_send_work); + schedule_delayed_work(&fcopy_timeout_work, + HV_UTIL_TIMEOUT * HZ); + return; + } + icmsghdr->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + vmbus_sendpacket(channel, recv_buffer, recvlen, requestid, + VM_PKT_DATA_INBAND, 0); +} + +/* Callback when data is received from userspace */ +static int fcopy_on_msg(void *msg, int len) +{ + int *val = (int *)msg; + + if (len != sizeof(int)) + return -EINVAL; + + if (fcopy_transaction.state == HVUTIL_DEVICE_INIT) + return fcopy_handle_handshake(*val); + + if (fcopy_transaction.state != HVUTIL_USERSPACE_REQ) + return -EINVAL; + + /* + * Complete the transaction by forwarding the result + * to the host. But first, cancel the timeout. + */ + if (cancel_delayed_work_sync(&fcopy_timeout_work)) { + fcopy_transaction.state = HVUTIL_USERSPACE_RECV; + fcopy_respond_to_host(*val); + hv_poll_channel(fcopy_transaction.recv_channel, + fcopy_poll_wrapper); + } + + return 0; +} + +static void fcopy_on_reset(void) +{ + /* + * The daemon has exited; reset the state. + */ + fcopy_transaction.state = HVUTIL_DEVICE_INIT; + + if (cancel_delayed_work_sync(&fcopy_timeout_work)) + fcopy_respond_to_host(HV_E_FAIL); +} + +int hv_fcopy_init(struct hv_util_service *srv) +{ + recv_buffer = srv->recv_buffer; + fcopy_transaction.recv_channel = srv->channel; + + /* + * When this driver loads, the user level daemon that + * processes the host requests may not yet be running. + * Defer processing channel callbacks until the daemon + * has registered. + */ + fcopy_transaction.state = HVUTIL_DEVICE_INIT; + + hvt = hvutil_transport_init(fcopy_devname, 0, 0, + fcopy_on_msg, fcopy_on_reset); + if (!hvt) + return -EFAULT; + + return 0; +} + +void hv_fcopy_deinit(void) +{ + fcopy_transaction.state = HVUTIL_DEVICE_DYING; + cancel_delayed_work_sync(&fcopy_timeout_work); + hvutil_transport_destroy(hvt); +} diff --git a/drivers/hv/hv_kvp.c b/drivers/hv/hv_kvp.c new file mode 100644 index 000000000..d6106e1a0 --- /dev/null +++ b/drivers/hv/hv_kvp.c @@ -0,0 +1,767 @@ +/* + * An implementation of key value pair (KVP) functionality for Linux. + * + * + * Copyright (C) 2010, Novell, Inc. + * Author : K. Y. Srinivasan <ksrinivasan@novell.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. 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., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/net.h> +#include <linux/nls.h> +#include <linux/connector.h> +#include <linux/workqueue.h> +#include <linux/hyperv.h> + +#include "hyperv_vmbus.h" +#include "hv_utils_transport.h" + +/* + * Pre win8 version numbers used in ws2008 and ws 2008 r2 (win7) + */ +#define WS2008_SRV_MAJOR 1 +#define WS2008_SRV_MINOR 0 +#define WS2008_SRV_VERSION (WS2008_SRV_MAJOR << 16 | WS2008_SRV_MINOR) + +#define WIN7_SRV_MAJOR 3 +#define WIN7_SRV_MINOR 0 +#define WIN7_SRV_VERSION (WIN7_SRV_MAJOR << 16 | WIN7_SRV_MINOR) + +#define WIN8_SRV_MAJOR 4 +#define WIN8_SRV_MINOR 0 +#define WIN8_SRV_VERSION (WIN8_SRV_MAJOR << 16 | WIN8_SRV_MINOR) + +#define KVP_VER_COUNT 3 +static const int kvp_versions[] = { + WIN8_SRV_VERSION, + WIN7_SRV_VERSION, + WS2008_SRV_VERSION +}; + +#define FW_VER_COUNT 2 +static const int fw_versions[] = { + UTIL_FW_VERSION, + UTIL_WS2K8_FW_VERSION +}; + +/* + * Global state maintained for transaction that is being processed. For a class + * of integration services, including the "KVP service", the specified protocol + * is a "request/response" protocol which means that there can only be single + * outstanding transaction from the host at any given point in time. We use + * this to simplify memory management in this driver - we cache and process + * only one message at a time. + * + * While the request/response protocol is guaranteed by the host, we further + * ensure this by serializing packet processing in this driver - we do not + * read additional packets from the VMBUS until the current packet is fully + * handled. + */ + +static struct { + int state; /* hvutil_device_state */ + int recv_len; /* number of bytes received. */ + struct hv_kvp_msg *kvp_msg; /* current message */ + struct vmbus_channel *recv_channel; /* chn we got the request */ + u64 recv_req_id; /* request ID. */ +} kvp_transaction; + +/* + * This state maintains the version number registered by the daemon. + */ +static int dm_reg_value; + +static void kvp_send_key(struct work_struct *dummy); + + +static void kvp_respond_to_host(struct hv_kvp_msg *msg, int error); +static void kvp_timeout_func(struct work_struct *dummy); +static void kvp_host_handshake_func(struct work_struct *dummy); +static void kvp_register(int); + +static DECLARE_DELAYED_WORK(kvp_timeout_work, kvp_timeout_func); +static DECLARE_DELAYED_WORK(kvp_host_handshake_work, kvp_host_handshake_func); +static DECLARE_WORK(kvp_sendkey_work, kvp_send_key); + +static const char kvp_devname[] = "vmbus/hv_kvp"; +static u8 *recv_buffer; +static struct hvutil_transport *hvt; +/* + * Register the kernel component with the user-level daemon. + * As part of this registration, pass the LIC version number. + * This number has no meaning, it satisfies the registration protocol. + */ +#define HV_DRV_VERSION "3.1" + +static void kvp_poll_wrapper(void *channel) +{ + /* Transaction is finished, reset the state here to avoid races. */ + kvp_transaction.state = HVUTIL_READY; + tasklet_schedule(&((struct vmbus_channel *)channel)->callback_event); +} + +static void kvp_register_done(void) +{ + /* + * If we're still negotiating with the host cancel the timeout + * work to not poll the channel twice. + */ + pr_debug("KVP: userspace daemon registered\n"); + cancel_delayed_work_sync(&kvp_host_handshake_work); + hv_poll_channel(kvp_transaction.recv_channel, kvp_poll_wrapper); +} + +static void +kvp_register(int reg_value) +{ + + struct hv_kvp_msg *kvp_msg; + char *version; + + kvp_msg = kzalloc(sizeof(*kvp_msg), GFP_KERNEL); + + if (kvp_msg) { + version = kvp_msg->body.kvp_register.version; + kvp_msg->kvp_hdr.operation = reg_value; + strcpy(version, HV_DRV_VERSION); + + hvutil_transport_send(hvt, kvp_msg, sizeof(*kvp_msg), + kvp_register_done); + kfree(kvp_msg); + } +} + +static void kvp_timeout_func(struct work_struct *dummy) +{ + /* + * If the timer fires, the user-mode component has not responded; + * process the pending transaction. + */ + kvp_respond_to_host(NULL, HV_E_FAIL); + + hv_poll_channel(kvp_transaction.recv_channel, kvp_poll_wrapper); +} + +static void kvp_host_handshake_func(struct work_struct *dummy) +{ + tasklet_schedule(&kvp_transaction.recv_channel->callback_event); +} + +static int kvp_handle_handshake(struct hv_kvp_msg *msg) +{ + switch (msg->kvp_hdr.operation) { + case KVP_OP_REGISTER: + dm_reg_value = KVP_OP_REGISTER; + pr_info("KVP: IP injection functionality not available\n"); + pr_info("KVP: Upgrade the KVP daemon\n"); + break; + case KVP_OP_REGISTER1: + dm_reg_value = KVP_OP_REGISTER1; + break; + default: + pr_info("KVP: incompatible daemon\n"); + pr_info("KVP: KVP version: %d, Daemon version: %d\n", + KVP_OP_REGISTER1, msg->kvp_hdr.operation); + return -EINVAL; + } + + /* + * We have a compatible daemon; complete the handshake. + */ + pr_debug("KVP: userspace daemon ver. %d connected\n", + msg->kvp_hdr.operation); + kvp_register(dm_reg_value); + + return 0; +} + + +/* + * Callback when data is received from user mode. + */ + +static int kvp_on_msg(void *msg, int len) +{ + struct hv_kvp_msg *message = (struct hv_kvp_msg *)msg; + struct hv_kvp_msg_enumerate *data; + int error = 0; + + if (len < sizeof(*message)) + return -EINVAL; + + /* + * If we are negotiating the version information + * with the daemon; handle that first. + */ + + if (kvp_transaction.state < HVUTIL_READY) { + return kvp_handle_handshake(message); + } + + /* We didn't send anything to userspace so the reply is spurious */ + if (kvp_transaction.state < HVUTIL_USERSPACE_REQ) + return -EINVAL; + + kvp_transaction.state = HVUTIL_USERSPACE_RECV; + + /* + * Based on the version of the daemon, we propagate errors from the + * daemon differently. + */ + + data = &message->body.kvp_enum_data; + + switch (dm_reg_value) { + case KVP_OP_REGISTER: + /* + * Null string is used to pass back error condition. + */ + if (data->data.key[0] == 0) + error = HV_S_CONT; + break; + + case KVP_OP_REGISTER1: + /* + * We use the message header information from + * the user level daemon to transmit errors. + */ + error = message->error; + break; + } + + /* + * Complete the transaction by forwarding the key value + * to the host. But first, cancel the timeout. + */ + if (cancel_delayed_work_sync(&kvp_timeout_work)) { + kvp_respond_to_host(message, error); + hv_poll_channel(kvp_transaction.recv_channel, kvp_poll_wrapper); + } + + return 0; +} + + +static int process_ob_ipinfo(void *in_msg, void *out_msg, int op) +{ + struct hv_kvp_msg *in = in_msg; + struct hv_kvp_ip_msg *out = out_msg; + int len; + + switch (op) { + case KVP_OP_GET_IP_INFO: + /* + * Transform all parameters into utf16 encoding. + */ + len = utf8s_to_utf16s((char *)in->body.kvp_ip_val.ip_addr, + strlen((char *)in->body.kvp_ip_val.ip_addr), + UTF16_HOST_ENDIAN, + (wchar_t *)out->kvp_ip_val.ip_addr, + MAX_IP_ADDR_SIZE); + if (len < 0) + return len; + + len = utf8s_to_utf16s((char *)in->body.kvp_ip_val.sub_net, + strlen((char *)in->body.kvp_ip_val.sub_net), + UTF16_HOST_ENDIAN, + (wchar_t *)out->kvp_ip_val.sub_net, + MAX_IP_ADDR_SIZE); + if (len < 0) + return len; + + len = utf8s_to_utf16s((char *)in->body.kvp_ip_val.gate_way, + strlen((char *)in->body.kvp_ip_val.gate_way), + UTF16_HOST_ENDIAN, + (wchar_t *)out->kvp_ip_val.gate_way, + MAX_GATEWAY_SIZE); + if (len < 0) + return len; + + len = utf8s_to_utf16s((char *)in->body.kvp_ip_val.dns_addr, + strlen((char *)in->body.kvp_ip_val.dns_addr), + UTF16_HOST_ENDIAN, + (wchar_t *)out->kvp_ip_val.dns_addr, + MAX_IP_ADDR_SIZE); + if (len < 0) + return len; + + len = utf8s_to_utf16s((char *)in->body.kvp_ip_val.adapter_id, + strlen((char *)in->body.kvp_ip_val.adapter_id), + UTF16_HOST_ENDIAN, + (wchar_t *)out->kvp_ip_val.adapter_id, + MAX_ADAPTER_ID_SIZE); + if (len < 0) + return len; + + out->kvp_ip_val.dhcp_enabled = + in->body.kvp_ip_val.dhcp_enabled; + out->kvp_ip_val.addr_family = + in->body.kvp_ip_val.addr_family; + } + + return 0; +} + +static void process_ib_ipinfo(void *in_msg, void *out_msg, int op) +{ + struct hv_kvp_ip_msg *in = in_msg; + struct hv_kvp_msg *out = out_msg; + + switch (op) { + case KVP_OP_SET_IP_INFO: + /* + * Transform all parameters into utf8 encoding. + */ + utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.ip_addr, + MAX_IP_ADDR_SIZE, + UTF16_LITTLE_ENDIAN, + (__u8 *)out->body.kvp_ip_val.ip_addr, + MAX_IP_ADDR_SIZE); + + utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.sub_net, + MAX_IP_ADDR_SIZE, + UTF16_LITTLE_ENDIAN, + (__u8 *)out->body.kvp_ip_val.sub_net, + MAX_IP_ADDR_SIZE); + + utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.gate_way, + MAX_GATEWAY_SIZE, + UTF16_LITTLE_ENDIAN, + (__u8 *)out->body.kvp_ip_val.gate_way, + MAX_GATEWAY_SIZE); + + utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.dns_addr, + MAX_IP_ADDR_SIZE, + UTF16_LITTLE_ENDIAN, + (__u8 *)out->body.kvp_ip_val.dns_addr, + MAX_IP_ADDR_SIZE); + + out->body.kvp_ip_val.dhcp_enabled = in->kvp_ip_val.dhcp_enabled; + + /* fallthrough */ + + case KVP_OP_GET_IP_INFO: + utf16s_to_utf8s((wchar_t *)in->kvp_ip_val.adapter_id, + MAX_ADAPTER_ID_SIZE, + UTF16_LITTLE_ENDIAN, + (__u8 *)out->body.kvp_ip_val.adapter_id, + MAX_ADAPTER_ID_SIZE); + + out->body.kvp_ip_val.addr_family = in->kvp_ip_val.addr_family; + } +} + + + + +static void +kvp_send_key(struct work_struct *dummy) +{ + struct hv_kvp_msg *message; + struct hv_kvp_msg *in_msg; + __u8 operation = kvp_transaction.kvp_msg->kvp_hdr.operation; + __u8 pool = kvp_transaction.kvp_msg->kvp_hdr.pool; + __u32 val32; + __u64 val64; + int rc; + + /* The transaction state is wrong. */ + if (kvp_transaction.state != HVUTIL_HOSTMSG_RECEIVED) + return; + + message = kzalloc(sizeof(*message), GFP_KERNEL); + if (!message) + return; + + message->kvp_hdr.operation = operation; + message->kvp_hdr.pool = pool; + in_msg = kvp_transaction.kvp_msg; + + /* + * The key/value strings sent from the host are encoded in + * in utf16; convert it to utf8 strings. + * The host assures us that the utf16 strings will not exceed + * the max lengths specified. We will however, reserve room + * for the string terminating character - in the utf16s_utf8s() + * function we limit the size of the buffer where the converted + * string is placed to HV_KVP_EXCHANGE_MAX_*_SIZE -1 to guarantee + * that the strings can be properly terminated! + */ + + switch (message->kvp_hdr.operation) { + case KVP_OP_SET_IP_INFO: + process_ib_ipinfo(in_msg, message, KVP_OP_SET_IP_INFO); + break; + case KVP_OP_GET_IP_INFO: + /* + * We only need to pass on the info of operation, adapter_id + * and addr_family to the userland kvp daemon. + */ + process_ib_ipinfo(in_msg, message, KVP_OP_GET_IP_INFO); + break; + case KVP_OP_SET: + switch (in_msg->body.kvp_set.data.value_type) { + case REG_SZ: + /* + * The value is a string - utf16 encoding. + */ + message->body.kvp_set.data.value_size = + utf16s_to_utf8s( + (wchar_t *)in_msg->body.kvp_set.data.value, + in_msg->body.kvp_set.data.value_size, + UTF16_LITTLE_ENDIAN, + message->body.kvp_set.data.value, + HV_KVP_EXCHANGE_MAX_VALUE_SIZE - 1) + 1; + break; + + case REG_U32: + /* + * The value is a 32 bit scalar. + * We save this as a utf8 string. + */ + val32 = in_msg->body.kvp_set.data.value_u32; + message->body.kvp_set.data.value_size = + sprintf(message->body.kvp_set.data.value, + "%d", val32) + 1; + break; + + case REG_U64: + /* + * The value is a 64 bit scalar. + * We save this as a utf8 string. + */ + val64 = in_msg->body.kvp_set.data.value_u64; + message->body.kvp_set.data.value_size = + sprintf(message->body.kvp_set.data.value, + "%llu", val64) + 1; + break; + + } + + /* + * The key is always a string - utf16 encoding. + */ + message->body.kvp_set.data.key_size = + utf16s_to_utf8s( + (wchar_t *)in_msg->body.kvp_set.data.key, + in_msg->body.kvp_set.data.key_size, + UTF16_LITTLE_ENDIAN, + message->body.kvp_set.data.key, + HV_KVP_EXCHANGE_MAX_KEY_SIZE - 1) + 1; + + break; + + case KVP_OP_GET: + message->body.kvp_get.data.key_size = + utf16s_to_utf8s( + (wchar_t *)in_msg->body.kvp_get.data.key, + in_msg->body.kvp_get.data.key_size, + UTF16_LITTLE_ENDIAN, + message->body.kvp_get.data.key, + HV_KVP_EXCHANGE_MAX_KEY_SIZE - 1) + 1; + break; + + case KVP_OP_DELETE: + message->body.kvp_delete.key_size = + utf16s_to_utf8s( + (wchar_t *)in_msg->body.kvp_delete.key, + in_msg->body.kvp_delete.key_size, + UTF16_LITTLE_ENDIAN, + message->body.kvp_delete.key, + HV_KVP_EXCHANGE_MAX_KEY_SIZE - 1) + 1; + break; + + case KVP_OP_ENUMERATE: + message->body.kvp_enum_data.index = + in_msg->body.kvp_enum_data.index; + break; + } + + kvp_transaction.state = HVUTIL_USERSPACE_REQ; + rc = hvutil_transport_send(hvt, message, sizeof(*message), NULL); + if (rc) { + pr_debug("KVP: failed to communicate to the daemon: %d\n", rc); + if (cancel_delayed_work_sync(&kvp_timeout_work)) { + kvp_respond_to_host(message, HV_E_FAIL); + kvp_transaction.state = HVUTIL_READY; + } + } + + kfree(message); +} + +/* + * Send a response back to the host. + */ + +static void +kvp_respond_to_host(struct hv_kvp_msg *msg_to_host, int error) +{ + struct hv_kvp_msg *kvp_msg; + struct hv_kvp_exchg_msg_value *kvp_data; + char *key_name; + char *value; + struct icmsg_hdr *icmsghdrp; + int keylen = 0; + int valuelen = 0; + u32 buf_len; + struct vmbus_channel *channel; + u64 req_id; + int ret; + + /* + * Copy the global state for completing the transaction. Note that + * only one transaction can be active at a time. + */ + + buf_len = kvp_transaction.recv_len; + channel = kvp_transaction.recv_channel; + req_id = kvp_transaction.recv_req_id; + + icmsghdrp = (struct icmsg_hdr *) + &recv_buffer[sizeof(struct vmbuspipe_hdr)]; + + if (channel->onchannel_callback == NULL) + /* + * We have raced with util driver being unloaded; + * silently return. + */ + return; + + icmsghdrp->status = error; + + /* + * If the error parameter is set, terminate the host's enumeration + * on this pool. + */ + if (error) { + /* + * Something failed or we have timed out; + * terminate the current host-side iteration. + */ + goto response_done; + } + + kvp_msg = (struct hv_kvp_msg *) + &recv_buffer[sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + switch (kvp_transaction.kvp_msg->kvp_hdr.operation) { + case KVP_OP_GET_IP_INFO: + ret = process_ob_ipinfo(msg_to_host, + (struct hv_kvp_ip_msg *)kvp_msg, + KVP_OP_GET_IP_INFO); + if (ret < 0) + icmsghdrp->status = HV_E_FAIL; + + goto response_done; + case KVP_OP_SET_IP_INFO: + goto response_done; + case KVP_OP_GET: + kvp_data = &kvp_msg->body.kvp_get.data; + goto copy_value; + + case KVP_OP_SET: + case KVP_OP_DELETE: + goto response_done; + + default: + break; + } + + kvp_data = &kvp_msg->body.kvp_enum_data.data; + key_name = msg_to_host->body.kvp_enum_data.data.key; + + /* + * The windows host expects the key/value pair to be encoded + * in utf16. Ensure that the key/value size reported to the host + * will be less than or equal to the MAX size (including the + * terminating character). + */ + keylen = utf8s_to_utf16s(key_name, strlen(key_name), UTF16_HOST_ENDIAN, + (wchar_t *) kvp_data->key, + (HV_KVP_EXCHANGE_MAX_KEY_SIZE / 2) - 2); + kvp_data->key_size = 2*(keylen + 1); /* utf16 encoding */ + +copy_value: + value = msg_to_host->body.kvp_enum_data.data.value; + valuelen = utf8s_to_utf16s(value, strlen(value), UTF16_HOST_ENDIAN, + (wchar_t *) kvp_data->value, + (HV_KVP_EXCHANGE_MAX_VALUE_SIZE / 2) - 2); + kvp_data->value_size = 2*(valuelen + 1); /* utf16 encoding */ + + /* + * If the utf8s to utf16s conversion failed; notify host + * of the error. + */ + if ((keylen < 0) || (valuelen < 0)) + icmsghdrp->status = HV_E_FAIL; + + kvp_data->value_type = REG_SZ; /* all our values are strings */ + +response_done: + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, buf_len, req_id, + VM_PKT_DATA_INBAND, 0); +} + +/* + * This callback is invoked when we get a KVP message from the host. + * The host ensures that only one KVP transaction can be active at a time. + * KVP implementation in Linux needs to forward the key to a user-mde + * component to retrieve the corresponding value. Consequently, we cannot + * respond to the host in the context of this callback. Since the host + * guarantees that at most only one transaction can be active at a time, + * we stash away the transaction state in a set of global variables. + */ + +void hv_kvp_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + + struct hv_kvp_msg *kvp_msg; + + struct icmsg_hdr *icmsghdrp; + int kvp_srv_version; + static enum {NEGO_NOT_STARTED, + NEGO_IN_PROGRESS, + NEGO_FINISHED} host_negotiatied = NEGO_NOT_STARTED; + + if (kvp_transaction.state < HVUTIL_READY) { + /* + * If userspace daemon is not connected and host is asking + * us to negotiate we need to delay to not lose messages. + * This is important for Failover IP setting. + */ + if (host_negotiatied == NEGO_NOT_STARTED) { + host_negotiatied = NEGO_IN_PROGRESS; + schedule_delayed_work(&kvp_host_handshake_work, + HV_UTIL_NEGO_TIMEOUT * HZ); + } + return; + } + if (kvp_transaction.state > HVUTIL_READY) + return; + + vmbus_recvpacket(channel, recv_buffer, PAGE_SIZE * 4, &recvlen, + &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdrp, + recv_buffer, fw_versions, FW_VER_COUNT, + kvp_versions, KVP_VER_COUNT, + NULL, &kvp_srv_version)) { + pr_info("KVP IC version %d.%d\n", + kvp_srv_version >> 16, + kvp_srv_version & 0xFFFF); + } + } else { + kvp_msg = (struct hv_kvp_msg *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + /* + * Stash away this global state for completing the + * transaction; note transactions are serialized. + */ + + kvp_transaction.recv_len = recvlen; + kvp_transaction.recv_req_id = requestid; + kvp_transaction.kvp_msg = kvp_msg; + + if (kvp_transaction.state < HVUTIL_READY) { + /* Userspace is not registered yet */ + kvp_respond_to_host(NULL, HV_E_FAIL); + return; + } + kvp_transaction.state = HVUTIL_HOSTMSG_RECEIVED; + + /* + * Get the information from the + * user-mode component. + * component. This transaction will be + * completed when we get the value from + * the user-mode component. + * Set a timeout to deal with + * user-mode not responding. + */ + schedule_work(&kvp_sendkey_work); + schedule_delayed_work(&kvp_timeout_work, + HV_UTIL_TIMEOUT * HZ); + + return; + + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + + host_negotiatied = NEGO_FINISHED; + hv_poll_channel(kvp_transaction.recv_channel, kvp_poll_wrapper); + } + +} + +static void kvp_on_reset(void) +{ + if (cancel_delayed_work_sync(&kvp_timeout_work)) + kvp_respond_to_host(NULL, HV_E_FAIL); + kvp_transaction.state = HVUTIL_DEVICE_INIT; +} + +int +hv_kvp_init(struct hv_util_service *srv) +{ + recv_buffer = srv->recv_buffer; + kvp_transaction.recv_channel = srv->channel; + + /* + * When this driver loads, the user level daemon that + * processes the host requests may not yet be running. + * Defer processing channel callbacks until the daemon + * has registered. + */ + kvp_transaction.state = HVUTIL_DEVICE_INIT; + + hvt = hvutil_transport_init(kvp_devname, CN_KVP_IDX, CN_KVP_VAL, + kvp_on_msg, kvp_on_reset); + if (!hvt) + return -EFAULT; + + return 0; +} + +void hv_kvp_deinit(void) +{ + kvp_transaction.state = HVUTIL_DEVICE_DYING; + cancel_delayed_work_sync(&kvp_host_handshake_work); + cancel_delayed_work_sync(&kvp_timeout_work); + cancel_work_sync(&kvp_sendkey_work); + hvutil_transport_destroy(hvt); +} diff --git a/drivers/hv/hv_snapshot.c b/drivers/hv/hv_snapshot.c new file mode 100644 index 000000000..6831efd73 --- /dev/null +++ b/drivers/hv/hv_snapshot.c @@ -0,0 +1,398 @@ +/* + * An implementation of host initiated guest snapshot. + * + * + * Copyright (C) 2013, Microsoft, Inc. + * Author : K. Y. Srinivasan <kys@microsoft.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/net.h> +#include <linux/nls.h> +#include <linux/connector.h> +#include <linux/workqueue.h> +#include <linux/hyperv.h> + +#include "hyperv_vmbus.h" +#include "hv_utils_transport.h" + +#define VSS_MAJOR 5 +#define VSS_MINOR 0 +#define VSS_VERSION (VSS_MAJOR << 16 | VSS_MINOR) + +#define VSS_VER_COUNT 1 +static const int vss_versions[] = { + VSS_VERSION +}; + +#define FW_VER_COUNT 1 +static const int fw_versions[] = { + UTIL_FW_VERSION +}; + +/* + * Timeout values are based on expecations from host + */ +#define VSS_FREEZE_TIMEOUT (15 * 60) + +/* + * Global state maintained for transaction that is being processed. For a class + * of integration services, including the "VSS service", the specified protocol + * is a "request/response" protocol which means that there can only be single + * outstanding transaction from the host at any given point in time. We use + * this to simplify memory management in this driver - we cache and process + * only one message at a time. + * + * While the request/response protocol is guaranteed by the host, we further + * ensure this by serializing packet processing in this driver - we do not + * read additional packets from the VMBUs until the current packet is fully + * handled. + */ + +static struct { + int state; /* hvutil_device_state */ + int recv_len; /* number of bytes received. */ + struct vmbus_channel *recv_channel; /* chn we got the request */ + u64 recv_req_id; /* request ID. */ + struct hv_vss_msg *msg; /* current message */ +} vss_transaction; + + +static void vss_respond_to_host(int error); + +/* + * This state maintains the version number registered by the daemon. + */ +static int dm_reg_value; + +static const char vss_devname[] = "vmbus/hv_vss"; +static __u8 *recv_buffer; +static struct hvutil_transport *hvt; + +static void vss_timeout_func(struct work_struct *dummy); +static void vss_handle_request(struct work_struct *dummy); + +static DECLARE_DELAYED_WORK(vss_timeout_work, vss_timeout_func); +static DECLARE_WORK(vss_handle_request_work, vss_handle_request); + +static void vss_poll_wrapper(void *channel) +{ + /* Transaction is finished, reset the state here to avoid races. */ + vss_transaction.state = HVUTIL_READY; + hv_vss_onchannelcallback(channel); +} + +/* + * Callback when data is received from user mode. + */ + +static void vss_timeout_func(struct work_struct *dummy) +{ + /* + * Timeout waiting for userspace component to reply happened. + */ + pr_warn("VSS: timeout waiting for daemon to reply\n"); + vss_respond_to_host(HV_E_FAIL); + + hv_poll_channel(vss_transaction.recv_channel, vss_poll_wrapper); +} + +static void vss_register_done(void) +{ + hv_poll_channel(vss_transaction.recv_channel, vss_poll_wrapper); + pr_debug("VSS: userspace daemon registered\n"); +} + +static int vss_handle_handshake(struct hv_vss_msg *vss_msg) +{ + u32 our_ver = VSS_OP_REGISTER1; + + switch (vss_msg->vss_hdr.operation) { + case VSS_OP_REGISTER: + /* Daemon doesn't expect us to reply */ + dm_reg_value = VSS_OP_REGISTER; + break; + case VSS_OP_REGISTER1: + /* Daemon expects us to reply with our own version */ + if (hvutil_transport_send(hvt, &our_ver, sizeof(our_ver), + vss_register_done)) + return -EFAULT; + dm_reg_value = VSS_OP_REGISTER1; + break; + default: + return -EINVAL; + } + pr_info("VSS: userspace daemon ver. %d connected\n", dm_reg_value); + return 0; +} + +static int vss_on_msg(void *msg, int len) +{ + struct hv_vss_msg *vss_msg = (struct hv_vss_msg *)msg; + + if (len != sizeof(*vss_msg)) { + pr_debug("VSS: Message size does not match length\n"); + return -EINVAL; + } + + if (vss_msg->vss_hdr.operation == VSS_OP_REGISTER || + vss_msg->vss_hdr.operation == VSS_OP_REGISTER1) { + /* + * Don't process registration messages if we're in the middle + * of a transaction processing. + */ + if (vss_transaction.state > HVUTIL_READY) { + pr_debug("VSS: Got unexpected registration request\n"); + return -EINVAL; + } + + return vss_handle_handshake(vss_msg); + } else if (vss_transaction.state == HVUTIL_USERSPACE_REQ) { + vss_transaction.state = HVUTIL_USERSPACE_RECV; + + if (vss_msg->vss_hdr.operation == VSS_OP_HOT_BACKUP) + vss_transaction.msg->vss_cf.flags = + VSS_HBU_NO_AUTO_RECOVERY; + + if (cancel_delayed_work_sync(&vss_timeout_work)) { + vss_respond_to_host(vss_msg->error); + /* Transaction is finished, reset the state. */ + hv_poll_channel(vss_transaction.recv_channel, + vss_poll_wrapper); + } + } else { + /* This is a spurious call! */ + pr_debug("VSS: Transaction not active\n"); + return -EINVAL; + } + return 0; +} + +static void vss_send_op(void) +{ + int op = vss_transaction.msg->vss_hdr.operation; + int rc; + struct hv_vss_msg *vss_msg; + + /* The transaction state is wrong. */ + if (vss_transaction.state != HVUTIL_HOSTMSG_RECEIVED) { + pr_debug("VSS: Unexpected attempt to send to daemon\n"); + return; + } + + vss_msg = kzalloc(sizeof(*vss_msg), GFP_KERNEL); + if (!vss_msg) + return; + + vss_msg->vss_hdr.operation = op; + + vss_transaction.state = HVUTIL_USERSPACE_REQ; + + schedule_delayed_work(&vss_timeout_work, op == VSS_OP_FREEZE ? + VSS_FREEZE_TIMEOUT * HZ : HV_UTIL_TIMEOUT * HZ); + + rc = hvutil_transport_send(hvt, vss_msg, sizeof(*vss_msg), NULL); + if (rc) { + pr_warn("VSS: failed to communicate to the daemon: %d\n", rc); + if (cancel_delayed_work_sync(&vss_timeout_work)) { + vss_respond_to_host(HV_E_FAIL); + vss_transaction.state = HVUTIL_READY; + } + } + + kfree(vss_msg); +} + +static void vss_handle_request(struct work_struct *dummy) +{ + switch (vss_transaction.msg->vss_hdr.operation) { + /* + * Initiate a "freeze/thaw" operation in the guest. + * We respond to the host once the operation is complete. + * + * We send the message to the user space daemon and the operation is + * performed in the daemon. + */ + case VSS_OP_THAW: + case VSS_OP_FREEZE: + case VSS_OP_HOT_BACKUP: + if (vss_transaction.state < HVUTIL_READY) { + /* Userspace is not registered yet */ + pr_debug("VSS: Not ready for request.\n"); + vss_respond_to_host(HV_E_FAIL); + return; + } + + pr_debug("VSS: Received request for op code: %d\n", + vss_transaction.msg->vss_hdr.operation); + vss_transaction.state = HVUTIL_HOSTMSG_RECEIVED; + vss_send_op(); + return; + case VSS_OP_GET_DM_INFO: + vss_transaction.msg->dm_info.flags = 0; + break; + default: + break; + } + + vss_respond_to_host(0); + hv_poll_channel(vss_transaction.recv_channel, vss_poll_wrapper); +} + +/* + * Send a response back to the host. + */ + +static void +vss_respond_to_host(int error) +{ + struct icmsg_hdr *icmsghdrp; + u32 buf_len; + struct vmbus_channel *channel; + u64 req_id; + + /* + * Copy the global state for completing the transaction. Note that + * only one transaction can be active at a time. + */ + + buf_len = vss_transaction.recv_len; + channel = vss_transaction.recv_channel; + req_id = vss_transaction.recv_req_id; + + icmsghdrp = (struct icmsg_hdr *) + &recv_buffer[sizeof(struct vmbuspipe_hdr)]; + + if (channel->onchannel_callback == NULL) + /* + * We have raced with util driver being unloaded; + * silently return. + */ + return; + + icmsghdrp->status = error; + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, buf_len, req_id, + VM_PKT_DATA_INBAND, 0); + +} + +/* + * This callback is invoked when we get a VSS message from the host. + * The host ensures that only one VSS transaction can be active at a time. + */ + +void hv_vss_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + struct hv_vss_msg *vss_msg; + int vss_srv_version; + + struct icmsg_hdr *icmsghdrp; + + if (vss_transaction.state > HVUTIL_READY) + return; + + vmbus_recvpacket(channel, recv_buffer, PAGE_SIZE * 2, &recvlen, + &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdrp, + recv_buffer, fw_versions, FW_VER_COUNT, + vss_versions, VSS_VER_COUNT, + NULL, &vss_srv_version)) { + + pr_info("VSS IC version %d.%d\n", + vss_srv_version >> 16, + vss_srv_version & 0xFFFF); + } + } else { + vss_msg = (struct hv_vss_msg *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + /* + * Stash away this global state for completing the + * transaction; note transactions are serialized. + */ + + vss_transaction.recv_len = recvlen; + vss_transaction.recv_req_id = requestid; + vss_transaction.msg = (struct hv_vss_msg *)vss_msg; + + schedule_work(&vss_handle_request_work); + return; + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } + +} + +static void vss_on_reset(void) +{ + if (cancel_delayed_work_sync(&vss_timeout_work)) + vss_respond_to_host(HV_E_FAIL); + vss_transaction.state = HVUTIL_DEVICE_INIT; +} + +int +hv_vss_init(struct hv_util_service *srv) +{ + if (vmbus_proto_version < VERSION_WIN8_1) { + pr_warn("Integration service 'Backup (volume snapshot)'" + " not supported on this host version.