diff options
Diffstat (limited to 'kernel/drivers/hv')
| -rw-r--r-- | kernel/drivers/hv/Kconfig | 22 | ||||
| -rw-r--r-- | kernel/drivers/hv/Makefile | 8 | ||||
| -rw-r--r-- | kernel/drivers/hv/channel.c | 956 | ||||
| -rw-r--r-- | kernel/drivers/hv/channel_mgmt.c | 862 | ||||
| -rw-r--r-- | kernel/drivers/hv/connection.c | 480 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv.c | 535 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv_balloon.c | 1616 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv_fcopy.c | 441 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv_kvp.c | 722 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv_snapshot.c | 303 | ||||
| -rw-r--r-- | kernel/drivers/hv/hv_util.c | 447 | ||||
| -rw-r--r-- | kernel/drivers/hv/hyperv_vmbus.h | 738 | ||||
| -rw-r--r-- | kernel/drivers/hv/ring_buffer.c | 566 | ||||
| -rw-r--r-- | kernel/drivers/hv/vmbus_drv.c | 1114 |
14 files changed, 8810 insertions, 0 deletions
diff --git a/kernel/drivers/hv/Kconfig b/kernel/drivers/hv/Kconfig new file mode 100644 index 000000000..0403b51d2 --- /dev/null +++ b/kernel/drivers/hv/Kconfig @@ -0,0 +1,22 @@ +menu "Microsoft Hyper-V guest support" + +config HYPERV + tristate "Microsoft Hyper-V client drivers" + depends on X86 && ACPI && PCI && X86_LOCAL_APIC && HYPERVISOR_GUEST + help + Select this option to run Linux as a Hyper-V client operating + system. + +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/kernel/drivers/hv/Makefile b/kernel/drivers/hv/Makefile new file mode 100644 index 000000000..5e4dfa4cf --- /dev/null +++ b/kernel/drivers/hv/Makefile @@ -0,0 +1,8 @@ +obj-$(CONFIG_HYPERV) += hv_vmbus.o +obj-$(CONFIG_HYPERV_UTILS) += hv_utils.o +obj-$(CONFIG_HYPERV_BALLOON) += hv_balloon.o + +hv_vmbus-y := vmbus_drv.o \ + hv.o connection.o channel.o \ + channel_mgmt.o ring_buffer.o +hv_utils-y := hv_util.o hv_kvp.o hv_snapshot.o hv_fcopy.o diff --git a/kernel/drivers/hv/channel.c b/kernel/drivers/hv/channel.c new file mode 100644 index 000000000..54da66dc7 --- /dev/null +++ b/kernel/drivers/hv/channel.c @@ -0,0 +1,956 @@ +/* + * 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 "hyperv_vmbus.h" + +#define NUM_PAGES_SPANNED(addr, len) \ +((PAGE_ALIGN(addr + len) >> PAGE_SHIFT) - (addr >> PAGE_SHIFT)) + +/* + * vmbus_setevent- Trigger an event notification on the specified + * channel. + */ +static void vmbus_setevent(struct vmbus_channel *channel) +{ + struct hv_monitor_page *monitorpage; + + if (channel->offermsg.monitor_allocated) { + /* Each u32 represents 32 channels */ + sync_set_bit(channel->offermsg.child_relid & 31, + (unsigned long *) vmbus_connection.send_int_page + + (channel->offermsg.child_relid >> 5)); + + /* 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); + } +} + +/* + * 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; + void *in, *out; + unsigned long flags; + int ret, err = 0; + unsigned long t; + + 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 */ + out = (void *)__get_free_pages(GFP_KERNEL|__GFP_ZERO, + get_order(send_ringbuffer_size + recv_ringbuffer_size)); + + if (!out) { + err = -ENOMEM; + goto error0; + } + + in = (void *)((unsigned long)out + send_ringbuffer_size); + + newchannel->ringbuffer_pages = out; + newchannel->ringbuffer_pagecount = (send_ringbuffer_size + + recv_ringbuffer_size) >> PAGE_SHIFT; + + ret = hv_ringbuffer_init( + &newchannel->outbound, out, send_ringbuffer_size); + + if (ret != 0) { + err = ret; + goto error0; + } + + ret = hv_ringbuffer_init( + &newchannel->inbound, in, recv_ringbuffer_size); + if (ret != 0) { + err = ret; + goto error0; + } + + + /* Establish the gpadl for the ring buffer */ + newchannel->ringbuffer_gpadlhandle = 0; + + ret = vmbus_establish_gpadl(newchannel, + newchannel->outbound.ring_buffer, + send_ringbuffer_size + + recv_ringbuffer_size, + &newchannel->ringbuffer_gpadlhandle); + + if (ret != 0) { + err = ret; + goto error0; + } + + /* 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_gpadl; + } + + init_completion(&open_info->waitevent); + + 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_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); + + ret = vmbus_post_msg(open_msg, + sizeof(struct vmbus_channel_open_channel)); + + if (ret != 0) { + err = ret; + goto error1; + } + + t = wait_for_completion_timeout(&open_info->waitevent, 5*HZ); + if (t == 0) { + err = -ETIMEDOUT; + goto error1; + } + + + if (open_info->response.open_result.status) + err = open_info->response.open_result.status; + + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&open_info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + if (err == 0) + newchannel->state = CHANNEL_OPENED_STATE; + + kfree(open_info); + return err; + +error1: + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&open_info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + +error_gpadl: + vmbus_teardown_gpadl(newchannel, newchannel->ringbuffer_gpadlhandle); + +error0: + free_pages((unsigned long)out, + get_order(send_ringbuffer_size + recv_ringbuffer_size)); + kfree(open_info); + newchannel->state = CHANNEL_OPEN_STATE; + return err; +} +EXPORT_SYMBOL_GPL(vmbus_open); + +/* + * create_gpadl_header - Creates a gpadl for the specified buffer + */ +static int create_gpadl_header(void *kbuffer, u32 size, + struct vmbus_channel_msginfo **msginfo, + u32 *messagecount) +{ + 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] = slow_virt_to_phys( + kbuffer + PAGE_SIZE * i) >> PAGE_SHIFT; + *msginfo = msgheader; + *messagecount = 1; + + 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; + (*messagecount)++; + 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 gurantees that this is ok. + */ + for (i = 0; i < pfncurr; i++) + gpadl_body->pfn[i] = slow_virt_to_phys( + kbuffer + PAGE_SIZE * (pfnsum + i)) >> + PAGE_SHIFT; + + /* 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; + 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] = slow_virt_to_phys( + kbuffer + PAGE_SIZE * i) >> PAGE_SHIFT; + + *msginfo = msgheader; + *messagecount = 1; + } + + return 0; +nomem: + kfree(msgheader); + kfree(msgbody); + return -ENOMEM; +} + +/* + * vmbus_establish_gpadl - Estabish 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; + u32 msgcount; + 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, &msgcount); + if (ret) + return ret; + + init_completion(&msginfo->waitevent); + + 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); + + ret = vmbus_post_msg(gpadlmsg, msginfo->msgsize - + sizeof(*msginfo)); + if (ret != 0) + goto cleanup; + + if (msgcount > 1) { + 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)); + if (ret != 0) + goto cleanup; + + } + } + wait_for_completion(&msginfo->waitevent); + + /* 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); + + 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); + + 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); + ret = vmbus_post_msg(msg, + sizeof(struct vmbus_channel_gpadl_teardown)); + + if (ret) + goto post_msg_err; + + wait_for_completion(&info->waitevent); + +post_msg_err: + 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; +} + +static int vmbus_close_internal(struct vmbus_channel *channel) +{ + struct vmbus_channel_close_channel *msg; + int ret; + + channel->state = CHANNEL_OPEN_STATE; + channel->sc_creation_callback = NULL; + /* Stop callback and cancel the timer 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(); + } + + /* 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)); + + 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. + */ + return ret; + } + + /* 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. + */ + return ret; + } + } + + /* Cleanup the ring buffers for this channel */ + hv_ringbuffer_cleanup(&channel->outbound); + hv_ringbuffer_cleanup(&channel->inbound); + + free_pages((unsigned long)channel->ringbuffer_pages, + get_order(channel->ringbuffer_pagecount * PAGE_SIZE)); + + /* + * If the channel has been rescinded; process device removal. + */ + if (channel->rescind) + hv_process_channel_removal(channel, + channel->offermsg.child_relid); + 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->state != CHANNEL_OPENED_STATE) + continue; + vmbus_close_internal(cur_channel); + } + /* + * Now close the primary. + */ + vmbus_close_internal(channel); +} +EXPORT_SYMBOL_GPL(vmbus_close); + +int vmbus_sendpacket_ctl(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u64 requestid, + enum vmbus_packet_type type, u32 flags, bool kick_q) +{ + 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 ret; + bool signal = false; + + + /* 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); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + + /* + * Signalling the host is conditional on many factors: + * 1. The ring state changed from being empty to non-empty. + * This is tracked by the variable "signal". + * 2. The variable kick_q tracks if more data will be placed + * on the ring. We will not signal if more data is + * to be placed. + * + * If we cannot write to the ring-buffer; signal the host + * even if we may not have written anything. This is a rare + * enough condition that it should not matter. + */ + if (((ret == 0) && kick_q && signal) || (ret)) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL(vmbus_sendpacket_ctl); + +/** + * 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) +{ + return vmbus_sendpacket_ctl(channel, buffer, bufferlen, requestid, + type, flags, true); +} +EXPORT_SYMBOL(vmbus_sendpacket); + +/* + * vmbus_sendpacket_pagebuffer_ctl - 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_ctl(struct vmbus_channel *channel, + struct hv_page_buffer pagebuffers[], + u32 pagecount, void *buffer, u32 bufferlen, + u64 requestid, + u32 flags, + bool kick_q) +{ + int ret; + int i; + struct vmbus_channel_packet_page_buffer desc; + u32 descsize; + u32 packetlen; + u32 packetlen_aligned; + struct kvec bufferlist[3]; + u64 aligned_data = 0; + bool signal = false; + + 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 = flags; + desc.dataoffset8 = descsize >> 3; /* in 8-bytes grandularity */ + desc.length8 = (u16)(packetlen_aligned >> 3); + desc.transactionid = requestid; + 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); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + + /* + * Signalling the host is conditional on many factors: + * 1. The ring state changed from being empty to non-empty. + * This is tracked by the variable "signal". + * 2. The variable kick_q tracks if more data will be placed + * on the ring. We will not signal if more data is + * to be placed. + * + * If we cannot write to the ring-buffer; signal the host + * even if we may not have written anything. This is a rare + * enough condition that it should not matter. + */ + if (((ret == 0) && kick_q && signal) || (ret)) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_pagebuffer_ctl); + +/* + * vmbus_sendpacket_pagebuffer - Send a range of single-page buffer + * packets using a GPADL Direct packet type. + */ +int vmbus_sendpacket_pagebuffer(struct vmbus_channel *channel, + struct hv_page_buffer pagebuffers[], + u32 pagecount, void *buffer, u32 bufferlen, + u64 requestid) +{ + u32 flags = VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED; + return vmbus_sendpacket_pagebuffer_ctl(channel, pagebuffers, pagecount, + buffer, bufferlen, requestid, + flags, true); + +} +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) +{ + int ret; + u32 packetlen; + u32 packetlen_aligned; + struct kvec bufferlist[3]; + u64 aligned_data = 0; + bool signal = false; + + 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 grandularity */ + desc->length8 = (u16)(packetlen_aligned >> 3); + desc->transactionid = requestid; + 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); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + + if (ret == 0 && signal) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_mpb_desc); + +/* + * vmbus_sendpacket_multipagebuffer - Send a multi-page buffer packet + * using a GPADL Direct packet type. + */ +int vmbus_sendpacket_multipagebuffer(struct vmbus_channel *channel, + struct hv_multipage_buffer *multi_pagebuffer, + void *buffer, u32 bufferlen, u64 requestid) +{ + int ret; + struct vmbus_channel_packet_multipage_buffer desc; + u32 descsize; + u32 packetlen; + u32 packetlen_aligned; + struct kvec bufferlist[3]; + u64 aligned_data = 0; + bool signal = false; + u32 pfncount = NUM_PAGES_SPANNED(multi_pagebuffer->offset, + multi_pagebuffer->len); + + if (pfncount > MAX_MULTIPAGE_BUFFER_COUNT) + return -EINVAL; + + /* + * Adjust the size down since vmbus_channel_packet_multipage_buffer is + * the largest size we support + */ + descsize = sizeof(struct vmbus_channel_packet_multipage_buffer) - + ((MAX_MULTIPAGE_BUFFER_COUNT - pfncount) * + sizeof(u64)); + 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 grandularity */ + desc.length8 = (u16)(packetlen_aligned >> 3); + desc.transactionid = requestid; + desc.rangecount = 1; + + desc.range.len = multi_pagebuffer->len; + desc.range.offset = multi_pagebuffer->offset; + + memcpy(desc.range.pfn_array, multi_pagebuffer->pfn_array, + pfncount * sizeof(u64)); + + 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); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3, &signal); + + if (ret == 0 && signal) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_multipagebuffer); + +/** + * 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. + */ +int vmbus_recvpacket(struct vmbus_channel *channel, void *buffer, + u32 bufferlen, u32 *buffer_actual_len, u64 *requestid) +{ + struct vmpacket_descriptor desc; + u32 packetlen; + u32 userlen; + int ret; + bool signal = false; + + *buffer_actual_len = 0; + *requestid = 0; + + + ret = hv_ringbuffer_peek(&channel->inbound, &desc, + sizeof(struct vmpacket_descriptor)); + if (ret != 0) + return 0; + + packetlen = desc.len8 << 3; + userlen = packetlen - (desc.offset8 << 3); + + *buffer_actual_len = userlen; + + if (userlen > bufferlen) { + + pr_err("Buffer too small - got %d needs %d\n", + bufferlen, userlen); + return -ETOOSMALL; + } + + *requestid = desc.trans_id; + + /* Copy over the packet to the user buffer */ + ret = hv_ringbuffer_read(&channel->inbound, buffer, userlen, + (desc.offset8 << 3), &signal); + + if (signal) + vmbus_setevent(channel); + + return 0; +} +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) +{ + struct vmpacket_descriptor desc; + u32 packetlen; + int ret; + bool signal = false; + + *buffer_actual_len = 0; + *requestid = 0; + + + ret = hv_ringbuffer_peek(&channel->inbound, &desc, + sizeof(struct vmpacket_descriptor)); + if (ret != 0) + return 0; + + + packetlen = desc.len8 << 3; + + *buffer_actual_len = packetlen; + + if (packetlen > bufferlen) + return -ENOBUFS; + + *requestid = desc.trans_id; + + /* Copy over the entire packet to the user buffer */ + ret = hv_ringbuffer_read(&channel->inbound, buffer, packetlen, 0, + &signal); + + if (signal) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_recvpacket_raw); diff --git a/kernel/drivers/hv/channel_mgmt.c b/kernel/drivers/hv/channel_mgmt.c new file mode 100644 index 000000000..0eeb1b3bc --- /dev/null +++ b/kernel/drivers/hv/channel_mgmt.c @@ -0,0 +1,862 @@ +/* + * 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/list.h> +#include <linux/module.h> +#include <linux/completion.h> +#include <linux/hyperv.h> + +#include "hyperv_vmbus.h" + +/** + * 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. + * @negop is of type &struct icmsg_negotiate. + * Set up and fill in default negotiate response message. + * + * The fw_version specifies the framework version that + * we can support and srv_version specifies the service + * version we can support. + * + * Mainly used by Hyper-V drivers. + */ +bool vmbus_prep_negotiate_resp(struct icmsg_hdr *icmsghdrp, + struct icmsg_negotiate *negop, u8 *buf, + int fw_version, int srv_version) +{ + int icframe_major, icframe_minor; + int icmsg_major, icmsg_minor; + int fw_major, fw_minor; + int srv_major, srv_minor; + int i; + bool found_match = false; + + icmsghdrp->icmsgsize = 0x10; + fw_major = (fw_version >> 16); + fw_minor = (fw_version & 0xFFFF); + + srv_major = (srv_version >> 16); + srv_minor = (srv_version & 0xFFFF); + + 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 < negop->icframe_vercnt; i++) { + if ((negop->icversion_data[i].major == fw_major) && + (negop->icversion_data[i].minor == fw_minor)) { + icframe_major = negop->icversion_data[i].major; + icframe_minor = negop->icversion_data[i].minor; + found_match = true; + } + } + + if (!found_match) + goto fw_error; + + found_match = false; + + for (i = negop->icframe_vercnt; + (i < negop->icframe_vercnt + negop->icmsg_vercnt); i++) { + if ((negop->icversion_data[i].major == srv_major) && + (negop->icversion_data[i].minor == srv_minor)) { + icmsg_major = negop->icversion_data[i].major; + icmsg_minor = negop->icversion_data[i].minor; + found_match = true; + } + } + + /* + * 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; + } + + 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) +{ + static atomic_t chan_num = ATOMIC_INIT(0); + struct vmbus_channel *channel; + + channel = kzalloc(sizeof(*channel), GFP_ATOMIC); + if (!channel) + return NULL; + + channel->id = atomic_inc_return(&chan_num); + spin_lock_init(&channel->inbound_lock); + spin_lock_init(&channel->lock); + + INIT_LIST_HEAD(&channel->sc_list); + INIT_LIST_HEAD(&channel->percpu_list); + + return channel; +} + +/* + * free_channel - Release the resources used by the vmbus channel object + */ +static void free_channel(struct vmbus_channel *channel) +{ + kfree(channel); +} + +static void percpu_channel_enq(void *arg) +{ + struct vmbus_channel *channel = arg; + int cpu = smp_processor_id(); + + list_add_tail(&channel->percpu_list, &hv_context.percpu_list[cpu]); +} + +static void percpu_channel_deq(void *arg) +{ + struct vmbus_channel *channel = arg; + + list_del(&channel->percpu_list); +} + + +void hv_process_channel_removal(struct vmbus_channel *channel, u32 relid) +{ + struct vmbus_channel_relid_released msg; + unsigned long flags; + struct vmbus_channel *primary_channel; + + memset(&msg, 0, sizeof(struct vmbus_channel_relid_released)); + msg.child_relid = relid; + msg.header.msgtype = CHANNELMSG_RELID_RELEASED; + vmbus_post_msg(&msg, sizeof(struct vmbus_channel_relid_released)); + + if (channel == NULL) + return; + + 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) { + spin_lock_irqsave(&vmbus_connection.channel_lock, flags); + list_del(&channel->listentry); + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + } else { + primary_channel = channel->primary_channel; + spin_lock_irqsave(&primary_channel->lock, flags); + list_del(&channel->sc_list); + spin_unlock_irqrestore(&primary_channel->lock, flags); + } + free_channel(channel); +} + +void vmbus_free_channels(void) +{ + struct vmbus_channel *channel; + + list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) { + vmbus_device_unregister(channel->device_obj); + free_channel(channel); + } +} + +/* + * 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; + bool fnew = true; + bool enq = false; + unsigned long flags; + + /* Make sure this is a new offer */ + spin_lock_irqsave(&vmbus_connection.channel_lock, flags); + + 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); + enq = true; + } + + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + + if (enq) { + 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(); + } + } + if (!fnew) { + /* + * Check to see if this is a sub-channel. + */ + if (newchannel->offermsg.offer.sub_channel_index != 0) { + /* + * 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); + + 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(); + } + + newchannel->state = CHANNEL_OPEN_STATE; + channel->num_sc++; + if (channel->sc_creation_callback != NULL) + channel->sc_creation_callback(newchannel); + + return; + } + + goto err_free_chan; + } + + /* + * 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; + + /* + * Start the process of binding this offer to the driver + * We need to set the DeviceObject field before calling + * vmbus_child_dev_add() + */ + 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; + + /* + * Add the new device to the bus. This will kick off device-driver + * binding which eventually