\n"); + return -ENOTSUPP; + } + recv_buffer = srv->recv_buffer; + vss_transaction.recv_channel = srv->channel; + + /* + * When this driver loads, the user level daemon that + * processes the host requests may not yet be running. + * Defer processing channel callbacks until the daemon + * has registered. + */ + vss_transaction.state = HVUTIL_DEVICE_INIT; + + hvt = hvutil_transport_init(vss_devname, CN_VSS_IDX, CN_VSS_VAL, + vss_on_msg, vss_on_reset); + if (!hvt) { + pr_warn("VSS: Failed to initialize transport\n"); + return -EFAULT; + } + + return 0; +} + +void hv_vss_deinit(void) +{ + vss_transaction.state = HVUTIL_DEVICE_DYING; + cancel_delayed_work_sync(&vss_timeout_work); + cancel_work_sync(&vss_handle_request_work); + hvutil_transport_destroy(hvt); +} diff --git a/drivers/hv/hv_trace.c b/drivers/hv/hv_trace.c new file mode 100644 index 000000000..38d359cf1 --- /dev/null +++ b/drivers/hv/hv_trace.c @@ -0,0 +1,6 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include "hyperv_vmbus.h" + +#define CREATE_TRACE_POINTS +#include "hv_trace.h" diff --git a/drivers/hv/hv_trace.h b/drivers/hv/hv_trace.h new file mode 100644 index 000000000..999f80a63 --- /dev/null +++ b/drivers/hv/hv_trace.h @@ -0,0 +1,329 @@ +// SPDX-License-Identifier: GPL-2.0 + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM hyperv + +#if !defined(_HV_TRACE_H) || defined(TRACE_HEADER_MULTI_READ) +#define _HV_TRACE_H + +#include <linux/tracepoint.h> + +DECLARE_EVENT_CLASS(vmbus_hdr_msg, + TP_PROTO(const struct vmbus_channel_message_header *hdr), + TP_ARGS(hdr), + TP_STRUCT__entry(__field(unsigned int, msgtype)), + TP_fast_assign(__entry->msgtype = hdr->msgtype;), + TP_printk("msgtype=%u", __entry->msgtype) +); + +DEFINE_EVENT(vmbus_hdr_msg, vmbus_on_msg_dpc, + TP_PROTO(const struct vmbus_channel_message_header *hdr), + TP_ARGS(hdr) +); + +DEFINE_EVENT(vmbus_hdr_msg, vmbus_on_message, + TP_PROTO(const struct vmbus_channel_message_header *hdr), + TP_ARGS(hdr) +); + +TRACE_EVENT(vmbus_onoffer, + TP_PROTO(const struct vmbus_channel_offer_channel *offer), + TP_ARGS(offer), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u8, monitorid) + __field(u16, is_ddc_int) + __field(u32, connection_id) + __array(char, if_type, 16) + __array(char, if_instance, 16) + __field(u16, chn_flags) + __field(u16, mmio_mb) + __field(u16, sub_idx) + ), + TP_fast_assign(__entry->child_relid = offer->child_relid; + __entry->monitorid = offer->monitorid; + __entry->is_ddc_int = offer->is_dedicated_interrupt; + __entry->connection_id = offer->connection_id; + memcpy(__entry->if_type, + &offer->offer.if_type.b, 16); + memcpy(__entry->if_instance, + &offer->offer.if_instance.b, 16); + __entry->chn_flags = offer->offer.chn_flags; + __entry->mmio_mb = offer->offer.mmio_megabytes; + __entry->sub_idx = offer->offer.sub_channel_index; + ), + TP_printk("child_relid 0x%x, monitorid 0x%x, is_dedicated %d, " + "connection_id 0x%x, if_type %pUl, if_instance %pUl, " + "chn_flags 0x%x, mmio_megabytes %d, sub_channel_index %d", + __entry->child_relid, __entry->monitorid, + __entry->is_ddc_int, __entry->connection_id, + __entry->if_type, __entry->if_instance, + __entry->chn_flags, __entry->mmio_mb, + __entry->sub_idx + ) + ); + +TRACE_EVENT(vmbus_onoffer_rescind, + TP_PROTO(const struct vmbus_channel_rescind_offer *offer), + TP_ARGS(offer), + TP_STRUCT__entry(__field(u32, child_relid)), + TP_fast_assign(__entry->child_relid = offer->child_relid), + TP_printk("child_relid 0x%x", __entry->child_relid) + ); + +TRACE_EVENT(vmbus_onopen_result, + TP_PROTO(const struct vmbus_channel_open_result *result), + TP_ARGS(result), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u32, openid) + __field(u32, status) + ), + TP_fast_assign(__entry->child_relid = result->child_relid; + __entry->openid = result->openid; + __entry->status = result->status; + ), + TP_printk("child_relid 0x%x, openid %d, status %d", + __entry->child_relid, __entry->openid, __entry->status + ) + ); + +TRACE_EVENT(vmbus_ongpadl_created, + TP_PROTO(const struct vmbus_channel_gpadl_created *gpadlcreated), + TP_ARGS(gpadlcreated), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u32, gpadl) + __field(u32, status) + ), + TP_fast_assign(__entry->child_relid = gpadlcreated->child_relid; + __entry->gpadl = gpadlcreated->gpadl; + __entry->status = gpadlcreated->creation_status; + ), + TP_printk("child_relid 0x%x, gpadl 0x%x, creation_status %d", + __entry->child_relid, __entry->gpadl, __entry->status + ) + ); + +TRACE_EVENT(vmbus_ongpadl_torndown, + TP_PROTO(const struct vmbus_channel_gpadl_torndown *gpadltorndown), + TP_ARGS(gpadltorndown), + TP_STRUCT__entry(__field(u32, gpadl)), + TP_fast_assign(__entry->gpadl = gpadltorndown->gpadl), + TP_printk("gpadl 0x%x", __entry->gpadl) + ); + +TRACE_EVENT(vmbus_onversion_response, + TP_PROTO(const struct vmbus_channel_version_response *response), + TP_ARGS(response), + TP_STRUCT__entry( + __field(u8, ver) + ), + TP_fast_assign(__entry->ver = response->version_supported; + ), + TP_printk("version_supported %d", __entry->ver) + ); + +TRACE_EVENT(vmbus_request_offers, + TP_PROTO(int ret), + TP_ARGS(ret), + TP_STRUCT__entry(__field(int, ret)), + TP_fast_assign(__entry->ret = ret), + TP_printk("sending ret %d", __entry->ret) + ); + +TRACE_EVENT(vmbus_open, + TP_PROTO(const struct vmbus_channel_open_channel *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u32, openid) + __field(u32, gpadlhandle) + __field(u32, target_vp) + __field(u32, offset) + __field(int, ret) + ), + TP_fast_assign( + __entry->child_relid = msg->child_relid; + __entry->openid = msg->openid; + __entry->gpadlhandle = msg->ringbuffer_gpadlhandle; + __entry->target_vp = msg->target_vp; + __entry->offset = msg->downstream_ringbuffer_pageoffset; + __entry->ret = ret; + ), + TP_printk("sending child_relid 0x%x, openid %d, " + "gpadlhandle 0x%x, target_vp 0x%x, offset 0x%x, ret %d", + __entry->child_relid, __entry->openid, + __entry->gpadlhandle, __entry->target_vp, + __entry->offset, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_close_internal, + TP_PROTO(const struct vmbus_channel_close_channel *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(int, ret) + ), + TP_fast_assign( + __entry->child_relid = msg->child_relid; + __entry->ret = ret; + ), + TP_printk("sending child_relid 0x%x, ret %d", __entry->child_relid, + __entry->ret) + ); + +TRACE_EVENT(vmbus_establish_gpadl_header, + TP_PROTO(const struct vmbus_channel_gpadl_header *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u32, gpadl) + __field(u16, range_buflen) + __field(u16, rangecount) + __field(int, ret) + ), + TP_fast_assign( + __entry->child_relid = msg->child_relid; + __entry->gpadl = msg->gpadl; + __entry->range_buflen = msg->range_buflen; + __entry->rangecount = msg->rangecount; + __entry->ret = ret; + ), + TP_printk("sending child_relid 0x%x, gpadl 0x%x, range_buflen %d " + "rangecount %d, ret %d", + __entry->child_relid, __entry->gpadl, + __entry->range_buflen, __entry->rangecount, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_establish_gpadl_body, + TP_PROTO(const struct vmbus_channel_gpadl_body *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, msgnumber) + __field(u32, gpadl) + __field(int, ret) + ), + TP_fast_assign( + __entry->msgnumber = msg->msgnumber; + __entry->gpadl = msg->gpadl; + __entry->ret = ret; + ), + TP_printk("sending msgnumber %d, gpadl 0x%x, ret %d", + __entry->msgnumber, __entry->gpadl, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_teardown_gpadl, + TP_PROTO(const struct vmbus_channel_gpadl_teardown *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(u32, gpadl) + __field(int, ret) + ), + TP_fast_assign( + __entry->child_relid = msg->child_relid; + __entry->gpadl = msg->gpadl; + __entry->ret = ret; + ), + TP_printk("sending child_relid 0x%x, gpadl 0x%x, ret %d", + __entry->child_relid, __entry->gpadl, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_negotiate_version, + TP_PROTO(const struct vmbus_channel_initiate_contact *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, ver) + __field(u32, target_vcpu) + __field(int, ret) + __field(u64, int_page) + __field(u64, mon_page1) + __field(u64, mon_page2) + ), + TP_fast_assign( + __entry->ver = msg->vmbus_version_requested; + __entry->target_vcpu = msg->target_vcpu; + __entry->int_page = msg->interrupt_page; + __entry->mon_page1 = msg->monitor_page1; + __entry->mon_page2 = msg->monitor_page2; + __entry->ret = ret; + ), + TP_printk("sending vmbus_version_requested %d, target_vcpu 0x%x, " + "pages %llx:%llx:%llx, ret %d", + __entry->ver, __entry->target_vcpu, __entry->int_page, + __entry->mon_page1, __entry->mon_page2, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_release_relid, + TP_PROTO(const struct vmbus_channel_relid_released *msg, int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __field(u32, child_relid) + __field(int, ret) + ), + TP_fast_assign( + __entry->child_relid = msg->child_relid; + __entry->ret = ret; + ), + TP_printk("sending child_relid 0x%x, ret %d", + __entry->child_relid, __entry->ret + ) + ); + +TRACE_EVENT(vmbus_send_tl_connect_request, + TP_PROTO(const struct vmbus_channel_tl_connect_request *msg, + int ret), + TP_ARGS(msg, ret), + TP_STRUCT__entry( + __array(char, guest_id, 16) + __array(char, host_id, 16) + __field(int, ret) + ), + TP_fast_assign( + memcpy(__entry->guest_id, &msg->guest_endpoint_id.b, 16); + memcpy(__entry->host_id, &msg->host_service_id.b, 16); + __entry->ret = ret; + ), + TP_printk("sending guest_endpoint_id %pUl, host_service_id %pUl, " + "ret %d", + __entry->guest_id, __entry->host_id, __entry->ret + ) + ); + +DECLARE_EVENT_CLASS(vmbus_channel, + TP_PROTO(const struct vmbus_channel *channel), + TP_ARGS(channel), + TP_STRUCT__entry(__field(u32, relid)), + TP_fast_assign(__entry->relid = channel->offermsg.child_relid), + TP_printk("relid 0x%x", __entry->relid) +); + +DEFINE_EVENT(vmbus_channel, vmbus_chan_sched, + TP_PROTO(const struct vmbus_channel *channel), + TP_ARGS(channel) +); + +DEFINE_EVENT(vmbus_channel, vmbus_setevent, + TP_PROTO(const struct vmbus_channel *channel), + TP_ARGS(channel) +); + +DEFINE_EVENT(vmbus_channel, vmbus_on_event, + TP_PROTO(const struct vmbus_channel *channel), + TP_ARGS(channel) +); + +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_FILE hv_trace +#endif /* _HV_TRACE_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/drivers/hv/hv_trace_balloon.h b/drivers/hv/hv_trace_balloon.h new file mode 100644 index 000000000..93082888a --- /dev/null +++ b/drivers/hv/hv_trace_balloon.h @@ -0,0 +1,48 @@ +#undef TRACE_SYSTEM +#define TRACE_SYSTEM hyperv + +#if !defined(_HV_TRACE_BALLOON_H) || defined(TRACE_HEADER_MULTI_READ) +#define _HV_TRACE_BALLOON_H + +#include <linux/tracepoint.h> + +TRACE_EVENT(balloon_status, + TP_PROTO(u64 available, u64 committed, + unsigned long vm_memory_committed, + unsigned long pages_ballooned, + unsigned long pages_added, + unsigned long pages_onlined), + TP_ARGS(available, committed, vm_memory_committed, + pages_ballooned, pages_added, pages_onlined), + TP_STRUCT__entry( + __field(u64, available) + __field(u64, committed) + __field(unsigned long, vm_memory_committed) + __field(unsigned long, pages_ballooned) + __field(unsigned long, pages_added) + __field(unsigned long, pages_onlined) + ), + TP_fast_assign( + __entry->available = available; + __entry->committed = committed; + __entry->vm_memory_committed = vm_memory_committed; + __entry->pages_ballooned = pages_ballooned; + __entry->pages_added = pages_added; + __entry->pages_onlined = pages_onlined; + ), + TP_printk("available %lld, committed %lld; vm_memory_committed %ld;" + " pages_ballooned %ld, pages_added %ld, pages_onlined %ld", + __entry->available, __entry->committed, + __entry->vm_memory_committed, __entry->pages_ballooned, + __entry->pages_added, __entry->pages_onlined + ) + ); + +#undef TRACE_INCLUDE_PATH +#define TRACE_INCLUDE_PATH . +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_FILE hv_trace_balloon +#endif /* _HV_TRACE_BALLOON_H */ + +/* This part must be outside protection */ +#include <trace/define_trace.h> diff --git a/drivers/hv/hv_util.c b/drivers/hv/hv_util.c new file mode 100644 index 000000000..2003314dc --- /dev/null +++ b/drivers/hv/hv_util.c @@ -0,0 +1,584 @@ +/* + * Copyright (c) 2010, 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> + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/kernel.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/sysctl.h> +#include <linux/reboot.h> +#include <linux/hyperv.h> +#include <linux/clockchips.h> +#include <linux/ptp_clock_kernel.h> +#include <asm/mshyperv.h> + +#include "hyperv_vmbus.h" + +#define SD_MAJOR 3 +#define SD_MINOR 0 +#define SD_VERSION (SD_MAJOR << 16 | SD_MINOR) + +#define SD_MAJOR_1 1 +#define SD_VERSION_1 (SD_MAJOR_1 << 16 | SD_MINOR) + +#define TS_MAJOR 4 +#define TS_MINOR 0 +#define TS_VERSION (TS_MAJOR << 16 | TS_MINOR) + +#define TS_MAJOR_1 1 +#define TS_VERSION_1 (TS_MAJOR_1 << 16 | TS_MINOR) + +#define TS_MAJOR_3 3 +#define TS_VERSION_3 (TS_MAJOR_3 << 16 | TS_MINOR) + +#define HB_MAJOR 3 +#define HB_MINOR 0 +#define HB_VERSION (HB_MAJOR << 16 | HB_MINOR) + +#define HB_MAJOR_1 1 +#define HB_VERSION_1 (HB_MAJOR_1 << 16 | HB_MINOR) + +static int sd_srv_version; +static int ts_srv_version; +static int hb_srv_version; + +#define SD_VER_COUNT 2 +static const int sd_versions[] = { + SD_VERSION, + SD_VERSION_1 +}; + +#define TS_VER_COUNT 3 +static const int ts_versions[] = { + TS_VERSION, + TS_VERSION_3, + TS_VERSION_1 +}; + +#define HB_VER_COUNT 2 +static const int hb_versions[] = { + HB_VERSION, + HB_VERSION_1 +}; + +#define FW_VER_COUNT 2 +static const int fw_versions[] = { + UTIL_FW_VERSION, + UTIL_WS2K8_FW_VERSION +}; + +static void shutdown_onchannelcallback(void *context); +static struct hv_util_service util_shutdown = { + .util_cb = shutdown_onchannelcallback, +}; + +static int hv_timesync_init(struct hv_util_service *srv); +static void hv_timesync_deinit(void); + +static void timesync_onchannelcallback(void *context); +static struct hv_util_service util_timesynch = { + .util_cb = timesync_onchannelcallback, + .util_init = hv_timesync_init, + .util_deinit = hv_timesync_deinit, +}; + +static void heartbeat_onchannelcallback(void *context); +static struct hv_util_service util_heartbeat = { + .util_cb = heartbeat_onchannelcallback, +}; + +static struct hv_util_service util_kvp = { + .util_cb = hv_kvp_onchannelcallback, + .util_init = hv_kvp_init, + .util_deinit = hv_kvp_deinit, +}; + +static struct hv_util_service util_vss = { + .util_cb = hv_vss_onchannelcallback, + .util_init = hv_vss_init, + .util_deinit = hv_vss_deinit, +}; + +static struct hv_util_service util_fcopy = { + .util_cb = hv_fcopy_onchannelcallback, + .util_init = hv_fcopy_init, + .util_deinit = hv_fcopy_deinit, +}; + +static void perform_shutdown(struct work_struct *dummy) +{ + orderly_poweroff(true); +} + +/* + * Perform the shutdown operation in a thread context. + */ +static DECLARE_WORK(shutdown_work, perform_shutdown); + +static void shutdown_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + bool execute_shutdown = false; + u8 *shut_txf_buf = util_shutdown.recv_buffer; + + struct shutdown_msg_data *shutdown_msg; + + struct icmsg_hdr *icmsghdrp; + + vmbus_recvpacket(channel, shut_txf_buf, + PAGE_SIZE, &recvlen, &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&shut_txf_buf[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdrp, shut_txf_buf, + fw_versions, FW_VER_COUNT, + sd_versions, SD_VER_COUNT, + NULL, &sd_srv_version)) { + pr_info("Shutdown IC version %d.%d\n", + sd_srv_version >> 16, + sd_srv_version & 0xFFFF); + } + } else { + shutdown_msg = + (struct shutdown_msg_data *)&shut_txf_buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + switch (shutdown_msg->flags) { + case 0: + case 1: + icmsghdrp->status = HV_S_OK; + execute_shutdown = true; + + pr_info("Shutdown request received -" + " graceful shutdown initiated\n"); + break; + default: + icmsghdrp->status = HV_E_FAIL; + execute_shutdown = false; + + pr_info("Shutdown request received -" + " Invalid request\n"); + break; + } + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, shut_txf_buf, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } + + if (execute_shutdown == true) + schedule_work(&shutdown_work); +} + +/* + * Set the host time in a process context. + */ +static struct work_struct adj_time_work; + +/* + * The last time sample, received from the host. PTP device responds to + * requests by using this data and the current partition-wide time reference + * count. + */ +static struct { + u64 host_time; + u64 ref_time; + spinlock_t lock; +} host_ts; + +static struct timespec64 hv_get_adj_host_time(void) +{ + struct timespec64 ts; + u64 newtime, reftime; + unsigned long flags; + + spin_lock_irqsave(&host_ts.lock, flags); + reftime = hyperv_cs->read(hyperv_cs); + newtime = host_ts.host_time + (reftime - host_ts.ref_time); + ts = ns_to_timespec64((newtime - WLTIMEDELTA) * 100); + spin_unlock_irqrestore(&host_ts.lock, flags); + + return ts; +} + +static void hv_set_host_time(struct work_struct *work) +{ + struct timespec64 ts = hv_get_adj_host_time(); + + do_settimeofday64(&ts); +} + +/* + * Synchronize time with host after reboot, restore, etc. + * + * ICTIMESYNCFLAG_SYNC flag bit indicates reboot, restore events of the VM. + * After reboot the flag ICTIMESYNCFLAG_SYNC is included in the first time + * message after the timesync channel is opened. Since the hv_utils module is + * loaded after hv_vmbus, the first message is usually missed. This bit is + * considered a hard request to discipline the clock. + * + * ICTIMESYNCFLAG_SAMPLE bit indicates a time sample from host. This is + * typically used as a hint to the guest. The guest is under no obligation + * to discipline the clock. + */ +static inline void adj_guesttime(u64 hosttime, u64 reftime, u8 adj_flags) +{ + unsigned long flags; + u64 cur_reftime; + + /* + * Save the adjusted time sample from the host and the snapshot + * of the current system time. + */ + spin_lock_irqsave(&host_ts.lock, flags); + + cur_reftime = hyperv_cs->read(hyperv_cs); + host_ts.host_time = hosttime; + host_ts.ref_time = cur_reftime; + + /* + * TimeSync v4 messages contain reference time (guest's Hyper-V + * clocksource read when the time sample was generated), we can + * improve the precision by adding the delta between now and the + * time of generation. For older protocols we set + * reftime == cur_reftime on call. + */ + host_ts.host_time += (cur_reftime - reftime); + + spin_unlock_irqrestore(&host_ts.lock, flags); + + /* Schedule work to do do_settimeofday64() */ + if (adj_flags & ICTIMESYNCFLAG_SYNC) + schedule_work(&adj_time_work); +} + +/* + * Time Sync Channel message handler. + */ +static void timesync_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + struct icmsg_hdr *icmsghdrp; + struct ictimesync_data *timedatap; + struct ictimesync_ref_data *refdata; + u8 *time_txf_buf = util_timesynch.recv_buffer; + + vmbus_recvpacket(channel, time_txf_buf, + PAGE_SIZE, &recvlen, &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&time_txf_buf[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdrp, time_txf_buf, + fw_versions, FW_VER_COUNT, + ts_versions, TS_VER_COUNT, + NULL, &ts_srv_version)) { + pr_info("TimeSync IC version %d.