invokes the device driver's AddDevice() + * method. + */ + if (vmbus_device_register(newchannel->device_obj) != 0) { + pr_err("unable to add child device object (relid %d)\n", + newchannel->offermsg.child_relid); + kfree(newchannel->device_obj); + goto err_deq_chan; + } + return; + +err_deq_chan: + spin_lock_irqsave(&vmbus_connection.channel_lock, flags); + list_del(&newchannel->listentry); + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + + 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(); + } + +err_free_chan: + free_channel(newchannel); +} + +enum { + IDE = 0, + SCSI, + NIC, + MAX_PERF_CHN, +}; + +/* + * This is an array of device_ids (device types) that are performance critical. + * We attempt to distribute the interrupt load for these devices across + * all available CPUs. + */ +static const struct hv_vmbus_device_id hp_devs[] = { + /* IDE */ + { HV_IDE_GUID, }, + /* Storage - SCSI */ + { HV_SCSI_GUID, }, + /* Network */ + { HV_NIC_GUID, }, + /* NetworkDirect Guest RDMA */ + { HV_ND_GUID, }, +}; + + +/* + * We use this state to statically distribute the channel interrupt load. + */ +static u32 next_vp; + +/* + * Starting with Win8, we can statically distribute the incoming + * channel interrupt load by binding a channel to VCPU. We + * implement here a simple round robin scheme for distributing + * the interrupt load. + * We will bind channels that are not performance critical to cpu 0 and + * performance critical channels (IDE, SCSI and Network) will be uniformly + * distributed across all available CPUs. + */ +static void init_vp_index(struct vmbus_channel *channel, const uuid_le *type_guid) +{ + u32 cur_cpu; + int i; + bool perf_chn = false; + u32 max_cpus = num_online_cpus(); + + for (i = IDE; i < MAX_PERF_CHN; i++) { + if (!memcmp(type_guid->b, hp_devs[i].guid, + sizeof(uuid_le))) { + perf_chn = true; + break; + } + } + if ((vmbus_proto_version == VERSION_WS2008) || + (vmbus_proto_version == VERSION_WIN7) || (!perf_chn)) { + /* + * 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. + */ + channel->target_cpu = 0; + channel->target_vp = 0; + return; + } + cur_cpu = (++next_vp % max_cpus); + channel->target_cpu = cur_cpu; + channel->target_vp = hv_context.vp_index[cur_cpu]; +} + +/* + * 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; + + /* Allocate the channel object and save this offer. */ + newchannel = alloc_channel(); + if (!newchannel) { + pr_err("Unable to allocate channel object\n"); + return; + } + + /* + * By default we setup state to enable batched + * reading. A specific service can choose to + * disable this prior to opening the channel. + */ + newchannel->batched_reading = true; + + /* + * Setup state for signalling the host. + */ + newchannel->sig_event = (struct hv_input_signal_event *) + (ALIGN((unsigned long) + &newchannel->sig_buf, + HV_HYPERCALL_PARAM_ALIGN)); + + newchannel->sig_event->connectionid.asu32 = 0; + newchannel->sig_event->connectionid.u.id = VMBUS_EVENT_CONNECTION_ID; + newchannel->sig_event->flag_number = 0; + newchannel->sig_event->rsvdz = 0; + + if (vmbus_proto_version != VERSION_WS2008) { + newchannel->is_dedicated_interrupt = + (offer->is_dedicated_interrupt != 0); + newchannel->sig_event->connectionid.u.id = + offer->connection_id; + } + + init_vp_index(newchannel, &offer->offer.if_type); + + 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; + unsigned long flags; + struct device *dev; + + rescind = (struct vmbus_channel_rescind_offer *)hdr; + channel = relid2channel(rescind->child_relid); + + if (channel == NULL) { + hv_process_channel_removal(NULL, rescind->child_relid); + return; + } + + spin_lock_irqsave(&channel->lock, flags); + channel->rescind = true; + spin_unlock_irqrestore(&channel->lock, flags); + + if (channel->device_obj) { + /* + * 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 { + hv_process_channel_removal(channel, + channel->offermsg.child_relid); + } +} + +/* + * 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; + + /* + * 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; + + /* + * 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; + + /* + * 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; + 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 */ +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}, +}; + +/* + * 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; + int size; + + hdr = (struct vmbus_channel_message_header *)msg->u.payload; + size = msg->header.payload_size; + + if (hdr->msgtype >= CHANNELMSG_COUNT) { + pr_err("Received invalid channel message type %d size %d\n", + hdr->msgtype, size); + print_hex_dump_bytes("", DUMP_PREFIX_NONE, + (unsigned char *)msg->u.payload, size); + return; + } + + if (channel_message_table[hdr->msgtype].message_handler) + channel_message_table[hdr->msgtype].message_handler(hdr); + else + pr_err("Unhandled channel message type %d\n", hdr->msgtype); +} + +/* + * 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)); + 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_context.vp_index[get_cpu()]; + put_cpu(); + 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); diff --git a/kernel/drivers/hv/connection.c b/kernel/drivers/hv/connection.c new file mode 100644 index 000000000..b27220a42 --- /dev/null +++ b/kernel/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/hyperv.h> +#include "hyperv_vmbus.h" + + +struct vmbus_connection vmbus_connection = { + .conn_state = DISCONNECTED, + .next_gpadl_handle = ATOMIC_INIT(0xE1E10), +}; + +/* + * 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_WS2008): + default: + return VERSION_INVAL; + } +} + +static int vmbus_negotiate_version(struct vmbus_channel_msginfo *msginfo, + __u32 version) +{ + int ret = 0; + struct vmbus_channel_initiate_contact *msg; + unsigned long flags; + + init_completion(&msginfo->waitevent); + + msg = (struct vmbus_channel_initiate_contact *)msginfo->msg; + + msg->header.msgtype = CHANNELMSG_INITIATE_CONTACT; + msg->vmbus_version_requested = version; + msg->interrupt_page = virt_to_phys(vmbus_connection.int_page); + msg->monitor_page1 = virt_to_phys(vmbus_connection.monitor_pages[0]); + msg->monitor_page2 = virt_to_phys(vmbus_connection.monitor_pages[1]); + if (version == VERSION_WIN8_1) { + msg->target_vcpu = hv_context.vp_index[get_cpu()]; + put_cpu(); + } + + /* + * 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)); + 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; + } 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; + } + + INIT_LIST_HEAD(&vmbus_connection.chn_msg_list); + spin_lock_init(&vmbus_connection.channelmsg_lock); + + INIT_LIST_HEAD(&vmbus_connection.chn_list); + spin_lock_init(&vmbus_connection.channel_lock); + + /* + * 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("Hyper-V Host Build:%d-%d.%d-%d-%d.%d; Vmbus version:%d.%d\n", + host_info_eax, host_info_ebx >> 16, + host_info_ebx & 0xFFFF, host_info_ecx, + host_info_edx >> 24, host_info_edx & 0xFFFFFF, + 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) +{ + if (vmbus_connection.work_queue) { + drain_workqueue(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; +} + +/* + * Map the given relid to the corresponding channel based on the + * per-cpu list of channels that have been affinitized to this CPU. + * This will be used in the channel callback path as we can do this + * mapping in a lock-free fashion. + */ +static struct vmbus_channel *pcpu_relid2channel(u32 relid) +{ + struct vmbus_channel *channel; + struct vmbus_channel *found_channel = NULL; + int cpu = smp_processor_id(); + struct list_head *pcpu_head = &hv_context.percpu_list[cpu]; + + list_for_each_entry(channel, pcpu_head, percpu_list) { + if (channel->offermsg.child_relid == relid) { + found_channel = channel; + break; + } + } + + return found_channel; +} + +/* + * 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; + unsigned long flags; + struct list_head *cur, *tmp; + struct vmbus_channel *cur_sc; + + spin_lock_irqsave(&vmbus_connection.channel_lock, flags); + 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; + } + } + } + } + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + + return found_channel; +} + +/* + * process_chn_event - Process a channel event notification + */ +static void process_chn_event(u32 relid) +{ + struct vmbus_channel *channel; + void *arg; + bool read_state; + u32 bytes_to_read; + + /* + * Find the channel based on this relid and invokes the + * channel callback to process the event + */ + channel = pcpu_relid2channel(relid); + + if (!channel) + return; + + /* + * A channel once created is persistent even when there + * is no driver handling the device. An unloading driver + * sets the onchannel_callback to NULL on the same CPU + * as where this interrupt is handled (in an interrupt context). + * Thus, checking and invoking the driver specific callback takes + * care of orderly unloading of the driver. + */ + + if (channel->onchannel_callback != NULL) { + arg = channel->channel_callback_context; + read_state = channel->batched_reading; + /* + * This callback reads the messages sent by the host. + * We can 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. + */ + + do { + if (read_state) + hv_begin_read(&channel->inbound); + channel->onchannel_callback(arg); + if (read_state) + bytes_to_read = hv_end_read(&channel->inbound); + else + bytes_to_read = 0; + } while (read_state && (bytes_to_read != 0)); + } +} + +/* + * vmbus_on_event - Handler for events + */ +void vmbus_on_event(unsigned long data) +{ + u32 dword; + u32 maxdword; + int bit; + u32 relid; + u32 *recv_int_page = NULL; + void *page_addr; + int cpu = smp_processor_id(); + union hv_synic_event_flags *event; + + if ((vmbus_proto_version == VERSION_WS2008) || + (vmbus_proto_version == VERSION_WIN7)) { + maxdword = MAX_NUM_CHANNELS_SUPPORTED >> 5; + 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. + */ + maxdword = HV_EVENT_FLAGS_DWORD_COUNT; + page_addr = hv_context.synic_event_page[cpu]; + event = (union hv_synic_event_flags *)page_addr + + VMBUS_MESSAGE_SINT; + recv_int_page = event->flags32; + } + + + + /* Check events */ + if (!recv_int_page) + return; + for (dword = 0; dword < maxdword; dword++) { + if (!recv_int_page[dword]) + continue; + for (bit = 0; bit < 32; bit++) { + if (sync_test_and_clear_bit(bit, + (unsigned long *)&recv_int_page[dword])) { + relid = (dword << 5) + bit; + + if (relid == 0) + /* + * Special case - vmbus + * channel protocol msg + */ + continue; + + process_chn_event(relid); + } + } + } +} + +/* + * vmbus_post_msg - Send a msg on the vmbus's message connection + */ +int vmbus_post_msg(void *buffer, size_t buflen) +{ + union hv_connection_id conn_id; + int ret = 0; + int retries = 0; + u32 msec = 1; + + conn_id.asu32 = 0; + conn_id.u.id = VMBUS_MESSAGE_CONNECTION_ID; + + /* + * hv_post_message() can have transient failures because of + * insufficient resources. Retry the operation a couple of + * times before giving up. + */ + while (retries < 20) { + ret = hv_post_message(conn_id, 1, buffer, buflen); + + switch (ret) { + case HV_STATUS_INVALID_CONNECTION_ID: + /* + * We could get this if we send messages too + * frequently. + */ + ret = -EAGAIN; + break; + case HV_STATUS_INSUFFICIENT_MEMORY: + case HV_STATUS_INSUFFICIENT_BUFFERS: + ret = -ENOMEM; + break; + case HV_STATUS_SUCCESS: + return ret; + default: + pr_err("hv_post_msg() failed; error code:%d\n", ret); + return -EINVAL; + } + + retries++; + msleep(msec); + if (msec < 2048) + msec *= 2; + } + return ret; +} + +/* + * vmbus_set_event - Send an event notification to the parent + */ +int vmbus_set_event(struct vmbus_channel *channel) +{ + u32 child_relid = channel->offermsg.child_relid; + + if (!channel->is_dedicated_interrupt) { + /* Each u32 represents 32 channels */ + sync_set_bit(child_relid & 31, + (unsigned long *)vmbus_connection.send_int_page + + (child_relid >> 5)); + } + + return hv_signal_event(channel->sig_event); +} diff --git a/kernel/drivers/hv/hv.c b/kernel/drivers/hv/hv.c new file mode 100644 index 000000000..d3943bcee --- /dev/null +++ b/kernel/drivers/hv/hv.c @@ -0,0 +1,535 @@ +/* + * 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/interrupt.h> +#include <linux/clockchips.h> +#include <asm/hyperv.h> +#include <asm/mshyperv.h> +#include "hyperv_vmbus.h" + +/* The one and only */ +struct hv_context hv_context = { + .synic_initialized = false, + .hypercall_page = NULL, +}; + +#define HV_TIMER_FREQUENCY (10 * 1000 * 1000) /* 100ns period */ +#define HV_MAX_MAX_DELTA_TICKS 0xffffffff +#define HV_MIN_DELTA_TICKS 1 + +/* + * query_hypervisor_info - Get version info of the windows hypervisor + */ +unsigned int host_info_eax; +unsigned int host_info_ebx; +unsigned int host_info_ecx; +unsigned int host_info_edx; + +static int query_hypervisor_info(void) +{ + unsigned int eax; + unsigned int ebx; + unsigned int ecx; + unsigned int edx; + unsigned int max_leaf; + unsigned int op; + + /* + * Its assumed that this is called after confirming that Viridian + * is present. Query id and revision. + */ + eax = 0; + ebx = 0; + ecx = 0; + edx = 0; + op = HVCPUID_VENDOR_MAXFUNCTION; + cpuid(op, &eax, &ebx, &ecx, &edx); + + max_leaf = eax; + + if (max_leaf >= HVCPUID_VERSION) { + eax = 0; + ebx = 0; + ecx = 0; + edx = 0; + op = HVCPUID_VERSION; + cpuid(op, &eax, &ebx, &ecx, &edx); + host_info_eax = eax; + host_info_ebx = ebx; + host_info_ecx = ecx; + host_info_edx = edx; + } + return max_leaf; +} + +/* + * do_hypercall- Invoke the specified hypercall + */ +static u64 do_hypercall(u64 control, void *input, void *output) +{ +#ifdef CONFIG_X86_64 + u64 hv_status = 0; + u64 input_address = (input) ? virt_to_phys(input) : 0; + u64 output_address = (output) ? virt_to_phys(output) : 0; + void *hypercall_page = hv_context.hypercall_page; + + __asm__ __volatile__("mov %0, %%r8" : : "r" (output_address) : "r8"); + __asm__ __volatile__("call *%3" : "=a" (hv_status) : + "c" (control), "d" (input_address), + "m" (hypercall_page)); + + return hv_status; + +#else + + u32 control_hi = control >> 32; + u32 control_lo = control & 0xFFFFFFFF; + u32 hv_status_hi = 1; + u32 hv_status_lo = 1; + u64 input_address = (input) ? virt_to_phys(input) : 0; + u32 input_address_hi = input_address >> 32; + u32 input_address_lo = input_address & 0xFFFFFFFF; + u64 output_address = (output) ? virt_to_phys(output) : 0; + u32 output_address_hi = output_address >> 32; + u32 output_address_lo = output_address & 0xFFFFFFFF; + void *hypercall_page = hv_context.hypercall_page; + + __asm__ __volatile__ ("call *%8" : "=d"(hv_status_hi), + "=a"(hv_status_lo) : "d" (control_hi), + "a" (control_lo), "b" (input_address_hi), + "c" (input_address_lo), "D"(output_address_hi), + "S"(output_address_lo), "m" (hypercall_page)); + + return hv_status_lo | ((u64)hv_status_hi << 32); +#endif /* !x86_64 */ +} + +/* + * hv_init - Main initialization routine. + * + * This routine must be called before any other routines in here are called + */ +int hv_init(void) +{ + int max_leaf; + union hv_x64_msr_hypercall_contents hypercall_msr; + void *virtaddr = NULL; + + memset(hv_context.synic_event_page, 0, sizeof(void *) * NR_CPUS); + memset(hv_context.synic_message_page, 0, + sizeof(void *) * NR_CPUS); + memset(hv_context.post_msg_page, 0, + sizeof(void *) * NR_CPUS); + memset(hv_context.vp_index, 0, + sizeof(int) * NR_CPUS); + memset(hv_context.event_dpc, 0, + sizeof(void *) * NR_CPUS); + memset(hv_context.clk_evt, 0, + sizeof(void *) * NR_CPUS); + + max_leaf = query_hypervisor_info(); + + /* + * Write our OS ID. + */ + hv_context.guestid = generate_guest_id(0, LINUX_VERSION_CODE, 0); + wrmsrl(HV_X64_MSR_GUEST_OS_ID, hv_context.guestid); + + /* See if the hypercall page is already set */ + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + + virtaddr = __vmalloc(PAGE_SIZE, GFP_KERNEL, PAGE_KERNEL_EXEC); + + if (!virtaddr) + goto cleanup; + + hypercall_msr.enable = 1; + + hypercall_msr.guest_physical_address = vmalloc_to_pfn(virtaddr); + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + + /* Confirm that hypercall page did get setup. */ + hypercall_msr.as_uint64 = 0; + rdmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + + if (!hypercall_msr.enable) + goto cleanup; + + hv_context.hypercall_page = virtaddr; + + return 0; + +cleanup: + if (virtaddr) { + if (hypercall_msr.enable) { + hypercall_msr.as_uint64 = 0; + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + } + + vfree(virtaddr); + } + + return -ENOTSUPP; +} + +/* + * hv_cleanup - Cleanup routine. + * + * This routine is called normally during driver unloading or exiting. + */ +void hv_cleanup(void) +{ + union hv_x64_msr_hypercall_contents hypercall_msr; + + /* Reset our OS id */ + wrmsrl(HV_X64_MSR_GUEST_OS_ID, 0); + + if (hv_context.hypercall_page) { + hypercall_msr.as_uint64 = 0; + wrmsrl(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64); + vfree(hv_context.hypercall_page); + hv_context.hypercall_page = NULL; + } +} + +/* + * 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; + u16 status; + + if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT) + return -EMSGSIZE; + + aligned_msg = (struct hv_input_post_message *) + hv_context.post_msg_page[get_cpu()]; + + 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 = do_hypercall(HVCALL_POST_MESSAGE, aligned_msg, NULL) + & 0xFFFF; + + put_cpu(); + return status; +} + + +/* + * hv_signal_event - + * Signal an event on the specified connection using the hypervisor event IPC. + * + * This involves a hypercall. + */ +u16 hv_signal_event(void *con_id) +{ + u16 status; + + status = (do_hypercall(HVCALL_SIGNAL_EVENT, con_id, NULL) & 0xFFFF); + + return status; +} + +static int hv_ce_set_next_event(unsigned long delta, + struct clock_event_device *evt) +{ + cycle_t current_tick; + + WARN_ON(evt->mode != CLOCK_EVT_MODE_ONESHOT); + + rdmsrl(HV_X64_MSR_TIME_REF_COUNT, current_tick); + current_tick += delta; + wrmsrl(HV_X64_MSR_STIMER0_COUNT, current_tick); + return 0; +} + +static void hv_ce_setmode(enum clock_event_mode mode, + struct clock_event_device *evt) +{ + union hv_timer_config timer_cfg; + + switch (mode) { + case CLOCK_EVT_MODE_PERIODIC: + /* unsupported */ + break; + + case CLOCK_EVT_MODE_ONESHOT: + timer_cfg.enable = 1; + timer_cfg.auto_enable = 1; + timer_cfg.sintx = VMBUS_MESSAGE_SINT; + wrmsrl(HV_X64_MSR_STIMER0_CONFIG, timer_cfg.as_uint64); + break; + + case CLOCK_EVT_MODE_UNUSED: + case CLOCK_EVT_MODE_SHUTDOWN: + wrmsrl(HV_X64_MSR_STIMER0_COUNT, 0); + wrmsrl(HV_X64_MSR_STIMER0_CONFIG, 0); + break; + case CLOCK_EVT_MODE_RESUME: + break; + } +} + +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_mode = hv_ce_setmode; + dev->set_next_event = hv_ce_set_next_event; +} + + +int hv_synic_alloc(void) +{ + size_t size = sizeof(struct tasklet_struct); + size_t ced_size = sizeof(struct clock_event_device); + int cpu; + + for_each_online_cpu(cpu) { + hv_context.event_dpc[cpu] = kmalloc(size, GFP_ATOMIC); + if (hv_context.event_dpc[cpu] == NULL) { + pr_err("Unable to allocate event dpc\n"); + goto err; + } + tasklet_init(hv_context.event_dpc[cpu], vmbus_on_event, cpu); + + hv_context.clk_evt[cpu] = kzalloc(ced_size, GFP_ATOMIC); + if (hv_context.clk_evt[cpu] == NULL) { + pr_err("Unable to allocate clock event device\n"); + goto err; + } + hv_init_clockevent_device(hv_context.clk_evt[cpu], cpu); + + hv_context.synic_message_page[cpu] = + (void *)get_zeroed_page(GFP_ATOMIC); + + if (hv_context.synic_message_page[cpu] == NULL) { + pr_err("Unable to allocate SYNIC message page\n"); + goto err; + } + + hv_context.synic_event_page[cpu] = + (void *)get_zeroed_page(GFP_ATOMIC); + + if (hv_context.synic_event_page[cpu] == NULL) { + pr_err("Unable to allocate SYNIC event page\n"); + goto err; + } + + hv_context.post_msg_page[cpu] = + (void *)get_zeroed_page(GFP_ATOMIC); + + if (hv_context.post_msg_page[cpu] == NULL) { + pr_err("Unable to allocate post msg page\n"); + goto err; + } + } + + return 0; +err: + return -ENOMEM; +} + +static void hv_synic_free_cpu(int cpu) +{ + kfree(hv_context.event_dpc[cpu]); + kfree(hv_context.clk_evt[cpu]); + if (hv_context.synic_event_page[cpu]) + free_page((unsigned long)hv_context.synic_event_page[cpu]); + if (hv_context.synic_message_page[cpu]) + free_page((unsigned long)hv_context.synic_message_page[cpu]); + if (hv_context.post_msg_page[cpu]) + free_page((unsigned