%d\n", + ts_srv_version >> 16, + ts_srv_version & 0xFFFF); + } + } else { + if (ts_srv_version > TS_VERSION_3) { + refdata = (struct ictimesync_ref_data *) + &time_txf_buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + adj_guesttime(refdata->parenttime, + refdata->vmreferencetime, + refdata->flags); + } else { + timedatap = (struct ictimesync_data *) + &time_txf_buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + adj_guesttime(timedatap->parenttime, + hyperv_cs->read(hyperv_cs), + timedatap->flags); + } + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, time_txf_buf, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } +} + +/* + * Heartbeat functionality. + * Every two seconds, Hyper-V send us a heartbeat request message. + * we respond to this message, and Hyper-V knows we are alive. + */ +static void heartbeat_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + struct icmsg_hdr *icmsghdrp; + struct heartbeat_msg_data *heartbeat_msg; + u8 *hbeat_txf_buf = util_heartbeat.recv_buffer; + + while (1) { + + vmbus_recvpacket(channel, hbeat_txf_buf, + PAGE_SIZE, &recvlen, &requestid); + + if (!recvlen) + break; + + icmsghdrp = (struct icmsg_hdr *)&hbeat_txf_buf[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + if (vmbus_prep_negotiate_resp(icmsghdrp, + hbeat_txf_buf, + fw_versions, FW_VER_COUNT, + hb_versions, HB_VER_COUNT, + NULL, &hb_srv_version)) { + + pr_info("Heartbeat IC version %d.%d\n", + hb_srv_version >> 16, + hb_srv_version & 0xFFFF); + } + } else { + heartbeat_msg = + (struct heartbeat_msg_data *)&hbeat_txf_buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + heartbeat_msg->seq_num += 1; + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, hbeat_txf_buf, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } +} + +static int util_probe(struct hv_device *dev, + const struct hv_vmbus_device_id *dev_id) +{ + struct hv_util_service *srv = + (struct hv_util_service *)dev_id->driver_data; + int ret; + + srv->recv_buffer = kmalloc(PAGE_SIZE * 4, GFP_KERNEL); + if (!srv->recv_buffer) + return -ENOMEM; + srv->channel = dev->channel; + if (srv->util_init) { + ret = srv->util_init(srv); + if (ret) { + ret = -ENODEV; + goto error1; + } + } + + /* + * The set of services managed by the util driver are not performance + * critical and do not need batched reading. Furthermore, some services + * such as KVP can only handle one message from the host at a time. + * Turn off batched reading for all util drivers before we open the + * channel. + */ + set_channel_read_mode(dev->channel, HV_CALL_DIRECT); + + hv_set_drvdata(dev, srv); + + ret = vmbus_open(dev->channel, 4 * PAGE_SIZE, 4 * PAGE_SIZE, NULL, 0, + srv->util_cb, dev->channel); + if (ret) + goto error; + + return 0; + +error: + if (srv->util_deinit) + srv->util_deinit(); +error1: + kfree(srv->recv_buffer); + return ret; +} + +static int util_remove(struct hv_device *dev) +{ + struct hv_util_service *srv = hv_get_drvdata(dev); + + if (srv->util_deinit) + srv->util_deinit(); + vmbus_close(dev->channel); + kfree(srv->recv_buffer); + + return 0; +} + +static const struct hv_vmbus_device_id id_table[] = { + /* Shutdown guid */ + { HV_SHUTDOWN_GUID, + .driver_data = (unsigned long)&util_shutdown + }, + /* Time synch guid */ + { HV_TS_GUID, + .driver_data = (unsigned long)&util_timesynch + }, + /* Heartbeat guid */ + { HV_HEART_BEAT_GUID, + .driver_data = (unsigned long)&util_heartbeat + }, + /* KVP guid */ + { HV_KVP_GUID, + .driver_data = (unsigned long)&util_kvp + }, + /* VSS GUID */ + { HV_VSS_GUID, + .driver_data = (unsigned long)&util_vss + }, + /* File copy GUID */ + { HV_FCOPY_GUID, + .driver_data = (unsigned long)&util_fcopy + }, + { }, +}; + +MODULE_DEVICE_TABLE(vmbus, id_table); + +/* The one and only one */ +static struct hv_driver util_drv = { + .name = "hv_util", + .id_table = id_table, + .probe = util_probe, + .remove = util_remove, + .driver = { + .probe_type = PROBE_PREFER_ASYNCHRONOUS, + }, +}; + +static int hv_ptp_enable(struct ptp_clock_info *info, + struct ptp_clock_request *request, int on) +{ + return -EOPNOTSUPP; +} + +static int hv_ptp_settime(struct ptp_clock_info *p, const struct timespec64 *ts) +{ + return -EOPNOTSUPP; +} + +static int hv_ptp_adjfreq(struct ptp_clock_info *ptp, s32 delta) +{ + return -EOPNOTSUPP; +} +static int hv_ptp_adjtime(struct ptp_clock_info *ptp, s64 delta) +{ + return -EOPNOTSUPP; +} + +static int hv_ptp_gettime(struct ptp_clock_info *info, struct timespec64 *ts) +{ + *ts = hv_get_adj_host_time(); + + return 0; +} + +static struct ptp_clock_info ptp_hyperv_info = { + .name = "hyperv", + .enable = hv_ptp_enable, + .adjtime = hv_ptp_adjtime, + .adjfreq = hv_ptp_adjfreq, + .gettime64 = hv_ptp_gettime, + .settime64 = hv_ptp_settime, + .owner = THIS_MODULE, +}; + +static struct ptp_clock *hv_ptp_clock; + +static int hv_timesync_init(struct hv_util_service *srv) +{ + /* TimeSync requires Hyper-V clocksource. */ + if (!hyperv_cs) + return -ENODEV; + + spin_lock_init(&host_ts.lock); + + INIT_WORK(&adj_time_work, hv_set_host_time); + + /* + * ptp_clock_register() returns NULL when CONFIG_PTP_1588_CLOCK is + * disabled but the driver is still useful without the PTP device + * as it still handles the ICTIMESYNCFLAG_SYNC case. + */ + hv_ptp_clock = ptp_clock_register(&ptp_hyperv_info, NULL); + if (IS_ERR_OR_NULL(hv_ptp_clock)) { + pr_err("cannot register PTP clock: %d\n", + PTR_ERR_OR_ZERO(hv_ptp_clock)); + hv_ptp_clock = NULL; + } + + return 0; +} + +static void hv_timesync_deinit(void) +{ + if (hv_ptp_clock) + ptp_clock_unregister(hv_ptp_clock); + cancel_work_sync(&adj_time_work); +} + +static int __init init_hyperv_utils(void) +{ + pr_info("Registering HyperV Utility Driver\n"); + + return vmbus_driver_register(&util_drv); +} + +static void exit_hyperv_utils(void) +{ + pr_info("De-Registered HyperV Utility Driver\n"); + + vmbus_driver_unregister(&util_drv); +} + +module_init(init_hyperv_utils); +module_exit(exit_hyperv_utils); + +MODULE_DESCRIPTION("Hyper-V Utilities"); +MODULE_LICENSE("GPL"); diff --git a/drivers/hv/hv_utils_transport.c b/drivers/hv/hv_utils_transport.c new file mode 100644 index 000000000..832777527 --- /dev/null +++ b/drivers/hv/hv_utils_transport.c @@ -0,0 +1,360 @@ +/* + * Kernel/userspace transport abstraction for Hyper-V util driver. + * + * Copyright (C) 2015, Vitaly Kuznetsov <vkuznets@redhat.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + */ + +#include <linux/slab.h> +#include <linux/fs.h> +#include <linux/poll.h> + +#include "hyperv_vmbus.h" +#include "hv_utils_transport.h" + +static DEFINE_SPINLOCK(hvt_list_lock); +static struct list_head hvt_list = LIST_HEAD_INIT(hvt_list); + +static void hvt_reset(struct hvutil_transport *hvt) +{ + kfree(hvt->outmsg); + hvt->outmsg = NULL; + hvt->outmsg_len = 0; + if (hvt->on_reset) + hvt->on_reset(); +} + +static ssize_t hvt_op_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + struct hvutil_transport *hvt; + int ret; + + hvt = container_of(file->f_op, struct hvutil_transport, fops); + + if (wait_event_interruptible(hvt->outmsg_q, hvt->outmsg_len > 0 || + hvt->mode != HVUTIL_TRANSPORT_CHARDEV)) + return -EINTR; + + mutex_lock(&hvt->lock); + + if (hvt->mode == HVUTIL_TRANSPORT_DESTROY) { + ret = -EBADF; + goto out_unlock; + } + + if (!hvt->outmsg) { + ret = -EAGAIN; + goto out_unlock; + } + + if (count < hvt->outmsg_len) { + ret = -EINVAL; + goto out_unlock; + } + + if (!copy_to_user(buf, hvt->outmsg, hvt->outmsg_len)) + ret = hvt->outmsg_len; + else + ret = -EFAULT; + + kfree(hvt->outmsg); + hvt->outmsg = NULL; + hvt->outmsg_len = 0; + + if (hvt->on_read) + hvt->on_read(); + hvt->on_read = NULL; + +out_unlock: + mutex_unlock(&hvt->lock); + return ret; +} + +static ssize_t hvt_op_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + struct hvutil_transport *hvt; + u8 *inmsg; + int ret; + + hvt = container_of(file->f_op, struct hvutil_transport, fops); + + inmsg = memdup_user(buf, count); + if (IS_ERR(inmsg)) + return PTR_ERR(inmsg); + + if (hvt->mode == HVUTIL_TRANSPORT_DESTROY) + ret = -EBADF; + else + ret = hvt->on_msg(inmsg, count); + + kfree(inmsg); + + return ret ? ret : count; +} + +static __poll_t hvt_op_poll(struct file *file, poll_table *wait) +{ + struct hvutil_transport *hvt; + + hvt = container_of(file->f_op, struct hvutil_transport, fops); + + poll_wait(file, &hvt->outmsg_q, wait); + + if (hvt->mode == HVUTIL_TRANSPORT_DESTROY) + return EPOLLERR | EPOLLHUP; + + if (hvt->outmsg_len > 0) + return EPOLLIN | EPOLLRDNORM; + + return 0; +} + +static int hvt_op_open(struct inode *inode, struct file *file) +{ + struct hvutil_transport *hvt; + int ret = 0; + bool issue_reset = false; + + hvt = container_of(file->f_op, struct hvutil_transport, fops); + + mutex_lock(&hvt->lock); + + if (hvt->mode == HVUTIL_TRANSPORT_DESTROY) { + ret = -EBADF; + } else if (hvt->mode == HVUTIL_TRANSPORT_INIT) { + /* + * Switching to CHARDEV mode. We switch bach to INIT when + * device gets released. + */ + hvt->mode = HVUTIL_TRANSPORT_CHARDEV; + } + else if (hvt->mode == HVUTIL_TRANSPORT_NETLINK) { + /* + * We're switching from netlink communication to using char + * device. Issue the reset first. + */ + issue_reset = true; + hvt->mode = HVUTIL_TRANSPORT_CHARDEV; + } else { + ret = -EBUSY; + } + + if (issue_reset) + hvt_reset(hvt); + + mutex_unlock(&hvt->lock); + + return ret; +} + +static void hvt_transport_free(struct hvutil_transport *hvt) +{ + misc_deregister(&hvt->mdev); + kfree(hvt->outmsg); + kfree(hvt); +} + +static int hvt_op_release(struct inode *inode, struct file *file) +{ + struct hvutil_transport *hvt; + int mode_old; + + hvt = container_of(file->f_op, struct hvutil_transport, fops); + + mutex_lock(&hvt->lock); + mode_old = hvt->mode; + if (hvt->mode != HVUTIL_TRANSPORT_DESTROY) + hvt->mode = HVUTIL_TRANSPORT_INIT; + /* + * Cleanup message buffers to avoid spurious messages when the daemon + * connects back. + */ + hvt_reset(hvt); + + if (mode_old == HVUTIL_TRANSPORT_DESTROY) + complete(&hvt->release); + + mutex_unlock(&hvt->lock); + + return 0; +} + +static void hvt_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp) +{ + struct hvutil_transport *hvt, *hvt_found = NULL; + + spin_lock(&hvt_list_lock); + list_for_each_entry(hvt, &hvt_list, list) { + if (hvt->cn_id.idx == msg->id.idx && + hvt->cn_id.val == msg->id.val) { + hvt_found = hvt; + break; + } + } + spin_unlock(&hvt_list_lock); + if (!hvt_found) { + pr_warn("hvt_cn_callback: spurious message received!\n"); + return; + } + + /* + * Switching to NETLINK mode. Switching to CHARDEV happens when someone + * opens the device. + */ + mutex_lock(&hvt->lock); + if (hvt->mode == HVUTIL_TRANSPORT_INIT) + hvt->mode = HVUTIL_TRANSPORT_NETLINK; + + if (hvt->mode == HVUTIL_TRANSPORT_NETLINK) + hvt_found->on_msg(msg->data, msg->len); + else + pr_warn("hvt_cn_callback: unexpected netlink message!\n"); + mutex_unlock(&hvt->lock); +} + +int hvutil_transport_send(struct hvutil_transport *hvt, void *msg, int len, + void (*on_read_cb)(void)) +{ + struct cn_msg *cn_msg; + int ret = 0; + + if (hvt->mode == HVUTIL_TRANSPORT_INIT || + hvt->mode == HVUTIL_TRANSPORT_DESTROY) { + return -EINVAL; + } else if (hvt->mode == HVUTIL_TRANSPORT_NETLINK) { + cn_msg = kzalloc(sizeof(*cn_msg) + len, GFP_ATOMIC); + if (!cn_msg) + return -ENOMEM; + cn_msg->id.idx = hvt->cn_id.idx; + cn_msg->id.val = hvt->cn_id.val; + cn_msg->len = len; + memcpy(cn_msg->data, msg, len); + ret = cn_netlink_send(cn_msg, 0, 0, GFP_ATOMIC); + kfree(cn_msg); + /* + * We don't know when netlink messages are delivered but unlike + * in CHARDEV mode we're not blocked and we can send next + * messages right away. + */ + if (on_read_cb) + on_read_cb(); + return ret; + } + /* HVUTIL_TRANSPORT_CHARDEV */ + mutex_lock(&hvt->lock); + if (hvt->mode != HVUTIL_TRANSPORT_CHARDEV) { + ret = -EINVAL; + goto out_unlock; + } + + if (hvt->outmsg) { + /* Previous message wasn't received */ + ret = -EFAULT; + goto out_unlock; + } + hvt->outmsg = kzalloc(len, GFP_KERNEL); + if (hvt->outmsg) { + memcpy(hvt->outmsg, msg, len); + hvt->outmsg_len = len; + hvt->on_read = on_read_cb; + wake_up_interruptible(&hvt->outmsg_q); + } else + ret = -ENOMEM; +out_unlock: + mutex_unlock(&hvt->lock); + return ret; +} + +struct hvutil_transport *hvutil_transport_init(const char *name, + u32 cn_idx, u32 cn_val, + int (*on_msg)(void *, int), + void (*on_reset)(void)) +{ + struct hvutil_transport *hvt; + + hvt = kzalloc(sizeof(*hvt), GFP_KERNEL); + if (!hvt) + return NULL; + + hvt->cn_id.idx = cn_idx; + hvt->cn_id.val = cn_val; + + hvt->mdev.minor = MISC_DYNAMIC_MINOR; + hvt->mdev.name = name; + + hvt->fops.owner = THIS_MODULE; + hvt->fops.read = hvt_op_read; + hvt->fops.write = hvt_op_write; + hvt->fops.poll = hvt_op_poll; + hvt->fops.open = hvt_op_open; + hvt->fops.release = hvt_op_release; + + hvt->mdev.fops = &hvt->fops; + + init_waitqueue_head(&hvt->outmsg_q); + mutex_init(&hvt->lock); + init_completion(&hvt->release); + + spin_lock(&hvt_list_lock); + list_add(&hvt->list, &hvt_list); + spin_unlock(&hvt_list_lock); + + hvt->on_msg = on_msg; + hvt->on_reset = on_reset; + + if (misc_register(&hvt->mdev)) + goto err_free_hvt; + + /* Use cn_id.idx/cn_id.val to determine if we need to setup netlink */ + if (hvt->cn_id.idx > 0 && hvt->cn_id.val > 0 && + cn_add_callback(&hvt->cn_id, name, hvt_cn_callback)) + goto err_free_hvt; + + return hvt; + +err_free_hvt: + spin_lock(&hvt_list_lock); + list_del(&hvt->list); + spin_unlock(&hvt_list_lock); + kfree(hvt); + return NULL; +} + +void hvutil_transport_destroy(struct hvutil_transport *hvt) +{ + int mode_old; + + mutex_lock(&hvt->lock); + mode_old = hvt->mode; + hvt->mode = HVUTIL_TRANSPORT_DESTROY; + wake_up_interruptible(&hvt->outmsg_q); + mutex_unlock(&hvt->lock); + + /* + * In case we were in 'chardev' mode we still have an open fd so we + * have to defer freeing the device. Netlink interface can be freed + * now. + */ + spin_lock(&hvt_list_lock); + list_del(&hvt->list); + spin_unlock(&hvt_list_lock); + if (hvt->cn_id.idx > 0 && hvt->cn_id.val > 0) + cn_del_callback(&hvt->cn_id); + + if (mode_old == HVUTIL_TRANSPORT_CHARDEV) + wait_for_completion(&hvt->release); + + hvt_transport_free(hvt); +} diff --git a/drivers/hv/hv_utils_transport.h b/drivers/hv/hv_utils_transport.h new file mode 100644 index 000000000..79afb626e --- /dev/null +++ b/drivers/hv/hv_utils_transport.h @@ -0,0 +1,55 @@ +/* + * Kernel/userspace transport abstraction for Hyper-V util driver. + * + * Copyright (C) 2015, Vitaly Kuznetsov <vkuznets@redhat.