long)hv_context.post_msg_page[cpu]); +} + +void hv_synic_free(void) +{ + int cpu; + + for_each_online_cpu(cpu) + hv_synic_free_cpu(cpu); +} + +/* + * hv_synic_init - Initialize the Synthethic 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. + */ +void hv_synic_init(void *arg) +{ + u64 version; + union hv_synic_simp simp; + union hv_synic_siefp siefp; + union hv_synic_sint shared_sint; + union hv_synic_scontrol sctrl; + u64 vp_index; + + int cpu = smp_processor_id(); + + if (!hv_context.hypercall_page) + return; + + /* Check the version */ + rdmsrl(HV_X64_MSR_SVERSION, version); + + /* Setup the Synic's message page */ + rdmsrl(HV_X64_MSR_SIMP, simp.as_uint64); + simp.simp_enabled = 1; + simp.base_simp_gpa = virt_to_phys(hv_context.synic_message_page[cpu]) + >> PAGE_SHIFT; + + wrmsrl(HV_X64_MSR_SIMP, simp.as_uint64); + + /* Setup the Synic's event page */ + rdmsrl(HV_X64_MSR_SIEFP, siefp.as_uint64); + siefp.siefp_enabled = 1; + siefp.base_siefp_gpa = virt_to_phys(hv_context.synic_event_page[cpu]) + >> PAGE_SHIFT; + + wrmsrl(HV_X64_MSR_SIEFP, siefp.as_uint64); + + /* Setup the shared SINT. */ + rdmsrl(HV_X64_MSR_SINT0 + VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + shared_sint.as_uint64 = 0; + shared_sint.vector = HYPERVISOR_CALLBACK_VECTOR; + shared_sint.masked = false; + shared_sint.auto_eoi = true; + + wrmsrl(HV_X64_MSR_SINT0 + VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + /* Enable the global synic bit */ + rdmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64); + sctrl.enable = 1; + + wrmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64); + + hv_context.synic_initialized = true; + + /* + * Setup the mapping between Hyper-V's notion + * of cpuid and Linux' notion of cpuid. + * This array will be indexed using Linux cpuid. + */ + rdmsrl(HV_X64_MSR_VP_INDEX, vp_index); + hv_context.vp_index[cpu] = (u32)vp_index; + + INIT_LIST_HEAD(&hv_context.percpu_list[cpu]); + + /* + * Register the per-cpu clockevent source. + */ + if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE) + clockevents_config_and_register(hv_context.clk_evt[cpu], + HV_TIMER_FREQUENCY, + HV_MIN_DELTA_TICKS, + HV_MAX_MAX_DELTA_TICKS); + return; +} + +/* + * hv_synic_clockevents_cleanup - Cleanup clockevent devices + */ +void hv_synic_clockevents_cleanup(void) +{ + int cpu; + + if (!(ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE)) + return; + + for_each_online_cpu(cpu) + clockevents_unbind_device(hv_context.clk_evt[cpu], cpu); +} + +/* + * hv_synic_cleanup - Cleanup routine for hv_synic_init(). + */ +void hv_synic_cleanup(void *arg) +{ + union hv_synic_sint shared_sint; + union hv_synic_simp simp; + union hv_synic_siefp siefp; + union hv_synic_scontrol sctrl; + int cpu = smp_processor_id(); + + if (!hv_context.synic_initialized) + return; + + /* Turn off clockevent device */ + if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE) + hv_ce_setmode(CLOCK_EVT_MODE_SHUTDOWN, + hv_context.clk_evt[cpu]); + + rdmsrl(HV_X64_MSR_SINT0 + VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + shared_sint.masked = 1; + + /* Need to correctly cleanup in the case of SMP!!! */ + /* Disable the interrupt */ + wrmsrl(HV_X64_MSR_SINT0 + VMBUS_MESSAGE_SINT, shared_sint.as_uint64); + + rdmsrl(HV_X64_MSR_SIMP, simp.as_uint64); + simp.simp_enabled = 0; + simp.base_simp_gpa = 0; + + wrmsrl(HV_X64_MSR_SIMP, simp.as_uint64); + + rdmsrl(HV_X64_MSR_SIEFP, siefp.as_uint64); + siefp.siefp_enabled = 0; + siefp.base_siefp_gpa = 0; + + wrmsrl(HV_X64_MSR_SIEFP, siefp.as_uint64); + + /* Disable the global synic bit */ + rdmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64); + sctrl.enable = 0; + wrmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64); + + hv_synic_free_cpu(cpu); +} diff --git a/kernel/drivers/hv/hv_balloon.c b/kernel/drivers/hv/hv_balloon.c new file mode 100644 index 000000000..cb5b7dc97 --- /dev/null +++ b/kernel/drivers/hv/hv_balloon.c @@ -0,0 +1,1616 @@ +/* + * 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> + +/* + * 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_WIN7 = DYNMEM_PROTOCOL_VERSION_1, + DYNMEM_PROTOCOL_VERSION_WIN8 = DYNMEM_PROTOCOL_VERSION_2, + + DYNMEM_PROTOCOL_VERSION_CURRENT = DYNMEM_PROTOCOL_VERSION_WIN8 +}; + + + +/* + * 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_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_end_pfn; + unsigned long ha_end_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; + + struct mutex ha_region_mutex; + + /* + * 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; +}; + +static struct hv_dynmem_device dm_device; + +static void post_status(struct hv_dynmem_device *dm); + +#ifdef CONFIG_MEMORY_HOTPLUG +static int hv_memory_notifier(struct notifier_block *nb, unsigned long val, + void *v) +{ + struct memory_notify *mem = (struct memory_notify *)v; + + switch (val) { + case MEM_GOING_ONLINE: + mutex_lock(&dm_device.ha_region_mutex); + break; + + case MEM_ONLINE: + dm_device.num_pages_onlined += mem->nr_pages; + case MEM_CANCEL_ONLINE: + mutex_unlock(&dm_device.ha_region_mutex); + if (dm_device.ha_waiting) { + dm_device.ha_waiting = false; + complete(&dm_device.ol_waitevent); + } + break; + + case MEM_OFFLINE: + mutex_lock(&dm_device.ha_region_mutex); + dm_device.num_pages_onlined -= mem->nr_pages; + mutex_unlock(&dm_device.ha_region_mutex); + break; + 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 +}; + + +static void hv_bring_pgs_online(unsigned long start_pfn, unsigned long size) +{ + int i; + + for (i = 0; i < size; i++) { + struct page *pg; + pg = pfn_to_page(start_pfn + i); + __online_page_set_limits(pg); + __online_page_increment_counters(pg); + __online_page_free(pg); + } +} + +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; + + for (i = 0; i < (size/HA_CHUNK); i++) { + start_pfn = start + (i * HA_CHUNK); + 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; + + init_completion(&dm_device.ol_waitevent); + dm_device.ha_waiting = true; + + mutex_unlock(&dm_device.ha_region_mutex); + 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_info("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; + } + has->ha_end_pfn -= HA_CHUNK; + has->covered_end_pfn -= processed_pfn; + mutex_lock(&dm_device.ha_region_mutex); + break; + } + + /* + * Wait for the memory block to be onlined. + * 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. + */ + wait_for_completion_timeout(&dm_device.ol_waitevent, 5*HZ); + mutex_lock(&dm_device.ha_region_mutex); + post_status(&dm_device); + } + + return; +} + +static void hv_online_page(struct page *pg) +{ + struct list_head *cur; + struct hv_hotadd_state *has; + unsigned long cur_start_pgp; + unsigned long cur_end_pgp; + + list_for_each(cur, &dm_device.ha_region_list) { + has = list_entry(cur, struct hv_hotadd_state, list); + cur_start_pgp = (unsigned long)pfn_to_page(has->start_pfn); + cur_end_pgp = (unsigned long)pfn_to_page(has->covered_end_pfn); + + if (((unsigned long)pg >= cur_start_pgp) && + ((unsigned long)pg < cur_end_pgp)) { + /* + * This frame is currently backed; online the + * page. + */ + __online_page_set_limits(pg); + __online_page_increment_counters(pg); + __online_page_free(pg); + } + } +} + +static bool pfn_covered(unsigned long start_pfn, unsigned long pfn_cnt) +{ + struct list_head *cur; + struct hv_hotadd_state *has; + unsigned long residual, new_inc; + + if (list_empty(&dm_device.ha_region_list)) + return false; + + list_for_each(cur, &dm_device.ha_region_list) { + has = list_entry(cur, struct hv_hotadd_state, list); + + /* + * If the pfn range we are dealing with is not in the current + * "hot add block", move on. + */ + if ((start_pfn >= has->end_pfn)) + continue; + /* + * 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; + } + + /* + * If the current start pfn is not where the covered_end + * is, update it. + */ + + if (has->covered_end_pfn != start_pfn) + has->covered_end_pfn = start_pfn; + + return true; + + } + + return false; +} + +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 list_head *cur; + struct hv_hotadd_state *has; + unsigned long pgs_ol = 0; + unsigned long old_covered_state; + + if (list_empty(&dm_device.ha_region_list)) + return 0; + + list_for_each(cur, &dm_device.ha_region_list) { + has = list_entry(cur, struct hv_hotadd_state, list); + + /* + * If the pfn range we are dealing with is not in the current + * "hot add block", move on. + */ + if ((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; + + /* + * Check if the corresponding memory block is already + * online by checking its last previously backed page. + * In case it is we need to bring rest (which was not + * backed previously) online too. + */ + if (start_pfn > has->start_pfn && + !PageReserved(pfn_to_page(start_pfn - 1))) + hv_bring_pgs_online(start_pfn, pgs_ol); + + has->covered_end_pfn += pgs_ol; + pfn_cnt -= 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; + } + hv_mem_hot_add(has->ha_end_pfn, size, pfn_cnt, has); + } + /* + * If we managed to online any pages that were given to us, + * we declare success. + */ + return has->covered_end_pfn - old_covered_state; + + } + + return 0; +} + +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; + + if (pfn_cnt == 0) + return 0; + + if (!dm_device.host_specified_ha_region) + if (pfn_covered(pg_start, pfn_cnt)) + 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); + + list_add_tail(&ha_region->list, &dm_device.ha_region_list); + ha_region->start_pfn = rg_start; + ha_region->ha_end_pfn = rg_start; + ha_region->covered_end_pfn = pg_start; + ha_region->end_pfn = rg_start + rg_size; + } + +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 + mutex_lock(&dm_device.ha_region_mutex); + 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; + mutex_unlock(&dm_device.ha_region_mutex); +#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_info("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: + pr_info("Received INFO_TYPE_MAX_PAGE_CNT\n"); + pr_info("Data Size is %d\n", info_hdr->data_size); + break; + default: + pr_info("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; + struct sysinfo val; + 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; + + si_meminfo(&val); + 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 = val.freeram; + 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(); + + /* + * 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; + + if (num_pages < alloc_unit) + return 0; + + for (i = 0; (i * alloc_unit) < num_pages; 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 num_pages; +} + + + +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; + struct sysinfo val; + unsigned long floor; + + /* The host balloons pages in 2M granularity. */ + WARN_ON_ONCE(num_pages % PAGES_IN_2M != 0); + + /* + * We will attempt 2M allocations. However, if we fail to + * allocate 2M chunks, we will go back to 4k allocations. + */ + alloc_unit = 512; + + si_meminfo(&val); + floor = compute_balloon_floor(); + + /* Refuse to balloon below the floor, keep the 2M granularity. */ + if (val.freeram < num_pages || val.freeram - num_pages < floor) { + num_pages = val.freeram > floor ? (val.freeram - floor) : 0; + num_pages -= num_pages % PAGES_IN_2M; + } + + 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) { + 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_info("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; + + for (i = 0; i < range_count; i++) { + free_balloon_pages(dm, &range_array[i]); + complete(&dm_device.config_event); + } + + 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; + + dm->next_version = 0; + 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_WIN7; + 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_info("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. + */ + 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_err("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; + + do_hot_add = hot_add; + + /* + * 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_WIN7; + init_completion(&dm_device.host_event); + init_completion(&dm_device.config_event); + INIT_LIST_HEAD(&dm_device.ha_region_list); + mutex_init(&dm_device.ha_region_mutex); + 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_WIN8; + version_req.is_last_attempt = 0; + + 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; + } + /* + * 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; + + 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 list_head *cur, *tmp; + struct hv_hotadd_state *has; + + 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 + list_for_each_safe(cur, tmp, &dm->ha_region_list) { + has = list_entry(cur, struct hv_hotadd_state, list); + list_del(&has->list); + kfree(has); + } + + 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, +}; + +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/kernel/drivers/hv/hv_fcopy.c b/kernel/drivers/hv/hv_fcopy.c new file mode 100644 index 000000000..cd453e4b2 --- /dev/null +++ b/kernel/drivers/hv/hv_fcopy.c @@ -0,0 +1,441 @@ +/* + * 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/semaphore.h> +#include <linux/fs.h> +#include <linux/nls.h> +#include <linux/workqueue.h> +#include <linux/cdev.h> +#include <linux/hyperv.h> +#include <linux/sched.h> +#include <linux/uaccess.h> +#include <linux/miscdevice.h> + +#include "hyperv_vmbus.h" + +#define WIN8_SRV_MAJOR 1 +#define WIN8_SRV_MINOR 1 +#define WIN8_SRV_VERSION (WIN8_SRV_MAJOR << 16 | WIN8_SRV_MINOR) + +/* + * 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. + * + * The transaction "active" state is set when we receive a request from the + * host and we cleanup this state when the transaction is completed - when we + * respond to the host with our response. When the transaction active state is + * set, we defer handling incoming packets. + */ + +static struct { + bool active; /* transaction status - active or not */ + int recv_len; /* number of bytes received. */ + struct hv_fcopy_hdr *fcopy_msg; /* current message */ + struct hv_start_fcopy message; /* sent to daemon */ + struct vmbus_channel *recv_channel; /* chn we got the request */ + u64 recv_req_id; /* request ID. */ + void *fcopy_context; /* for the channel callback */ + struct semaphore read_sema; +} fcopy_transaction; + +static bool opened; /* currently device opened */ + +/* + * Before we can accept copy messages from the host, we need + * to handshake with the user level daemon. This state tracks + * if we are in the handshake phase. + */ +static bool in_hand_shake = true; +static void fcopy_send_data(void); +static void fcopy_respond_to_host(int error); +static void fcopy_work_func(struct work_struct *dummy); +static DECLARE_DELAYED_WORK(fcopy_work, fcopy_work_func); +static u8 *recv_buffer; + +static void fcopy_work_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); + + /* In the case the user-space daemon crashes, hangs or is killed, we + * need to down the semaphore, otherwise, after the daemon starts next + * time, the obsolete data in fcopy_transaction.message or + * fcopy_transaction.fcopy_msg will be used immediately. + * + * NOTE: fcopy_read() happens to get the semaphore (very rare)? We're + * still OK, because we've reported the failure to the host. + */ + if (down_trylock(&fcopy_transaction.read_sema)) + ; + +} + +static int fcopy_handle_handshake(u32 version) +{ + switch (version) { + case FCOPY_CURRENT_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_info("FCP: user-mode registering done. Daemon version: %d\n", + version); + fcopy_transaction.active = false; + if (fcopy_transaction.fcopy_context) + hv_fcopy_onchannelcallback(fcopy_transaction.fcopy_context); + in_hand_shake = false; + return 0; +} + +static void fcopy_send_data(void) +{ + struct hv_start_fcopy *smsg_out = &fcopy_transaction.message; + int operation = fcopy_transaction.fcopy_msg->operation; + struct hv_start_fcopy *smsg_in; + + /* + * 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: + memset(smsg_out, 0, sizeof(struct hv_start_fcopy)); + 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; + break; + + default: + break; + } + up(&fcopy_transaction.read_sema); + return; +} + +/* + * 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; + + fcopy_transaction.active = false; + + 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; + struct icmsg_negotiate *negop = NULL; + int util_fw_version; + int fcopy_srv_version; + + if (fcopy_transaction.active) { + /* + * We will defer processing this callback once + * the current transaction is complete. + */ + fcopy_transaction.fcopy_context = context; + 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) { + util_fw_version = UTIL_FW_VERSION; + fcopy_srv_version = WIN8_SRV_VERSION; + vmbus_prep_negotiate_resp(icmsghdr, negop, recv_buffer, + util_fw_version, fcopy_srv_version); + } 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.active = true; + fcopy_transaction.recv_len = recvlen; + fcopy_transaction.recv_channel = channel; + fcopy_transaction.recv_req_id = requestid; + fcopy_transaction.fcopy_msg = fcopy_msg; + + /* + * Send the information to the user-level daemon. + */ + schedule_delayed_work(&fcopy_work, 5*HZ); + fcopy_send_data(); + return; + } + icmsghdr->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + vmbus_sendpacket(channel, recv_buffer, recvlen, requestid, + VM_PKT_DATA_INBAND, 0); +} + +/* + * Create a char device that can support read/write for passing + * the payload. + */ + +static ssize_t fcopy_read(struct file *file, char __user *buf, + size_t count, loff_t *ppos) +{ + void *src; + size_t copy_size; + int operation; + + /* + * Wait until there is something to be read. + */ + if (down_interruptible(&fcopy_transaction.read_sema)) + return -EINTR; + + /* + * The channel may be rescinded and in this case, we will wakeup the + * the thread blocked on the semaphore and we will use the opened + * state to correctly handle this case. + */ + if (!opened) + return -ENODEV; + + operation = fcopy_transaction.fcopy_msg->operation; + + if (operation == START_FILE_COPY) { + src = &fcopy_transaction.message; + copy_size = sizeof(struct hv_start_fcopy); + if (count < copy_size) + return 0; + } else { + src = fcopy_transaction.fcopy_msg; + copy_size = sizeof(struct hv_do_fcopy); + if (count < copy_size) + return 0; + } + if (copy_to_user(buf, src, copy_size)) + return -EFAULT; + + return copy_size; +} + +static ssize_t fcopy_write(struct file *file, const char __user *buf, + size_t count, loff_t *ppos) +{ + int response = 0; + + if (count != sizeof(int)) + return -EINVAL; + + if (copy_from_user(&response, buf, sizeof(int))) + return -EFAULT; + + if (in_hand_shake) { + if (fcopy_handle_handshake(response)) + return -EINVAL; + return sizeof(int); + } + + /* + * Complete the transaction by forwarding the result + * to the host. But first, cancel the timeout. + */ + if (cancel_delayed_work_sync(&fcopy_work)) + fcopy_respond_to_host(response); + + return sizeof(int); +} + +static int fcopy_open(struct inode *inode, struct file *f) +{ + /* + * The user level daemon that will open this device is + * really an extension of this driver. We can have only + * active open at a time. + */ + if (opened) + return -EBUSY; + + /* + * The daemon is alive; setup the state. + */ + opened = true; + return 0; +} + +/* XXX: there are still some tricky corner cases, e.g., + * 1) In a SMP guest, when fcopy_release() runs between + * schedule_delayed_work() and fcopy_send_data(), there is + * still a chance an obsolete message will be queued. + * + * 2) When the fcopy daemon is running, if we unload the driver, + * we'll notice a kernel oops when we kill the daemon later. + */ +static int fcopy_release(struct inode *inode, struct file *f) +{ + /* + * The daemon has exited; reset the state. + */ + in_hand_shake = true; + opened = false; + + if (cancel_delayed_work_sync(&fcopy_work)) { + /* We haven't up()-ed the semaphore(very rare)? */ + if (down_trylock(&fcopy_transaction.read_sema)) + ; + fcopy_respond_to_host(HV_E_FAIL); + } + return 0; +} + + +static const struct file_operations fcopy_fops = { + .read = fcopy_read, + .write = fcopy_write, + .release = fcopy_release, + .open = fcopy_open, +}; + +static struct miscdevice fcopy_misc = { + .minor = MISC_DYNAMIC_MINOR, + .name = "vmbus/hv_fcopy", + .fops = &fcopy_fops, +}; + +static int fcopy_dev_init(void) +{ + return