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * 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, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + */ + +#ifndef _HV_UTILS_TRANSPORT_H +#define _HV_UTILS_TRANSPORT_H + +#include <linux/connector.h> +#include <linux/miscdevice.h> + +enum hvutil_transport_mode { + HVUTIL_TRANSPORT_INIT = 0, + HVUTIL_TRANSPORT_NETLINK, + HVUTIL_TRANSPORT_CHARDEV, + HVUTIL_TRANSPORT_DESTROY, +}; + +struct hvutil_transport { + int mode; /* hvutil_transport_mode */ + struct file_operations fops; /* file operations */ + struct miscdevice mdev; /* misc device */ + struct cb_id cn_id; /* CN_*_IDX/CN_*_VAL */ + struct list_head list; /* hvt_list */ + int (*on_msg)(void *, int); /* callback on new user message */ + void (*on_reset)(void); /* callback when userspace drops */ + void (*on_read)(void); /* callback on message read */ + u8 *outmsg; /* message to the userspace */ + int outmsg_len; /* its length */ + wait_queue_head_t outmsg_q; /* poll/read wait queue */ + struct mutex lock; /* protects struct members */ + struct completion release; /* synchronize with fd release */ +}; + +struct hvutil_transport *hvutil_transport_init(const char *name, + u32 cn_idx, u32 cn_val, + int (*on_msg)(void *, int), + void (*on_reset)(void)); +int hvutil_transport_send(struct hvutil_transport *hvt, void *msg, int len, + void (*on_read_cb)(void)); +void hvutil_transport_destroy(struct hvutil_transport *hvt); + +#endif /* _HV_UTILS_TRANSPORT_H */ diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h new file mode 100644 index 000000000..c4ad51889 --- /dev/null +++ b/drivers/hv/hyperv_vmbus.h @@ -0,0 +1,446 @@ +/* + * + * Copyright (c) 2011, 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> + * + */ + +#ifndef _HYPERV_VMBUS_H +#define _HYPERV_VMBUS_H + +#include <linux/list.h> +#include <linux/bitops.h> +#include <asm/sync_bitops.h> +#include <asm/hyperv-tlfs.h> +#include <linux/atomic.h> +#include <linux/hyperv.h> +#include <linux/interrupt.h> + +#include "hv_trace.h" + +/* + * Timeout for services such as KVP and fcopy. + */ +#define HV_UTIL_TIMEOUT 30 + +/* + * Timeout for guest-host handshake for services. + */ +#define HV_UTIL_NEGO_TIMEOUT 55 + +/* Define synthetic interrupt controller flag constants. */ +#define HV_EVENT_FLAGS_COUNT (256 * 8) +#define HV_EVENT_FLAGS_LONG_COUNT (256 / sizeof(unsigned long)) + +/* + * Timer configuration register. + */ +union hv_timer_config { + u64 as_uint64; + struct { + u64 enable:1; + u64 periodic:1; + u64 lazy:1; + u64 auto_enable:1; + u64 apic_vector:8; + u64 direct_mode:1; + u64 reserved_z0:3; + u64 sintx:4; + u64 reserved_z1:44; + }; +}; + + +/* Define the synthetic interrupt controller event flags format. */ +union hv_synic_event_flags { + unsigned long flags[HV_EVENT_FLAGS_LONG_COUNT]; +}; + +/* Define SynIC control register. */ +union hv_synic_scontrol { + u64 as_uint64; + struct { + u64 enable:1; + u64 reserved:63; + }; +}; + +/* Define synthetic interrupt source. */ +union hv_synic_sint { + u64 as_uint64; + struct { + u64 vector:8; + u64 reserved1:8; + u64 masked:1; + u64 auto_eoi:1; + u64 reserved2:46; + }; +}; + +/* Define the format of the SIMP register */ +union hv_synic_simp { + u64 as_uint64; + struct { + u64 simp_enabled:1; + u64 preserved:11; + u64 base_simp_gpa:52; + }; +}; + +/* Define the format of the SIEFP register */ +union hv_synic_siefp { + u64 as_uint64; + struct { + u64 siefp_enabled:1; + u64 preserved:11; + u64 base_siefp_gpa:52; + }; +}; + +/* Definitions for the monitored notification facility */ +union hv_monitor_trigger_group { + u64 as_uint64; + struct { + u32 pending; + u32 armed; + }; +}; + +struct hv_monitor_parameter { + union hv_connection_id connectionid; + u16 flagnumber; + u16 rsvdz; +}; + +union hv_monitor_trigger_state { + u32 asu32; + + struct { + u32 group_enable:4; + u32 rsvdz:28; + }; +}; + +/* struct hv_monitor_page Layout */ +/* ------------------------------------------------------ */ +/* | 0 | TriggerState (4 bytes) | Rsvd1 (4 bytes) | */ +/* | 8 | TriggerGroup[0] | */ +/* | 10 | TriggerGroup[1] | */ +/* | 18 | TriggerGroup[2] | */ +/* | 20 | TriggerGroup[3] | */ +/* | 28 | Rsvd2[0] | */ +/* | 30 | Rsvd2[1] | */ +/* | 38 | Rsvd2[2] | */ +/* | 40 | NextCheckTime[0][0] | NextCheckTime[0][1] | */ +/* | ... | */ +/* | 240 | Latency[0][0..3] | */ +/* | 340 | Rsvz3[0] | */ +/* | 440 | Parameter[0][0] | */ +/* | 448 | Parameter[0][1] | */ +/* | ... | */ +/* | 840 | Rsvd4[0] | */ +/* ------------------------------------------------------ */ +struct hv_monitor_page { + union hv_monitor_trigger_state trigger_state; + u32 rsvdz1; + + union hv_monitor_trigger_group trigger_group[4]; + u64 rsvdz2[3]; + + s32 next_checktime[4][32]; + + u16 latency[4][32]; + u64 rsvdz3[32]; + + struct hv_monitor_parameter parameter[4][32]; + + u8 rsvdz4[1984]; +}; + +#define HV_HYPERCALL_PARAM_ALIGN sizeof(u64) + +/* Definition of the hv_post_message hypercall input structure. */ +struct hv_input_post_message { + union hv_connection_id connectionid; + u32 reserved; + u32 message_type; + u32 payload_size; + u64 payload[HV_MESSAGE_PAYLOAD_QWORD_COUNT]; +}; + + +enum { + VMBUS_MESSAGE_CONNECTION_ID = 1, + VMBUS_MESSAGE_CONNECTION_ID_4 = 4, + VMBUS_MESSAGE_PORT_ID = 1, + VMBUS_EVENT_CONNECTION_ID = 2, + VMBUS_EVENT_PORT_ID = 2, + VMBUS_MONITOR_CONNECTION_ID = 3, + VMBUS_MONITOR_PORT_ID = 3, + VMBUS_MESSAGE_SINT = 2, +}; + +/* + * Per cpu state for channel handling + */ +struct hv_per_cpu_context { + void *synic_message_page; + void *synic_event_page; + /* + * buffer to post messages to the host. + */ + void *post_msg_page; + + /* + * Starting with win8, we can take channel interrupts on any CPU; + * we will manage the tasklet that handles events messages on a per CPU + * basis. + */ + struct tasklet_struct msg_dpc; + + /* + * To optimize the mapping of relid to channel, maintain + * per-cpu list of the channels based on their CPU affinity. + */ + struct list_head chan_list; + struct clock_event_device *clk_evt; +}; + +struct hv_context { + /* We only support running on top of Hyper-V + * So at this point this really can only contain the Hyper-V ID + */ + u64 guestid; + + void *tsc_page; + + bool synic_initialized; + + struct hv_per_cpu_context __percpu *cpu_context; + + /* + * To manage allocations in a NUMA node. + * Array indexed by numa node ID. + */ + struct cpumask *hv_numa_map; +}; + +extern struct hv_context hv_context; + +/* Hv Interface */ + +extern int hv_init(void); + +extern int hv_post_message(union hv_connection_id connection_id, + enum hv_message_type message_type, + void *payload, size_t payload_size); + +extern int hv_synic_alloc(void); + +extern void hv_synic_free(void); + +extern int hv_synic_init(unsigned int cpu); + +extern int hv_synic_cleanup(unsigned int cpu); + +extern void hv_synic_clockevents_cleanup(void); + +/* Interface */ + + +int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info, + struct page *pages, u32 pagecnt); + +void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info); + +int hv_ringbuffer_write(struct vmbus_channel *channel, + const struct kvec *kv_list, u32 kv_count); + +int hv_ringbuffer_read(struct vmbus_channel *channel, + void *buffer, u32 buflen, u32 *buffer_actual_len, + u64 *requestid, bool raw); + +/* + * Maximum channels is determined by the size of the interrupt page + * which is PAGE_SIZE. 1/2 of PAGE_SIZE is for send endpoint interrupt + * and the other is receive endpoint interrupt + */ +#define MAX_NUM_CHANNELS ((PAGE_SIZE >> 1) << 3) /* 16348 channels */ + +/* The value here must be in multiple of 32 */ +/* TODO: Need to make this configurable */ +#define MAX_NUM_CHANNELS_SUPPORTED 256 + + +enum vmbus_connect_state { + DISCONNECTED, + CONNECTING, + CONNECTED, + DISCONNECTING +}; + +#define MAX_SIZE_CHANNEL_MESSAGE HV_MESSAGE_PAYLOAD_BYTE_COUNT + +struct vmbus_connection { + /* + * CPU on which the initial host contact was made. + */ + int connect_cpu; + + u32 msg_conn_id; + + atomic_t offer_in_progress; + + enum vmbus_connect_state conn_state; + + atomic_t next_gpadl_handle; + + struct completion unload_event; + /* + * Represents channel interrupts. Each bit position represents a + * channel. When a channel sends an interrupt via VMBUS, it finds its + * bit in the sendInterruptPage, set it and calls Hv to generate a port + * event. The other end receives the port event and parse the + * recvInterruptPage to see which bit is set + */ + void *int_page; + void *send_int_page; + void *recv_int_page; + + /* + * 2 pages - 1st page for parent->child notification and 2nd + * is child->parent notification + */ + struct hv_monitor_page *monitor_pages[2]; + struct list_head chn_msg_list; + spinlock_t channelmsg_lock; + + /* List of channels */ + struct list_head chn_list; + struct mutex channel_mutex; + + /* + * An offer message is handled first on the work_queue, and then + * is further handled on handle_primary_chan_wq or + * handle_sub_chan_wq. + */ + struct workqueue_struct *work_queue; + struct workqueue_struct *handle_primary_chan_wq; + struct workqueue_struct *handle_sub_chan_wq; +}; + + +struct vmbus_msginfo { + /* Bookkeeping stuff */ + struct list_head msglist_entry; + + /* The message itself */ + unsigned char msg[0]; +}; + + +extern struct vmbus_connection vmbus_connection; + +static inline void vmbus_send_interrupt(u32 relid) +{ + sync_set_bit(relid, vmbus_connection.send_int_page); +} + +enum vmbus_message_handler_type { + /* The related handler can sleep. */ + VMHT_BLOCKING = 0, + + /* The related handler must NOT sleep. */ + VMHT_NON_BLOCKING = 1, +}; + +struct vmbus_channel_message_table_entry { + enum vmbus_channel_message_type message_type; + enum vmbus_message_handler_type handler_type; + void (*message_handler)(struct vmbus_channel_message_header *msg); +}; + +extern const struct vmbus_channel_message_table_entry + channel_message_table[CHANNELMSG_COUNT]; + + +/* General vmbus interface */ + +struct hv_device *vmbus_device_create(const uuid_le *type, + const uuid_le *instance, + struct vmbus_channel *channel); + +int vmbus_device_register(struct hv_device *child_device_obj); +void vmbus_device_unregister(struct hv_device *device_obj); +int vmbus_add_channel_kobj(struct hv_device *device_obj, + struct vmbus_channel *channel); + +void vmbus_remove_channel_attr_group(struct vmbus_channel *channel); + +struct vmbus_channel *relid2channel(u32 relid); + +void vmbus_free_channels(void); + +/* Connection interface */ + +int vmbus_connect(void); +void vmbus_disconnect(void); + +int vmbus_post_msg(void *buffer, size_t buflen, bool can_sleep); + +void vmbus_on_event(unsigned long data); +void vmbus_on_msg_dpc(unsigned long data); + +int hv_kvp_init(struct hv_util_service *srv); +void hv_kvp_deinit(void); +void hv_kvp_onchannelcallback(void *context); + +int hv_vss_init(struct hv_util_service *srv); +void hv_vss_deinit(void); +void hv_vss_onchannelcallback(void *context); + +int hv_fcopy_init(struct hv_util_service *srv); +void hv_fcopy_deinit(void); +void hv_fcopy_onchannelcallback(void *context); +void vmbus_initiate_unload(bool crash); + +static inline void hv_poll_channel(struct vmbus_channel *channel, + void (*cb)(void *)) +{ + if (!channel) + return; + + if (in_interrupt() && (channel->target_cpu == smp_processor_id())) { + cb(channel); + return; + } + smp_call_function_single(channel->target_cpu, cb, channel, true); +} + +enum hvutil_device_state { + HVUTIL_DEVICE_INIT = 0, /* driver is loaded, waiting for userspace */ + HVUTIL_READY, /* userspace is registered */ + HVUTIL_HOSTMSG_RECEIVED, /* message from the host was received */ + HVUTIL_USERSPACE_REQ, /* request to userspace was sent */ + HVUTIL_USERSPACE_RECV, /* reply from userspace was received */ + HVUTIL_DEVICE_DYING, /* driver unload is in progress */ +}; + +#endif /* _HYPERV_VMBUS_H */ diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c new file mode 100644 index 000000000..d97b30af9 --- /dev/null +++ b/drivers/hv/ring_buffer.c @@ -0,0 +1,543 @@ +/* + * + * 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/kernel.h> +#include <linux/mm.h> +#include <linux/hyperv.h> +#include <linux/uio.h> +#include <linux/vmalloc.h> +#include <linux/slab.h> +#include <linux/prefetch.h> + +#include "hyperv_vmbus.h" + +#define VMBUS_PKT_TRAILER 8 + +/* + * When we write to the ring buffer, check if the host needs to + * be signaled. Here is the details of this protocol: + * + * 1. The host guarantees that while it is draining the + * ring buffer, it will set the interrupt_mask to + * indicate it does not need to be interrupted when + * new data is placed. + * + * 2. The host guarantees that it will completely drain + * the ring buffer before exiting the read loop. Further, + * once the ring buffer is empty, it will clear the + * interrupt_mask and re-check to see if new data has + * arrived. + * + * KYS: Oct. 30, 2016: + * It looks like Windows hosts have logic to deal with DOS attacks that + * can be triggered if it receives interrupts when it is not expecting + * the interrupt. The host expects interrupts only when the ring + * transitions from empty to non-empty (or full to non full on the guest + * to host ring). + * So, base the signaling decision solely on the ring state until the + * host logic is fixed. + */ + +static void hv_signal_on_write(u32 old_write, struct vmbus_channel *channel) +{ + struct hv_ring_buffer_info *rbi = &channel->outbound; + + virt_mb(); + if (READ_ONCE(rbi->ring_buffer->interrupt_mask)) + return; + + /* check interrupt_mask before read_index */ + virt_rmb(); + /* + * This is the only case we need to signal when the + * ring transitions from being empty to non-empty. + */ + if (old_write == READ_ONCE(rbi->ring_buffer->read_index)) + vmbus_setevent(channel); +} + +/* Get the next write location for the specified ring buffer. */ +static inline u32 +hv_get_next_write_location(struct hv_ring_buffer_info *ring_info) +{ + u32 next = ring_info->ring_buffer->write_index; + + return next; +} + +/* Set the next write location for the specified ring buffer. */ +static inline void +hv_set_next_write_location(struct hv_ring_buffer_info *ring_info, + u32 next_write_location) +{ + ring_info->ring_buffer->write_index = next_write_location; +} + +/* Set the next read location for the specified ring buffer. */ +static inline void +hv_set_next_read_location(struct hv_ring_buffer_info *ring_info, + u32 next_read_location) +{ + ring_info->ring_buffer->read_index = next_read_location; + ring_info->priv_read_index = next_read_location; +} + +/* Get the size of the ring buffer. */ +static inline u32 +hv_get_ring_buffersize(const struct hv_ring_buffer_info *ring_info) +{ + return ring_info->ring_datasize; +} + +/* Get the read and write indices as u64 of the specified ring buffer. */ +static inline u64 +hv_get_ring_bufferindices(struct hv_ring_buffer_info *ring_info) +{ + return (u64)ring_info->ring_buffer->write_index << 32; +} + +/* + * Helper routine to copy from source to ring buffer. + * Assume there is enough room. Handles wrap-around in dest case only!! + */ +static u32 hv_copyto_ringbuffer( + struct hv_ring_buffer_info *ring_info, + u32 start_write_offset, + const void *src, + u32 srclen) +{ + void *ring_buffer = hv_get_ring_buffer(ring_info); + u32 ring_buffer_size = hv_get_ring_buffersize(ring_info); + + memcpy(ring_buffer + start_write_offset, src, srclen); + + start_write_offset += srclen; + if (start_write_offset >= ring_buffer_size) + start_write_offset -= ring_buffer_size; + + return start_write_offset; +} + +/* + * + * hv_get_ringbuffer_availbytes() + * + * Get number of bytes available to read and to write to + * for the specified ring buffer + */ +static void +hv_get_ringbuffer_availbytes(const struct hv_ring_buffer_info *rbi, + u32 *read, u32 *write) +{ + u32 read_loc, write_loc, dsize; + + /* Capture the read/write indices before they changed */ + read_loc = READ_ONCE(rbi->ring_buffer->read_index); + write_loc = READ_ONCE(rbi->ring_buffer->write_index); + dsize = rbi->ring_datasize; + + *write = write_loc >= read_loc ? dsize - (write_loc - read_loc) : + read_loc - write_loc; + *read = dsize - *write; +} + +/* Get various debug metrics for the specified ring buffer. */ +int hv_ringbuffer_get_debuginfo(const struct hv_ring_buffer_info *ring_info, + struct hv_ring_buffer_debug_info *debug_info) +{ + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + + if (!ring_info->ring_buffer) + return -EINVAL; + + hv_get_ringbuffer_availbytes(ring_info, + &bytes_avail_toread, + &bytes_avail_towrite); + debug_info->bytes_avail_toread = bytes_avail_toread; + debug_info->bytes_avail_towrite = bytes_avail_towrite; + debug_info->current_read_index = ring_info->ring_buffer->read_index; + debug_info->current_write_index = ring_info->ring_buffer->write_index; + debug_info->current_interrupt_mask + = ring_info->ring_buffer->interrupt_mask; + return 0; +} +EXPORT_SYMBOL_GPL(hv_ringbuffer_get_debuginfo); + +/* Initialize the ring buffer. */ +int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info, + struct page *pages, u32 page_cnt) +{ + int i; + struct page **pages_wraparound; + + BUILD_BUG_ON((sizeof(struct hv_ring_buffer) != PAGE_SIZE)); + + memset(ring_info, 0, sizeof(struct hv_ring_buffer_info)); + + /* + * First page holds struct hv_ring_buffer, do wraparound mapping for + * the rest. + */ + pages_wraparound = kcalloc(page_cnt * 2 - 1, sizeof(struct page *), + GFP_KERNEL); + if (!pages_wraparound) + return -ENOMEM; + + pages_wraparound[0] = pages; + for (i = 0; i < 2 * (page_cnt - 1); i++) + pages_wraparound[i + 1] = &pages[i % (page_cnt - 1) + 1]; + + ring_info->ring_buffer = (struct hv_ring_buffer *) + vmap(pages_wraparound, page_cnt * 2 - 1, VM_MAP, PAGE_KERNEL); + + kfree(pages_wraparound); + + + if (!ring_info->ring_buffer) + return -ENOMEM; + + ring_info->ring_buffer->read_index = + ring_info->ring_buffer->write_index = 0; + + /* Set the feature bit for enabling flow control. */ + ring_info->ring_buffer->feature_bits.value = 1; + + ring_info->ring_size = page_cnt << PAGE_SHIFT; + ring_info->ring_size_div10_reciprocal = + reciprocal_value(ring_info->ring_size / 10); + ring_info->ring_datasize = ring_info->ring_size - + sizeof(struct hv_ring_buffer); + + spin_lock_init(&ring_info->ring_lock); + + return 0; +} + +/* Cleanup the ring buffer. */ +void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info) +{ + vunmap(ring_info->ring_buffer); +} + +/* Write to the ring buffer. */ +int hv_ringbuffer_write(struct vmbus_channel *channel, + const struct kvec *kv_list, u32 kv_count) +{ + int i; + u32 bytes_avail_towrite; + u32 totalbytes_towrite = sizeof(u64); + u32 next_write_location; + u32 old_write; + u64 prev_indices; + unsigned long flags; + struct hv_ring_buffer_info *outring_info = &channel->outbound; + + if (channel->rescind) + return -ENODEV; + + for (i = 0; i < kv_count; i++) + totalbytes_towrite += kv_list[i].iov_len; + + spin_lock_irqsave(&outring_info->ring_lock, flags); + + bytes_avail_towrite = hv_get_bytes_to_write(outring_info); + + /* + * If there is only room for the packet, assume it is full. + * Otherwise, the next time around, we think the ring buffer + * is empty since the read index == write index. + */ + if (bytes_avail_towrite <= totalbytes_towrite) { + spin_unlock_irqrestore(&outring_info->ring_lock, flags); + return -EAGAIN; + } + + /* Write to the ring buffer */ + next_write_location = hv_get_next_write_location(outring_info); + + old_write = next_write_location; + + for (i = 0; i < kv_count; i++) { + next_write_location = hv_copyto_ringbuffer(outring_info, + next_write_location, + kv_list[i].iov_base, + kv_list[i].iov_len); + } + + /* Set previous packet start */ + prev_indices = hv_get_ring_bufferindices(outring_info); + + next_write_location = hv_copyto_ringbuffer(outring_info, + next_write_location, + &prev_indices, + sizeof(u64)); + + /* Issue a full memory barrier before updating the write index */ + virt_mb(); + + /* Now, update the write location */ + hv_set_next_write_location(outring_info, next_write_location); + + + spin_unlock_irqrestore(&outring_info->ring_lock, flags); + + hv_signal_on_write(old_write, channel); + + if (channel->rescind) + return -ENODEV; + + return 0; +} + +int hv_ringbuffer_read(struct vmbus_channel *channel, + void *buffer, u32 buflen, u32 *buffer_actual_len, + u64 *requestid, bool raw) +{ + struct vmpacket_descriptor *desc; + u32 packetlen, offset; + + if (unlikely(buflen == 0)) + return -EINVAL; + + *buffer_actual_len = 0; + *requestid = 0; + + /* Make sure there is something to read */ + desc = hv_pkt_iter_first(channel); + if (desc == NULL) { + /* + * No error is set when there is even no header, drivers are + * supposed to analyze buffer_actual_len. + */ + return 0; + } + + offset = raw ? 0 : (desc->offset8 << 3); + packetlen = (desc->len8 << 3) - offset; + *buffer_actual_len = packetlen; + *requestid = desc->trans_id; + + if (unlikely(packetlen > buflen)) + return -ENOBUFS; + + /* since ring is double mapped, only one copy is necessary */ + memcpy(buffer, (const char *)desc + offset, packetlen); + + /* Advance ring index to next packet descriptor */ + __hv_pkt_iter_next(channel, desc); + + /* Notify host of update */ + hv_pkt_iter_close(channel); + + return 0; +} + +/* + * Determine number of bytes available in ring buffer after + * the current iterator (priv_read_index) location. + * + * This is similar to hv_get_bytes_to_read but with private + * read index instead. + */ +static u32 hv_pkt_iter_avail(const struct hv_ring_buffer_info *rbi) +{ + u32 priv_read_loc = rbi->priv_read_index; + u32 write_loc; + + /* + * The Hyper-V host writes the packet data, then uses + * store_release() to update the write_index. Use load_acquire() + * here to prevent loads of the packet data from being re-ordered + * before the read of the write_index and potentially getting + * stale data. + */ + write_loc = virt_load_acquire(&rbi->ring_buffer->write_index); + + if (write_loc >= priv_read_loc) + return write_loc - priv_read_loc; + else + return (rbi->ring_datasize - priv_read_loc) + write_loc; +} + +/* + * Get first vmbus packet from ring buffer after read_index + * + * If ring buffer is empty, returns NULL and no other action needed. + */ +struct vmpacket_descriptor *hv_pkt_iter_first(struct vmbus_channel *channel) +{ + struct hv_ring_buffer_info *rbi = &channel->inbound; + struct vmpacket_descriptor *desc; + + if (hv_pkt_iter_avail(rbi) < sizeof(struct vmpacket_descriptor)) + return NULL; + + desc = hv_get_ring_buffer(rbi) + rbi->priv_read_index; + if (desc) + prefetch((char *)desc + (desc->len8 << 3)); + + return desc; +} +EXPORT_SYMBOL_GPL(hv_pkt_iter_first); + +/* + * Get next vmbus packet from ring buffer. + * + * Advances the current location (priv_read_index) and checks for more + * data. If the end of the ring buffer is reached, then return NULL. + */ +struct vmpacket_descriptor * +__hv_pkt_iter_next(struct vmbus_channel *channel, + const struct vmpacket_descriptor *desc) +{ + struct hv_ring_buffer_info *rbi = &channel->inbound; + u32 packetlen = desc->len8 << 3; + u32 dsize = rbi->ring_datasize; + + /* bump offset to next potential packet */ + rbi->priv_read_index += packetlen + VMBUS_PKT_TRAILER; + if (rbi->priv_read_index >= dsize) + rbi->priv_read_index -= dsize; + + /* more data? */ + return hv_pkt_iter_first(channel); +} +EXPORT_SYMBOL_GPL(__hv_pkt_iter_next); + +/* How many bytes were read in this iterator cycle */ +static u32 hv_pkt_iter_bytes_read(const struct hv_ring_buffer_info *rbi, + u32 start_read_index) +{ + if (rbi->priv_read_index >= start_read_index) + return rbi->priv_read_index - start_read_index; + else + return rbi->ring_datasize - start_read_index + + rbi->priv_read_index; +} + +/* + * Update host ring buffer after iterating over packets. If the host has + * stopped queuing new entries because it found the ring buffer full, and + * sufficient space is being freed up, signal the host. But be careful to + * only signal the host when necessary, both for performance reasons and + * because Hyper-V protects itself by throttling guests that signal + * inappropriately. + * + * Determining when to signal is tricky. There are three key data inputs + * that must be handled in this order to avoid race conditions: + * + * 1. Update the read_index + * 2. Read the pending_send_sz + * 3. Read the current write_index + * + * The interrupt_mask is not used to determine when to signal. The + * interrupt_mask is used only on the guest->host ring buffer when + * sending requests to the host. The host does not use it on the host-> + * guest ring buffer to indicate whether it should be signaled. + */ +void hv_pkt_iter_close(struct vmbus_channel *channel) +{ + struct hv_ring_buffer_info *rbi = &channel->inbound; + u32 curr_write_sz, pending_sz, bytes_read, start_read_index; + + /* + * Make sure all reads are done before we update the read index since + * the writer may start writing to the read area once the read index + * is updated. + */ + virt_rmb(); + start_read_index = rbi->ring_buffer->read_index; + rbi->ring_buffer->read_index = rbi->priv_read_index; + + /* + * Older versions of Hyper-V (before WS2102 and Win8) do not + * implement pending_send_sz and simply poll if the host->guest + * ring buffer is full. No signaling is needed or expected. + */ + if (!rbi->ring_buffer->feature_bits.feat_pending_send_sz) + return; + + /* + * Issue a full memory barrier before making the signaling decision. + * If reading pending_send_sz were to be reordered and happen + * before we commit the new read_index, a race could occur. If the + * host were to set the pending_send_sz after we have sampled + * pending_send_sz, and the ring buffer blocks before we commit the + * read index, we could miss sending the interrupt. Issue a full + * memory barrier to address this. + */ + virt_mb(); + + /* + * If the pending_send_sz is zero, then the ring buffer is not + * blocked and there is no need to signal. This is far by the + * most common case, so exit quickly for best performance. + */ + pending_sz = READ_ONCE(rbi->ring_buffer->pending_send_sz); + if (!pending_sz) + return; + + /* + * Ensure the read of write_index in hv_get_bytes_to_write() + * happens after the read of pending_send_sz. + */ + virt_rmb(); + curr_write_sz = hv_get_bytes_to_write(rbi); + bytes_read = hv_pkt_iter_bytes_read(rbi, start_read_index); + + /* + * We want to signal the host only if we're transitioning + * from a "not enough free space" state to a "enough free + * space" state. For example, it's possible that this function + * could run and free up enough space to signal the host, and then + * run again and free up additional space before the host has a + * chance to clear the pending_send_sz. The 2nd invocation would + * be a null transition from "enough free space" to "enough free + * space", which doesn't warrant a signal. + * + * Exactly filling the ring buffer is treated as "not enough + * space". The ring buffer always must have at least one byte + * empty so the empty and full conditions are distinguishable. + * hv_get_bytes_to_write() doesn't fully tell the truth in + * this regard. + * + * So first check if we were in the "enough free space" state + * before we began the iteration. If so, the host was not + * blocked, and there's no need to signal. + */ + if (curr_write_sz - bytes_read > pending_sz) + return; + + /* + * Similarly, if the new state is "not enough space", then + * there's no need to signal. + */ + if (curr_write_sz <= pending_sz) + return; + + vmbus_setevent(channel); +} +EXPORT_SYMBOL_GPL(hv_pkt_iter_close); diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c new file mode 100644 index 000000000..fca092cfe --- /dev/null +++ b/drivers/hv/vmbus_drv.c @@ -0,0 +1,2088 @@ +/* + * 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/interrupt.h> +#include <linux/sysctl.h> +#include <linux/slab.h> +#include <linux/acpi.h> +#include <linux/completion.h> +#include <linux/hyperv.h> +#include <linux/kernel_stat.h> +#include <linux/clockchips.h> +#include <linux/cpu.h> +#include <linux/sched/task_stack.h> + +#include <asm/mshyperv.h> +#include <linux/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 "hyperv_vmbus.h" + +struct vmbus_dynid { + struct list_head node; + struct hv_vmbus_device_id id; +}; + +static struct acpi_device *hv_acpi_dev; + +static struct completion probe_event; + +static int hyperv_cpuhp_online; + +static void *hv_panic_page; + +/* + * 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 int hyperv_report_reg(void) +{ + return !sysctl_record_panic_msg || !hv_panic_page; +} + +static int hyperv_panic_event(struct notifier_block *nb, unsigned long val, + void *args) +{ + struct pt_regs *regs; + + vmbus_initiate_unload(true); + + /* + * Hyper-V should be notified only once about a panic. If we will be + * doing hyperv_report_panic_msg() later with kmsg data, don't do + * the notification here. + */ + if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE + && hyperv_report_reg()) { + regs = current_pt_regs(); + hyperv_report_panic(regs, val, false); + } + return NOTIFY_DONE; +} + +static int hyperv_die_event(struct notifier_block *nb, unsigned long val, + void *args) +{ + struct die_args *die = (struct die_args *)args; + struct pt_regs *regs = die->regs; + + /* + * Hyper-V should be notified only once about a panic. If we will be + * doing hyperv_report_panic_msg() later with kmsg data, don't do + * the notification here. + */ + if (hyperv_report_reg()) + hyperv_report_panic(regs, val, 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 const char *fb_mmio_name = "fb_range"; +static struct resource *fb_mmio; +static struct resource *hyperv_mmio; +static DEFINE_SEMAPHORE(hyperv_mmio_lock); + +static int vmbus_exists(void) +{ + if (hv_acpi_dev == 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; +} + +static u8 channel_monitor_offset(const struct vmbus_channel *channel) +{ + return (u8)channel->offermsg.monitorid % 32; +} + +static u32 channel_pending(const struct vmbus_channel *channel, + const struct hv_monitor_page *monitor_page) +{ + u8 monitor_group = channel_monitor_group(channel); + + return monitor_page->trigger_group[monitor_group].pending; +} + +static u32 channel_latency(const struct vmbus_channel *channel, + const struct hv_monitor_page *monitor_page) +{ + u8 monitor_group = channel_monitor_group(channel); + u8 monitor_offset = channel_monitor_offset(channel); + + return monitor_page->latency[monitor_group][monitor_offset]; +} + +static u32 channel_conn_id(struct vmbus_channel *channel, + struct hv_monitor_page *monitor_page) +{ + 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; +} + +static ssize_t id_show(struct device *dev, struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", hv_dev->channel->offermsg.child_relid); +} +static DEVICE_ATTR_RO(id); + +static ssize_t state_show(struct device *dev, struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", hv_dev->channel->state); +} +static DEVICE_ATTR_RO(state); + +static ssize_t monitor_id_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", hv_dev->channel->offermsg.monitorid); +} +static DEVICE_ATTR_RO(monitor_id); + +static ssize_t class_id_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "{%pUl}\n", + hv_dev->channel->offermsg.offer.if_type.b); +} +static DEVICE_ATTR_RO(class_id); + +static ssize_t device_id_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "{%pUl}\n", + hv_dev->channel->offermsg.offer.if_instance.b); +} +static DEVICE_ATTR_RO(device_id); + +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); +} +static DEVICE_ATTR_RO(modalias); + +#ifdef CONFIG_NUMA +static ssize_t numa_node_show(struct device *dev, + struct device_attribute *attr, char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + + return sprintf(buf, "%d\n", hv_dev->channel->numa_node); +} +static DEVICE_ATTR_RO(numa_node); +#endif + +static ssize_t server_monitor_pending_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_pending(hv_dev->channel, + vmbus_connection.monitor_pages[1])); +} +static DEVICE_ATTR_RO(server_monitor_pending); + +static ssize_t client_monitor_pending_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_pending(hv_dev->channel, + vmbus_connection.monitor_pages[1])); +} +static DEVICE_ATTR_RO(client_monitor_pending); + +static ssize_t server_monitor_latency_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_latency(hv_dev->channel, + vmbus_connection.monitor_pages[0])); +} +static DEVICE_ATTR_RO(server_monitor_latency); + +static ssize_t client_monitor_latency_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_latency(hv_dev->channel, + vmbus_connection.monitor_pages[1])); +} +static DEVICE_ATTR_RO(client_monitor_latency); + +static ssize_t server_monitor_conn_id_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_conn_id(hv_dev->channel, + vmbus_connection.monitor_pages[0])); +} +static DEVICE_ATTR_RO(server_monitor_conn_id); + +static ssize_t client_monitor_conn_id_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + + if (!hv_dev->channel) + return -ENODEV; + return sprintf(buf, "%d\n", + channel_conn_id(hv_dev->channel, + vmbus_connection.monitor_pages[1])); +} +static DEVICE_ATTR_RO(client_monitor_conn_id); + +static ssize_t out_intr_mask_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", outbound.current_interrupt_mask); +} +static DEVICE_ATTR_RO(out_intr_mask); + +static ssize_t out_read_index_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sprintf(buf, "%d\n", outbound.current_read_index); +} +static DEVICE_ATTR_RO(out_read_index); + +static ssize_t out_write_index_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sprintf(buf, "%d\n", outbound.current_write_index); +} +static DEVICE_ATTR_RO(out_write_index); + +static ssize_t out_read_bytes_avail_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sprintf(buf, "%d\n", outbound.bytes_avail_toread); +} +static DEVICE_ATTR_RO(out_read_bytes_avail); + +static ssize_t out_write_bytes_avail_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, + &outbound); + if (ret < 0) + return ret; + return sprintf(buf, "%d\n", outbound.bytes_avail_towrite); +} +static DEVICE_ATTR_RO(out_write_bytes_avail); + +static ssize_t in_intr_mask_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", inbound.current_interrupt_mask); +} +static DEVICE_ATTR_RO(in_intr_mask); + +static ssize_t in_read_index_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", inbound.current_read_index); +} +static DEVICE_ATTR_RO(in_read_index); + +static ssize_t in_write_index_show(struct device *dev, + struct device_attribute *dev_attr, char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", inbound.current_write_index); +} +static DEVICE_ATTR_RO(in_write_index); + +static ssize_t in_read_bytes_avail_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", inbound.bytes_avail_toread); +} +static DEVICE_ATTR_RO(in_read_bytes_avail); + +static ssize_t in_write_bytes_avail_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + + ret = hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + if (ret < 0) + return ret; + + return sprintf(buf, "%d\n", inbound.bytes_avail_towrite); +} +static DEVICE_ATTR_RO(in_write_bytes_avail); + +static ssize_t channel_vp_mapping_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + 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; + struct list_head *cur; + + if (!channel) + return -ENODEV; + + tot_written = snprintf(buf, buf_size, "%u:%u\n", + channel->offermsg.child_relid, channel->target_cpu); + + spin_lock_irqsave(&channel->lock, flags); + + 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; + } + + spin_unlock_irqrestore(&channel->lock, flags); + + return tot_written; +} +static DEVICE_ATTR_RO(channel_vp_mapping); + +static ssize_t vendor_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + return sprintf(buf, "0x%x\n", hv_dev->vendor_id); +} +static DEVICE_ATTR_RO(vendor); + +static ssize_t device_show(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + return sprintf(buf, "0x%x\n", hv_dev->device_id); +} +static DEVICE_ATTR_RO(device); + +/* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */ +static struct attribute *vmbus_dev_attrs[] = { + &dev_attr_id.attr, + &dev_attr_state.attr, + &dev_attr_monitor_id.attr, + &dev_attr_class_id.attr, + &dev_attr_device_id.attr, + &dev_attr_modalias.attr, +#ifdef CONFIG_NUMA + &dev_attr_numa_node.attr, +#endif + &dev_attr_server_monitor_pending.attr, + &dev_attr_client_monitor_pending.attr, + &dev_attr_server_monitor_latency.attr, + &dev_attr_client_monitor_latency.attr, + &dev_attr_server_monitor_conn_id.attr, + &dev_attr_client_monitor_conn_id.attr, + &dev_attr_out_intr_mask.attr, + &dev_attr_out_read_index.attr, + &dev_attr_out_write_index.attr, + &dev_attr_out_read_bytes_avail.attr, + &dev_attr_out_write_bytes_avail.attr, + &dev_attr_in_intr_mask.attr, + &dev_attr_in_read_index.attr, + &dev_attr_in_write_index.attr, + &dev_attr_in_read_bytes_avail.attr, + &dev_attr_in_write_bytes_avail.attr, + &dev_attr_channel_vp_mapping.attr, + &dev_attr_vendor.attr, + &dev_attr_device.attr, + NULL, +}; + +/* + * 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); + +/* + * vmbus_uevent - add uevent for our device + * + * This routine is invoked when a device is added or removed on the vmbus to + * generate a uevent to udev in the userspace. The udev will then look at its + * rule and the uevent generated here to load the appropriate driver + * + * The alias string will be of the form vmbus:guid where guid is the string + * 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) +{ + 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; + +static inline bool is_null_guid(const uuid_le *guid) +{ + if (uuid_le_cmp(*guid, null_guid)) + return false; + return true; +} + +/* + * 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, + const uuid_le *guid) +{ + const struct hv_vmbus_device_id *id = NULL; + struct vmbus_dynid *dynid; + + /* Look at the dynamic ids first, before the static ones */ + spin_lock(&drv->dynids.lock); + list_for_each_entry(dynid, &drv->dynids.list, node) { + if (!uuid_le_cmp(dynid->id.guid, *guid)) { + id = &dynid->id; + break; + } + } + spin_unlock(&drv->dynids.lock); + + if (id) + return id; + + id = drv->id_table; + if (id == NULL) + return NULL; /* empty device table */ + + for (; !is_null_guid(&id->guid); id++) + if (!uuid_le_cmp(id->guid, *guid)) + return id; + + return NULL; +} + +/* 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) +{ + struct vmbus_dynid *dynid; + + dynid = kzalloc(sizeof(*dynid), GFP_KERNEL); + if (!dynid) + return -ENOMEM; + + dynid->id.guid = *guid; + + spin_lock(&drv->dynids.lock); + list_add_tail(&dynid->node, &drv->dynids.list); + spin_unlock(&drv->dynids.lock); + + return driver_attach(&drv->driver); +} + +static void vmbus_free_dynids(struct hv_driver *drv) +{ + struct vmbus_dynid *dynid, *n; + + spin_lock(&drv->dynids.lock); + list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) { + list_del(&dynid->node); + kfree(dynid); + } + spin_unlock(&drv->dynids.lock); +} + +/* + * store_new_id - sysfs frontend to vmbus_add_dynid() + * + * Allow GUIDs to be added to an existing driver via sysfs. + */ +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; + ssize_t retval; + + retval = uuid_le_to_bin(buf, &guid); + if (retval) + return retval; + + if (hv_vmbus_get_id(drv, &guid)) + return -EEXIST; + + retval = vmbus_add_dynid(drv, &guid); + if (retval) + return retval; + return count; +} +static DRIVER_ATTR_WO(new_id); + +/* + * store_remove_id - remove a PCI device ID from this driver + * + * Removes a dynamic pci device ID to this driver. + */ +static ssize_t remove_id_store(struct device_driver *driver, const char *buf, + size_t count) +{ + struct hv_driver *drv = drv_to_hv_drv(driver); + struct vmbus_dynid *dynid, *n; + uuid_le guid; + ssize_t retval; + + retval = uuid_le_to_bin(buf, &guid); + if (retval) + return retval; + + retval = -ENODEV; + spin_lock(&drv->dynids.lock); + 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)) { + list_del(&dynid->node); + kfree(dynid); + retval = count; + break; + } + } + spin_unlock(&drv->dynids.lock); + + return retval; +} +static DRIVER_ATTR_WO(remove_id); + +static struct attribute *vmbus_drv_attrs[] = { + &driver_attr_new_id.attr, + &driver_attr_remove_id.attr, + NULL, +}; +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) +{ + 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. */ + if (is_hvsock_channel(hv_dev->channel)) + return drv->hvsock; + + if (hv_vmbus_get_id(drv, &hv_dev->dev_type)) + return 1; + + return 0; +} + +/* + * vmbus_probe - Add the new vmbus's child device + */ +static int vmbus_probe(struct device *child_device) +{ + int ret = 0; + struct hv_driver *drv = + drv_to_hv_drv(child_device->driver); + struct hv_device *dev = device_to_hv_device(child_device); + const struct hv_vmbus_device_id *dev_id; + + dev_id = hv_vmbus_get_id(drv, &dev->dev_type); + if (drv->probe) { + ret = drv->probe(dev, dev_id); + if (ret != 0) + pr_err("probe failed for device %s (%d)\n", + dev_name(child_device), ret); + + } else { + pr_err("probe not set for driver %s\n", + dev_name(child_device)); + ret = -ENODEV; + } + return ret; +} + +/* + * vmbus_remove - Remove a vmbus device + */ +static int vmbus_remove(struct device *child_device) +{ + struct hv_driver *drv; + struct hv_device *dev = device_to_hv_device(child_device); + + if (child_device->driver) { + drv = drv_to_hv_drv(child_device->driver); + if (drv->remove) + drv->remove(dev); + } + + return 0; +} + + +/* + * vmbus_shutdown - Shutdown a vmbus device + */ +static void vmbus_shutdown(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; + + drv = drv_to_hv_drv(child_device->driver); + + if (drv->shutdown) + drv->shutdown(dev); +} + + +/* + * vmbus_device_release - Final callback release of the vmbus child device + */ +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; + + mutex_lock(&vmbus_connection.channel_mutex); + hv_process_channel_removal(channel->offermsg.child_relid); + mutex_unlock(&vmbus_connection.channel_mutex); + kfree(hv_dev); + +} + +/* The one and only one */ +static struct bus_type hv_bus = { + .name = "vmbus", + .match = vmbus_match, + .shutdown = vmbus_shutdown, + .remove = vmbus_remove, + .probe = vmbus_probe, + .uevent = vmbus_uevent, + .dev_groups = vmbus_dev_groups, + .drv_groups = vmbus_drv_groups, +}; + +struct onmessage_work_context { + struct work_struct work; + struct hv_message msg; +}; + +static void vmbus_onmessage_work(struct work_struct *work) +{ + struct onmessage_work_context *ctx; + + /* Do not process messages if we're in DISCONNECTED state */ + if (vmbus_connection.conn_state == DISCONNECTED) + return; + + ctx = container_of(work, struct onmessage_work_context, + work); + vmbus_onmessage(&ctx->msg); + kfree(ctx); +} + +static void hv_process_timer_expiration(struct hv_message *msg, + struct hv_per_cpu_context *hv_cpu) +{ + 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 vmbus_channel_message_header *hdr; + const struct vmbus_channel_message_table_entry *entry; + struct onmessage_work_context *ctx; + u32 message_type = msg->header.message_type; + + if (message_type == HVMSG_NONE) + /* no msg */ + return; + + hdr = (struct vmbus_channel_message_header *)msg->u.payload; + + trace_vmbus_on_msg_dpc(hdr); + + if (hdr->msgtype >= CHANNELMSG_COUNT) { + WARN_ONCE(1, "unknown msgtype=%d\n", hdr->msgtype); + goto msg_handled; + } + + entry = &channel_message_table[hdr->msgtype]; + + if (!entry->message_handler) + goto msg_handled; + + if (entry->handler_type == VMHT_BLOCKING) { + ctx = kmalloc(sizeof(*ctx), GFP_ATOMIC); + if (ctx == NULL) + return; + + INIT_WORK(&ctx->work, vmbus_onmessage_work); + memcpy(&ctx->msg, msg, sizeof(*msg)); + + /* + * 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. + */ + switch (hdr->msgtype) { + case CHANNELMSG_RESCIND_CHANNELOFFER: + /* + * If we are handling the rescind message; + * schedule the work on the global work queue. + */ + schedule_work_on(vmbus_connection.connect_cpu, + &ctx->work); + break; + + case CHANNELMSG_OFFERCHANNEL: + atomic_inc(&vmbus_connection.offer_in_progress); + queue_work_on(vmbus_connection.connect_cpu, + vmbus_connection.work_queue, + &ctx->work); + break; + + default: + queue_work(vmbus_connection.work_queue, &ctx->work); + } + } else + entry->message_handler(hdr); + +msg_handled: + vmbus_signal_eom(msg, message_type); +} + + +/* + * Direct callback for channels using other deferred processing + */ +static void vmbus_channel_isr(struct vmbus_channel *channel) +{ + void (*callback_fn)(void *); + + callback_fn = READ_ONCE(channel->onchannel_callback); + if (likely(callback_fn != NULL)) + (*callback_fn)(channel->channel_callback_context); +} + +/* + * Schedule all channels with events pending + */ +static void vmbus_chan_sched(struct hv_per_cpu_context *hv_cpu) +{ + unsigned long *recv_int_page; + u32 maxbits, relid; + + 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; + + maxbits = HV_EVENT_FLAGS_COUNT; + recv_int_page = event->flags; + } + + if (unlikely(!recv_int_page)) + return; + + for_each_set_bit(relid, recv_int_page, maxbits) { + struct vmbus_channel *channel; + + if (!sync_test_and_clear_bit(relid, recv_int_page)) + continue; + + /* Special case - vmbus channel protocol msg */ + if (relid == 0) + continue; + + 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; + + if (channel->rescind) + continue; + + trace_vmbus_chan_sched(channel); + + ++channel->interrupts; + + switch (channel->callback_mode) { + case HV_CALL_ISR: + vmbus_channel_isr(channel); + break; + + case HV_CALL_BATCHED: + hv_begin_read(&channel->inbound); + /* fallthrough */ + case HV_CALL_DIRECT: + tasklet_schedule(&channel->callback_event); + } + } + + rcu_read_unlock(); + } +} + +static void vmbus_isr(void) +{ + 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)) + 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; + + /* 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 + tasklet_schedule(&hv_cpu->msg_dpc); + } + + add_interrupt_randomness(HYPERVISOR_CALLBACK_VECTOR); +} + +/* + * 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) +{ + size_t bytes_written; + phys_addr_t panic_pa; + + /* We are only interested in panics. */ + if ((reason != KMSG_DUMP_PANIC) || (!sysctl_record_panic_msg)) + return; + + panic_pa = virt_to_phys(hv_panic_page); + + /* + * Write dump contents to the page. No need to synchronize; panic should + * be single-threaded. + */ + kmsg_dump_get_buffer(dumper, true, hv_panic_page, PAGE_SIZE, + &bytes_written); + if (bytes_written) + hyperv_report_panic_msg(panic_pa, bytes_written); +} + +static struct kmsg_dumper hv_kmsg_dumper = { + .dump = hv_kmsg_dump, +}; + +static struct ctl_table_header *hv_ctl_table_hdr; +static int zero; +static int one = 1; + +/* + * 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 + }, + {} +}; + +static struct ctl_table hv_root_table[] = { + { + .procname = "kernel", + .mode = 0555, + .child = hv_ctl_table + }, + {} +}; + +/* + * vmbus_bus_init -Main vmbus driver initialization routine. + * + * Here, we + * - initialize the vmbus driver context + * - invoke the vmbus hv main init routine + * - retrieve the channel offers + */ +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); + return ret; + } + + ret = bus_register(&hv_bus); + 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. + */ + 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 (ret) + goto err_connect; + + /* + * Only register if the crash MSRs are available + */ + 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); + free_page( + (unsigned long)hv_panic_page); + hv_panic_page = NULL; + } + } else + pr_err("Hyper-V: panic message page memory " + "allocation failed"); + } + + register_die_notifier(&hyperv_die_block); + } + + /* + * Always register the panic notifier because we need to unload + * the VMbus channel connection to prevent any VMbus + * activity after the VM panics. + */ + atomic_notifier_chain_register(&panic_notifier_list, + &hyperv_panic_block); + + vmbus_request_offers(); + + return 0; + +err_connect: + cpuhp_remove_state(hyperv_cpuhp_online); +err_alloc: + hv_synic_free(); + hv_remove_vmbus_irq(); + + bus_unregister(&hv_bus); + unregister_sysctl_table(hv_ctl_table_hdr); + hv_ctl_table_hdr = NULL; + return ret; +} + +/** + * __vmbus_child_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 + * + * Registers the given driver with Linux through the 'driver_register()' call + * and sets up the hyper-v vmbus handling for this driver. + * It will return the state of the 'driver_register()' call. + * + */ +int __vmbus_driver_register(struct hv_driver *hv_driver, struct module *owner, const char *mod_name) +{ + int ret; + + pr_info("registering driver %s\n", hv_driver->name); + + ret = vmbus_exists(); + if (ret < 0) + return ret; + + hv_driver->driver.name = hv_driver->name; + hv_driver->driver.owner = owner; + hv_driver->driver.mod_name = mod_name; + hv_driver->driver.bus = &hv_bus; + + spin_lock_init(&hv_driver->dynids.lock); + INIT_LIST_HEAD(&hv_driver->dynids.list); + + ret = driver_register(&hv_driver->driver); + + return ret; +} +EXPORT_SYMBOL_GPL(__vmbus_driver_register); + +/** + * vmbus_driver_unregister() - Unregister a vmbus's driver + * @hv_driver: Pointer to driver structure you want to + * un-register + * + * Un-register the given driver that was previous registered with a call to + * vmbus_driver_register() + */ +void vmbus_driver_unregister(struct hv_driver *hv_driver) +{ + pr_info("unregistering driver %s\n", hv_driver->name); + + if (!vmbus_exists()) { + driver_unregister(&hv_driver->driver); + vmbus_free_dynids(hv_driver); + } +} +EXPORT_SYMBOL_GPL(vmbus_driver_unregister); + + +/* + * Called when last reference to channel is gone. + */ +static void vmbus_chan_release(struct kobject *kobj) +{ + struct vmbus_channel *channel + = container_of(kobj, struct vmbus_channel, kobj); + + kfree_rcu(channel, rcu); +} + +struct vmbus_chan_attribute { + struct attribute attr; + ssize_t (*show)(const struct vmbus_channel *chan, char *buf); + ssize_t (*store)(struct vmbus_channel *chan, + const char *buf, size_t count); +}; +#define VMBUS_CHAN_ATTR(_name, _mode, _show, _store) \ + struct vmbus_chan_attribute chan_attr_##_name \ + = __ATTR(_name, _mode, _show, _store) +#define VMBUS_CHAN_ATTR_RW(_name) \ + struct vmbus_chan_attribute chan_attr_##_name = __ATTR_RW(_name) +#define VMBUS_CHAN_ATTR_RO(_name) \ + struct vmbus_chan_attribute chan_attr_##_name = __ATTR_RO(_name) +#define VMBUS_CHAN_ATTR_WO(_name) \ + struct vmbus_chan_attribute chan_attr_##_name = __ATTR_WO(_name) + +static ssize_t vmbus_chan_attr_show(struct kobject *kobj, + struct attribute *attr, char *buf) +{ + const struct vmbus_chan_attribute *attribute + = container_of(attr, struct vmbus_chan_attribute, attr); + const 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 const struct sysfs_ops vmbus_chan_sysfs_ops = { + .show = vmbus_chan_attr_show, +}; + +static ssize_t out_mask_show(const struct vmbus_channel *channel, char *buf) +{ + const struct hv_ring_buffer_info *rbi = &channel->outbound; + + return sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); +} +static VMBUS_CHAN_ATTR_RO(out_mask); + +static ssize_t in_mask_show(const struct vmbus_channel *channel, char *buf) +{ + const struct hv_ring_buffer_info *rbi = &channel->inbound; + + return sprintf(buf, "%u\n", rbi->ring_buffer->interrupt_mask); +} +static VMBUS_CHAN_ATTR_RO(in_mask); + +static ssize_t read_avail_show(const struct vmbus_channel *channel, char *buf) +{ + const struct hv_ring_buffer_info *rbi = &channel->inbound; + + return sprintf(buf, "%u\n", hv_get_bytes_to_read(rbi)); +} +static VMBUS_CHAN_ATTR_RO(read_avail); + +static ssize_t write_avail_show(const struct vmbus_channel *channel, char *buf) +{ + const struct hv_ring_buffer_info *rbi = &channel->outbound; + + return sprintf(buf, "%u\n", hv_get_bytes_to_write(rbi)); +} +static VMBUS_CHAN_ATTR_RO(write_avail); + +static ssize_t show_target_cpu(const 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, + 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 ssize_t channel_latency_show(const 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 ssize_t channel_interrupts_show(const 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 ssize_t channel_events_show(const 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 ssize_t subchannel_monitor_id_show(const 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 ssize_t subchannel_id_show(const struct vmbus_channel *channel, + char *buf) +{ + return sprintf(buf, "%u\n", + channel->offermsg.offer.sub_channel_index); +} +static VMBUS_CHAN_ATTR_RO(subchannel_id); + +static struct attribute *vmbus_chan_attrs[] = { + &chan_attr_out_mask.attr, + &chan_attr_in_mask.attr, + &chan_attr_read_avail.attr, + &chan_attr_write_avail.attr, + &chan_attr_cpu.attr, + &chan_attr_pending.attr, + &chan_attr_latency.attr, + &chan_attr_interrupts.attr, + &chan_attr_events.attr, + &chan_attr_monitor_id.attr, + &chan_attr_subchannel_id.attr, + 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 struct attribute_group vmbus_chan_group = { + .attrs = vmbus_chan_attrs, + .is_visible = vmbus_chan_attr_is_visible +}; + +static struct kobj_type vmbus_chan_ktype = { + .sysfs_ops = &vmbus_chan_sysfs_ops, + .release = vmbus_chan_release, +}; + +/* + * 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; + + kobj->kset = dev->channels_kset; + ret = kobject_init_and_add(kobj, &vmbus_chan_ktype, NULL, + "%u", relid); + 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); + + return 0; +} + +/* + * 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 vmbus_channel *channel) +{ + struct hv_device *child_device_obj; + + child_device_obj = kzalloc(sizeof(struct hv_device), GFP_KERNEL); + if (!child_device_obj) { + pr_err("Unable to allocate device object for child device\n"); + return NULL; + } + + 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 */ + + + return child_device_obj; +} + +/* + * vmbus_device_register - Register the child device + */ +int vmbus_device_register(struct hv_device *child_device_obj) +{ + struct kobject *kobj = &child_device_obj->device.kobj; + int ret; + + dev_set_name(&child_device_obj->device, "%pUl", + child_device_obj->channel->offermsg.offer.if_instance.b); + + child_device_obj->device.bus = &hv_bus; + child_device_obj->device.parent = &hv_acpi_dev->dev; + child_device_obj->device.release = vmbus_device_release; + + /* + * Register with the LDM. This will kick off the driver/device + * binding...which will eventually call vmbus_match() and vmbus_probe() + */ + ret = device_register(&child_device_obj->device); + if (ret) { + pr_err("Unable to register child device\n"); + return ret; + } + + child_device_obj->channels_kset = kset_create_and_add("channels", + NULL, kobj); + if (!child_device_obj->channels_kset) { + ret = -ENOMEM; + goto err_dev_unregister; + } + + ret = vmbus_add_channel_kobj(child_device_obj, + child_device_obj->channel); + if (ret) { + pr_err("Unable to register primary channeln"); + goto err_kset_unregister; + } + + return 0; + +err_kset_unregister: + kset_unregister(child_device_obj->channels_kset); + +err_dev_unregister: + device_unregister(&child_device_obj->device); + return ret; +} + +/* + * vmbus_device_unregister - Remove the specified child device + * from the vmbus. + */ +void vmbus_device_unregister(struct hv_device *device_obj) +{ + pr_debug("child device %s unregistered\n", + dev_name(&device_obj->device)); + + kset_unregister(device_obj->channels_kset); + + /* + * Kick off the process of unregistering the device. + * This will call vmbus_remove() and eventually vmbus_device_release() + */ + device_unregister(&device_obj->device); +} + + +/* + * 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; + resource_size_t end = 0; + struct resource *new_res; + struct resource **old_res = &hyperv_mmio; + struct resource **prev_res = NULL; + + switch (res->type) { + + /* + * "Address" descriptors are for bus windows. Ignore + * "memory" descriptors, which are for registers on + * devices. + */ + case ACPI_RESOURCE_TYPE_ADDRESS32: + start = res->data.address32.address.minimum; + end = res->data.address32.address.maximum; + break; + + case ACPI_RESOURCE_TYPE_ADDRESS64: + start = res->data.address64.address.minimum; + end = res->data.address64.address.maximum; + break; + + default: + /* Unused resource type */ + return AE_OK; + + } + /* + * Ignore ranges that are below 1MB, as they're not + * necessary or useful here. + */ + if (end < 0x100000) + return AE_OK; + + new_res = kzalloc(sizeof(*new_res), GFP_ATOMIC); + if (!new_res) + return AE_NO_MEMORY; + + /* If this range overlaps the virtual TPM, truncate it. */ + if (end > VTPM_BASE_ADDRESS && start < VTPM_BASE_ADDRESS) + end = VTPM_BASE_ADDRESS; + + new_res->name = "hyperv mmio"; + new_res->flags = IORESOURCE_MEM; + new_res->start = start; + new_res->end = end; + + /* + * If two ranges are adjacent, merge them. + */ + do { + if (!*old_res) { + *old_res = new_res; + break; + } + + if (((*old_res)->end + 1) == new_res->start) { + (*old_res)->end = new_res->end; + kfree(new_res); + break; + } + + if ((*old_res)->start == new_res->end + 1) { + (*old_res)->start = new_res->start; + kfree(new_res); + break; + } + + if ((*old_res)->start > new_res->end) { + new_res->sibling = *old_res; + if (prev_res) + (*prev_res)->sibling = new_res; + *old_res = new_res; + break; + } + + prev_res = old_res; + old_res = &(*old_res)->sibling; + + } while (1); + + return AE_OK; +} + +static int vmbus_acpi_remove(struct acpi_device *device) +{ + struct resource *cur_res; + struct resource *next_res; + + if (hyperv_mmio) { + if (fb_mmio) { + __release_region(hyperv_mmio, fb_mmio->start, + resource_size(fb_mmio)); + fb_mmio = NULL; + } + + for (cur_res = hyperv_mmio; cur_res; cur_res = next_res) { + next_res = cur_res->sibling; + kfree(cur_res); + } + } + + return 0; +} + +static void vmbus_reserve_fb(void) +{ + int size; + /* + * 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); + } + } +} + +/** + * vmbus_allocate_mmio() - Pick a memory-mapped I/O range. + * @new: If successful, supplied a pointer to the + * allocated MMIO space. + * @device_obj: Identifies the caller + * @min: Minimum guest physical address of the + * allocation + * @max: Maximum guest physical address + * @size: Size of the range to be allocated + * @align: Alignment of the range to be allocated + * @fb_overlap_ok: Whether this allocation can be allowed + * to overlap the video frame buffer. + * + * This function walks the resources granted to VMBus by the + * _CRS object in the ACPI namespace underneath the parent + * "bridge" whether that's a root PCI bus in the Generation 1 + * case or a Module Device in the Generation 2 case. It then + * attempts to allocate from the global MMIO pool in a way that + * matches the constraints supplied in these parameters and by + * that _CRS. + * + * Return: 0 on success, -errno on failure + */ +int vmbus_allocate_mmio(struct resource **new, struct hv_device *device_obj, + resource_size_t min, resource_size_t max, + resource_size_t size, resource_size_t align, + bool fb_overlap_ok) +{ + struct resource *iter, *shadow; + resource_size_t range_min, range_max, start; + const char *dev_n = dev_name(&device_obj->device); + int retval; + + retval = -ENXIO; + down(&hyperv_mmio_lock); + + /* + * If overlaps with frame buffers are allowed, then first attempt to + * make the allocation from within the reserved region. Because it + * is already reserved, no shadow allocation is necessary. + */ + if (fb_overlap_ok && fb_mmio && !(min > fb_mmio->end) && + !(max < fb_mmio->start)) { + + range_min = fb_mmio->start; + range_max = fb_mmio->end; + start = (range_min + align - 1) & ~(align - 1); + for (; start + size - 1 <= range_max; start += align) { + *new = request_mem_region_exclusive(start, size, dev_n); + if (*new) { + retval = 0; + goto exit; + } + } + } + + for (iter = hyperv_mmio; iter; iter = iter->sibling) { + if ((iter->start >= max) || (iter->end <= min)) + continue; + + range_min = iter->start; + range_max = iter->end; + start = (range_min + align - 1) & ~(align - 1); + for (; start + size - 1 <= range_max; start += align) { + shadow = __request_region(iter, start, size, NULL, + IORESOURCE_BUSY); + if (!shadow) + continue; + + *new = request_mem_region_exclusive(start, size, dev_n); + if (*new) { + shadow->name = (char *)*new; + retval = 0; + goto exit; + } + + __release_region(iter, start, size); + } + } + +exit: + up(&hyperv_mmio_lock); + return retval; +} +EXPORT_SYMBOL_GPL(vmbus_allocate_mmio); + +/** + * vmbus_free_mmio() - Free a memory-mapped I/O range. + * @start: Base address of region to release. + * @size: Size of the range to be allocated + * + * This function releases anything requested by + * vmbus_mmio_allocate(). + */ +void vmbus_free_mmio(resource_size_t start, resource_size_t size) +{ + struct resource *iter; + + down(&hyperv_mmio_lock); + for (iter = hyperv_mmio; iter; iter = iter->sibling) { + if ((iter->start >= start + size) || (iter->end <= start)) + continue; + + __release_region(iter, start, size); + } + release_mem_region(start, size); + up(&hyperv_mmio_lock); + +} +EXPORT_SYMBOL_GPL(vmbus_free_mmio); + +static int vmbus_acpi_add(struct acpi_device *device) +{ + acpi_status result; + int ret_val = -ENODEV; + struct acpi_device *ancestor; + + hv_acpi_dev = device; + + result = acpi_walk_resources(device->handle, METHOD_NAME__CRS, + vmbus_walk_resources, NULL); + + if (ACPI_FAILURE(result)) + goto acpi_walk_err; + /* + * 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) { + result = acpi_walk_resources(ancestor->handle, METHOD_NAME__CRS, + vmbus_walk_resources, NULL); + + if (ACPI_FAILURE(result)) + continue; + if (hyperv_mmio) { + vmbus_reserve_fb(); + break; + } + } + ret_val = 0; + +acpi_walk_err: + complete(&probe_event); + if (ret_val) + vmbus_acpi_remove(device); + return ret_val; +} + +static const 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, + }, +}; + +static void hv_kexec_handler(void) +{ + hv_synic_clockevents_cleanup(); + vmbus_initiate_unload(false); + /* 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) +{ + 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. + */ + hv_synic_cleanup(smp_processor_id()); + hyperv_cleanup(); +}; + +static int __init hv_acpi_init(void) +{ + int ret, t; + + if (!hv_is_hyperv_initialized()) + return -ENODEV; + + init_completion(&probe_event); + + /* + * Get ACPI resources first. + */ + ret = acpi_bus_register_driver(&vmbus_acpi_driver); + + if (ret) + return ret; + + t = wait_for_completion_timeout(&probe_event, 5*HZ); + if (t == 0) { + ret = -ETIMEDOUT; + goto cleanup; + } + + ret = vmbus_bus_init(); + if (ret) + goto cleanup; + + hv_setup_kexec_handler(hv_kexec_handler); + hv_setup_crash_handler(hv_crash_handler); + + return 0; + +cleanup: + acpi_bus_unregister_driver(&vmbus_acpi_driver); + hv_acpi_dev = NULL; + return ret; +} + +static void __exit vmbus_exit(void) +{ + int cpu; + + hv_remove_kexec_handler(); + hv_remove_crash_handler(); + vmbus_connection.conn_state = DISCONNECTED; + hv_synic_clockevents_cleanup(); + vmbus_disconnect(); + hv_remove_vmbus_irq(); + 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); + } + vmbus_free_channels(); + + if (ms_hyperv.misc_features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) { + kmsg_dump_unregister(&hv_kmsg_dumper); + unregister_die_notifier(&hyperv_die_block); + } + + /* + * The panic notifier is always registered, hence we should + * also unconditionally unregister it here as well. + */ + atomic_notifier_chain_unregister(&panic_notifier_list, + &hyperv_panic_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); +} + + +MODULE_LICENSE("GPL"); + +subsys_initcall(hv_acpi_init); +module_exit(vmbus_exit); |