misc_register(&fcopy_misc); +} + +static void fcopy_dev_deinit(void) +{ + + /* + * The device is going away - perhaps because the + * host has rescinded the channel. Setup state so that + * user level daemon can gracefully exit if it is blocked + * on the read semaphore. + */ + opened = false; + /* + * Signal the semaphore as the device is + * going away. + */ + up(&fcopy_transaction.read_sema); + misc_deregister(&fcopy_misc); +} + +int hv_fcopy_init(struct hv_util_service *srv) +{ + recv_buffer = srv->recv_buffer; + + /* + * 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.active = true; + sema_init(&fcopy_transaction.read_sema, 0); + + return fcopy_dev_init(); +} + +void hv_fcopy_deinit(void) +{ + cancel_delayed_work_sync(&fcopy_work); + fcopy_dev_deinit(); +} diff --git a/kernel/drivers/hv/hv_kvp.c b/kernel/drivers/hv/hv_kvp.c new file mode 100644 index 000000000..beb8105c0 --- /dev/null +++ b/kernel/drivers/hv/hv_kvp.c @@ -0,0 +1,722 @@ +/* + * 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> + + +/* + * 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) + +/* + * Global state maintained for transaction that is being processed. + * Note that only one transaction can be active at any point in time. + * + * This state is set when we receive a request from the host; we + * cleanup this state when the transaction is completed - when we respond + * to the host with the key value. + */ + +static struct { + bool active; /* transaction status - active or not */ + 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. */ + void *kvp_context; /* for the channel callback */ +} kvp_transaction; + +/* + * Before we can accept KVP messages from the host, we need + * to handshake with the user level daemon. This state tracks + * if we are in the handshake phase. + */ +static bool in_hand_shake = true; + +/* + * 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_work_func(struct work_struct *dummy); +static void kvp_register(int); + +static DECLARE_DELAYED_WORK(kvp_work, kvp_work_func); +static DECLARE_WORK(kvp_sendkey_work, kvp_send_key); + +static struct cb_id kvp_id = { CN_KVP_IDX, CN_KVP_VAL }; +static const char kvp_name[] = "kvp_kernel_module"; +static u8 *recv_buffer; +/* + * 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_register(int reg_value) +{ + + struct cn_msg *msg; + struct hv_kvp_msg *kvp_msg; + char *version; + + msg = kzalloc(sizeof(*msg) + sizeof(struct hv_kvp_msg), GFP_ATOMIC); + + if (msg) { + kvp_msg = (struct hv_kvp_msg *)msg->data; + version = kvp_msg->body.kvp_register.version; + msg->id.idx = CN_KVP_IDX; + msg->id.val = CN_KVP_VAL; + + kvp_msg->kvp_hdr.operation = reg_value; + strcpy(version, HV_DRV_VERSION); + msg->len = sizeof(struct hv_kvp_msg); + cn_netlink_send(msg, 0, 0, GFP_ATOMIC); + kfree(msg); + } +} +static void +kvp_work_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); +} + +static void poll_channel(struct vmbus_channel *channel) +{ + if (channel->target_cpu != smp_processor_id()) + smp_call_function_single(channel->target_cpu, + hv_kvp_onchannelcallback, + channel, true); + else + hv_kvp_onchannelcallback(channel); +} + + +static int kvp_handle_handshake(struct hv_kvp_msg *msg) +{ + int ret = 1; + + 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); + ret = 0; + } + + if (ret) { + /* + * We have a compatible daemon; complete the handshake. + */ + pr_info("KVP: user-mode registering done.\n"); + kvp_register(dm_reg_value); + kvp_transaction.active = false; + if (kvp_transaction.kvp_context) + poll_channel(kvp_transaction.kvp_context); + } + return ret; +} + + +/* + * Callback when data is received from user mode. + */ + +static void +kvp_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp) +{ + struct hv_kvp_msg *message; + struct hv_kvp_msg_enumerate *data; + int error = 0; + + message = (struct hv_kvp_msg *)msg->data; + + /* + * If we are negotiating the version information + * with the daemon; handle that first. + */ + + if (in_hand_shake) { + if (kvp_handle_handshake(message)) + in_hand_shake = false; + return; + } + + /* + * 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_work)) + kvp_respond_to_host(message, error); +} + + +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_IP_ADDR_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; + + default: + 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 cn_msg *msg; + 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; + + msg = kzalloc(sizeof(*msg) + sizeof(struct hv_kvp_msg) , GFP_ATOMIC); + if (!msg) + return; + + msg->id.idx = CN_KVP_IDX; + msg->id.val = CN_KVP_VAL; + + message = (struct hv_kvp_msg *)msg->data; + 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 gaurantee + * 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: + 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; + + } + case KVP_OP_GET: + 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_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; + } + + msg->len = sizeof(struct hv_kvp_msg); + rc = cn_netlink_send(msg, 0, 0, GFP_ATOMIC); + if (rc) { + pr_debug("KVP: failed to communicate to the daemon: %d\n", rc); + if (cancel_delayed_work_sync(&kvp_work)) + kvp_respond_to_host(message, HV_E_FAIL); + } + + kfree(msg); + + return; +} + +/* + * 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; + + /* + * If a transaction is not active; log and return. + */ + + if (!kvp_transaction.active) { + /* + * This is a spurious call! + */ + pr_warn("KVP: Transaction not active\n"); + return; + } + /* + * 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; + + kvp_transaction.active = false; + + 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 timedout; + * 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); + poll_channel(channel); +} + +/* + * 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 retrive the corresponding value. Consequently, we cannot + * respond to the host in the conext 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; + struct icmsg_negotiate *negop = NULL; + int util_fw_version; + int kvp_srv_version; + + if (kvp_transaction.active) { + /* + * We will defer processing this callback once + * the current transaction is complete. + */ + kvp_transaction.kvp_context = context; + 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) { + /* + * Based on the host, select appropriate + * framework and service versions we will + * negotiate. + */ + switch (vmbus_proto_version) { + case (VERSION_WS2008): + util_fw_version = UTIL_WS2K8_FW_VERSION; + kvp_srv_version = WS2008_SRV_VERSION; + break; + case (VERSION_WIN7): + util_fw_version = UTIL_FW_VERSION; + kvp_srv_version = WIN7_SRV_VERSION; + break; + default: + util_fw_version = UTIL_FW_VERSION; + kvp_srv_version = WIN8_SRV_VERSION; + } + vmbus_prep_negotiate_resp(icmsghdrp, negop, + recv_buffer, util_fw_version, + kvp_srv_version); + + } 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_channel = channel; + kvp_transaction.recv_req_id = requestid; + kvp_transaction.active = true; + kvp_transaction.kvp_msg = kvp_msg; + + /* + * 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_work, 5*HZ); + + return; + + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } + +} + +int +hv_kvp_init(struct hv_util_service *srv) +{ + int err; + + err = cn_add_callback(&kvp_id, kvp_name, kvp_cn_callback); + if (err) + return err; + recv_buffer = srv->recv_buffer; + + /* + * 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.active = true; + + return 0; +} + +void hv_kvp_deinit(void) +{ + cn_del_callback(&kvp_id); + cancel_delayed_work_sync(&kvp_work); + cancel_work_sync(&kvp_sendkey_work); +} diff --git a/kernel/drivers/hv/hv_snapshot.c b/kernel/drivers/hv/hv_snapshot.c new file mode 100644 index 000000000..9d5e0d1ef --- /dev/null +++ b/kernel/drivers/hv/hv_snapshot.c @@ -0,0 +1,303 @@ +/* + * 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> + +#define VSS_MAJOR 5 +#define VSS_MINOR 0 +#define VSS_VERSION (VSS_MAJOR << 16 | VSS_MINOR) + +#define VSS_USERSPACE_TIMEOUT (msecs_to_jiffies(10 * 1000)) + +/* + * Global state maintained for transaction that is being processed. + * Note that only one transaction can be active at any point in time. + * + * This state is set when we receive a request from the host; we + * cleanup this state when the transaction is completed - when we respond + * to the host with the key value. + */ + +static struct { + bool active; /* transaction status - active or not */ + 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); + +static struct cb_id vss_id = { CN_VSS_IDX, CN_VSS_VAL }; +static const char vss_name[] = "vss_kernel_module"; +static __u8 *recv_buffer; + +static void vss_send_op(struct work_struct *dummy); +static void vss_timeout_func(struct work_struct *dummy); + +static DECLARE_DELAYED_WORK(vss_timeout_work, vss_timeout_func); +static DECLARE_WORK(vss_send_op_work, vss_send_op); + +/* + * 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); +} + +static void +vss_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp) +{ + struct hv_vss_msg *vss_msg; + + vss_msg = (struct hv_vss_msg *)msg->data; + + if (vss_msg->vss_hdr.operation == VSS_OP_REGISTER) { + pr_info("VSS daemon registered\n"); + vss_transaction.active = false; + if (vss_transaction.recv_channel != NULL) + hv_vss_onchannelcallback(vss_transaction.recv_channel); + return; + + } + if (cancel_delayed_work_sync(&vss_timeout_work)) + vss_respond_to_host(vss_msg->error); +} + + +static void vss_send_op(struct work_struct *dummy) +{ + int op = vss_transaction.msg->vss_hdr.operation; + int rc; + struct cn_msg *msg; + struct hv_vss_msg *vss_msg; + + msg = kzalloc(sizeof(*msg) + sizeof(*vss_msg), GFP_ATOMIC); + if (!msg) + return; + + vss_msg = (struct hv_vss_msg *)msg->data; + + msg->id.idx = CN_VSS_IDX; + msg->id.val = CN_VSS_VAL; + + vss_msg->vss_hdr.operation = op; + msg->len = sizeof(struct hv_vss_msg); + + rc = cn_netlink_send(msg, 0, 0, GFP_ATOMIC); + 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); + } + kfree(msg); + + return; +} + +/* + * 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; + + /* + * If a transaction is not active; log and return. + */ + + if (!vss_transaction.active) { + /* + * This is a spurious call! + */ + pr_warn("VSS: Transaction not active\n"); + return; + } + /* + * 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; + vss_transaction.active = false; + + 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; + + + struct icmsg_hdr *icmsghdrp; + struct icmsg_negotiate *negop = NULL; + + if (vss_transaction.active) { + /* + * We will defer processing this callback once + * the current transaction is complete. + */ + vss_transaction.recv_channel = channel; + 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) { + vmbus_prep_negotiate_resp(icmsghdrp, negop, + recv_buffer, UTIL_FW_VERSION, + VSS_VERSION); + } 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_channel = channel; + vss_transaction.recv_req_id = requestid; + vss_transaction.active = true; + vss_transaction.msg = (struct hv_vss_msg *)vss_msg; + + switch (vss_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_FREEZE: + case VSS_OP_THAW: + schedule_work(&vss_send_op_work); + schedule_delayed_work(&vss_timeout_work, + VSS_USERSPACE_TIMEOUT); + return; + + case VSS_OP_HOT_BACKUP: + vss_msg->vss_cf.flags = + VSS_HBU_NO_AUTO_RECOVERY; + vss_respond_to_host(0); + return; + + case VSS_OP_GET_DM_INFO: + vss_msg->dm_info.flags = 0; + vss_respond_to_host(0); + return; + + default: + vss_respond_to_host(0); + return; + + } + + } + + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION + | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, + recvlen, requestid, + VM_PKT_DATA_INBAND, 0); + } + +} + +int +hv_vss_init(struct hv_util_service *srv) +{ + int err; + + err = cn_add_callback(&vss_id, vss_name, vss_cn_callback); + if (err) + return err; + recv_buffer = srv->recv_buffer; + + /* + * 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.active = true; + return 0; +} + +void hv_vss_deinit(void) +{ + cn_del_callback(&vss_id); + cancel_delayed_work_sync(&vss_timeout_work); + cancel_work_sync(&vss_send_op_work); +} diff --git a/kernel/drivers/hv/hv_util.c b/kernel/drivers/hv/hv_util.c new file mode 100644 index 000000000..7994ec2e4 --- /dev/null +++ b/kernel/drivers/hv/hv_util.c @@ -0,0 +1,447 @@ +/* + * 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 "hyperv_vmbus.h" + +#define SD_MAJOR 3 +#define SD_MINOR 0 +#define SD_VERSION (SD_MAJOR << 16 | SD_MINOR) + +#define SD_WS2008_MAJOR 1 +#define SD_WS2008_VERSION (SD_WS2008_MAJOR << 16 | SD_MINOR) + +#define TS_MAJOR 3 +#define TS_MINOR 0 +#define TS_VERSION (TS_MAJOR << 16 | TS_MINOR) + +#define TS_WS2008_MAJOR 1 +#define TS_WS2008_VERSION (TS_WS2008_MAJOR << 16 | TS_MINOR) + +#define HB_MAJOR 3 +#define HB_MINOR 0 +#define HB_VERSION (HB_MAJOR << 16 | HB_MINOR) + +#define HB_WS2008_MAJOR 1 +#define HB_WS2008_VERSION (HB_WS2008_MAJOR << 16 | HB_MINOR) + +static int sd_srv_version; +static int ts_srv_version; +static int hb_srv_version; +static int util_fw_version; + +static void shutdown_onchannelcallback(void *context); +static struct hv_util_service util_shutdown = { + .util_cb = shutdown_onchannelcallback, +}; + +static void timesync_onchannelcallback(void *context); +static struct hv_util_service util_timesynch = { + .util_cb = timesync_onchannelcallback, +}; + +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; + struct icmsg_negotiate *negop = NULL; + + 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) { + vmbus_prep_negotiate_resp(icmsghdrp, negop, + shut_txf_buf, util_fw_version, + sd_srv_version); + } 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 guest time to host UTC time. + */ +static inline void do_adj_guesttime(u64 hosttime) +{ + s64 host_tns; + struct timespec host_ts; + + host_tns = (hosttime - WLTIMEDELTA) * 100; + host_ts = ns_to_timespec(host_tns); + + do_settimeofday(&host_ts); +} + +/* + * Set the host time in a process context. + */ + +struct adj_time_work { + struct work_struct work; + u64 host_time; +}; + +static void hv_set_host_time(struct work_struct *work) +{ + struct adj_time_work *wrk; + + wrk = container_of(work, struct adj_time_work, work); + do_adj_guesttime(wrk->host_time); + kfree(wrk); +} + +/* + * 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. The other + * thing is, systime is automatically set to emulated hardware clock which may + * not be UTC time or in the same time zone. So, to override these effects, we + * use the first 50 time samples for initial system time setting. + */ +static inline void adj_guesttime(u64 hosttime, u8 flags) +{ + struct adj_time_work *wrk; + static s32 scnt = 50; + + wrk = kmalloc(sizeof(struct adj_time_work), GFP_ATOMIC); + if (wrk == NULL) + return; + + wrk->host_time = hosttime; + if ((flags & ICTIMESYNCFLAG_SYNC) != 0) { + INIT_WORK(&wrk->work, hv_set_host_time); + schedule_work(&wrk->work); + return; + } + + if ((flags & ICTIMESYNCFLAG_SAMPLE) != 0 && scnt > 0) { + scnt--; + INIT_WORK(&wrk->work, hv_set_host_time); + schedule_work(&wrk->work); + } else + kfree(wrk); +} + +/* + * 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; + u8 *time_txf_buf = util_timesynch.recv_buffer; + struct icmsg_negotiate *negop = NULL; + + 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) { + vmbus_prep_negotiate_resp(icmsghdrp, negop, + time_txf_buf, + util_fw_version, + ts_srv_version); + } else { + timedatap = (struct ictimesync_data *)&time_txf_buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + adj_guesttime(timedatap->parenttime, 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; + struct icmsg_negotiate *negop = NULL; + + vmbus_recvpacket(channel, hbeat_txf_buf, + PAGE_SIZE, &recvlen, &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&hbeat_txf_buf[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + vmbus_prep_negotiate_resp(icmsghdrp, negop, + hbeat_txf_buf, util_fw_version, + hb_srv_version); + } 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; + 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_state(dev->channel, false); + + hv_set_drvdata(dev, srv); + + /* + * Based on the host; initialize the framework and + * service version numbers we will negotiate. + */ + switch (vmbus_proto_version) { + case (VERSION_WS2008): + util_fw_version = UTIL_WS2K8_FW_VERSION; + sd_srv_version = SD_WS2008_VERSION; + ts_srv_version = TS_WS2008_VERSION; + hb_srv_version = HB_WS2008_VERSION; + break; + + default: + util_fw_version = UTIL_FW_VERSION; + sd_srv_version = SD_VERSION; + ts_srv_version = TS_VERSION; + hb_srv_version = HB_VERSION; + } + + 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, +}; + +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/kernel/drivers/hv/hyperv_vmbus.h b/kernel/drivers/hv/hyperv_vmbus.h new file mode 100644 index 000000000..887287ad4 --- /dev/null +++ b/kernel/drivers/hv/hyperv_vmbus.h @@ -0,0 +1,738 @@ +/* + * + * 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 <asm/sync_bitops.h> +#include <linux/atomic.h> +#include <linux/hyperv.h> + +/* + * The below CPUID leaves are present if VersionAndFeatures.HypervisorPresent + * is set by CPUID(HVCPUID_VERSION_FEATURES). + */ +enum hv_cpuid_function { + HVCPUID_VERSION_FEATURES = 0x00000001, + HVCPUID_VENDOR_MAXFUNCTION = 0x40000000, + HVCPUID_INTERFACE = 0x40000001, + + /* + * The remaining functions depend on the value of + * HVCPUID_INTERFACE + */ + HVCPUID_VERSION = 0x40000002, + HVCPUID_FEATURES = 0x40000003, + HVCPUID_ENLIGHTENMENT_INFO = 0x40000004, + HVCPUID_IMPLEMENTATION_LIMITS = 0x40000005, +}; + +#define HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE 0x400 + +#define HV_X64_MSR_CRASH_P0 0x40000100 +#define HV_X64_MSR_CRASH_P1 0x40000101 +#define HV_X64_MSR_CRASH_P2 0x40000102 +#define HV_X64_MSR_CRASH_P3 0x40000103 +#define HV_X64_MSR_CRASH_P4 0x40000104 +#define HV_X64_MSR_CRASH_CTL 0x40000105 + +#define HV_CRASH_CTL_CRASH_NOTIFY (1ULL << 63) + +/* Define version of the synthetic interrupt controller. */ +#define HV_SYNIC_VERSION (1) + +/* Define the expected SynIC version. */ +#define HV_SYNIC_VERSION_1 (0x1) + +/* Define synthetic interrupt controller message constants. */ +#define HV_MESSAGE_SIZE (256) +#define HV_MESSAGE_PAYLOAD_BYTE_COUNT (240) +#define HV_MESSAGE_PAYLOAD_QWORD_COUNT (30) +#define HV_ANY_VP (0xFFFFFFFF) + +/* Define synthetic interrupt controller flag constants. */ +#define HV_EVENT_FLAGS_COUNT (256 * 8) +#define HV_EVENT_FLAGS_BYTE_COUNT (256) +#define HV_EVENT_FLAGS_DWORD_COUNT (256 / sizeof(u32)) + +/* Define hypervisor message types. */ +enum hv_message_type { + HVMSG_NONE = 0x00000000, + + /* Memory access messages. */ + HVMSG_UNMAPPED_GPA = 0x80000000, + HVMSG_GPA_INTERCEPT = 0x80000001, + + /* Timer notification messages. */ + HVMSG_TIMER_EXPIRED = 0x80000010, + + /* Error messages. */ + HVMSG_INVALID_VP_REGISTER_VALUE = 0x80000020, + HVMSG_UNRECOVERABLE_EXCEPTION = 0x80000021, + HVMSG_UNSUPPORTED_FEATURE = 0x80000022, + + /* Trace buffer complete messages. */ + HVMSG_EVENTLOG_BUFFERCOMPLETE = 0x80000040, + + /* Platform-specific processor intercept messages. */ + HVMSG_X64_IOPORT_INTERCEPT = 0x80010000, + HVMSG_X64_MSR_INTERCEPT = 0x80010001, + HVMSG_X64_CPUID_INTERCEPT = 0x80010002, + HVMSG_X64_EXCEPTION_INTERCEPT = 0x80010003, + HVMSG_X64_APIC_EOI = 0x80010004, + HVMSG_X64_LEGACY_FP_ERROR = 0x80010005 +}; + +/* Define the number of synthetic interrupt sources. */ +#define HV_SYNIC_SINT_COUNT (16) +#define HV_SYNIC_STIMER_COUNT (4) + +/* Define invalid partition identifier. */ +#define HV_PARTITION_ID_INVALID ((u64)0x0) + +/* Define port identifier type. */ +union hv_port_id { + u32 asu32; + struct { + u32 id:24; + u32 reserved:8; + } u ; +}; + +/* Define port type. */ +enum hv_port_type { + HVPORT_MSG = 1, + HVPORT_EVENT = 2, + HVPORT_MONITOR = 3 +}; + +/* Define port information structure. */ +struct hv_port_info { + enum hv_port_type port_type; + u32 padding; + union { + struct { + u32 target_sint; + u32 target_vp; + u64 rsvdz; + } message_port_info; + struct { + u32 target_sint; + u32 target_vp; + u16 base_flag_bumber; + u16 flag_count; + u32 rsvdz; + } event_port_info; + struct { + u64 monitor_address; + u64 rsvdz; + } monitor_port_info; + }; +}; + +struct hv_connection_info { + enum hv_port_type port_type; + u32 padding; + union { + struct { + u64 rsvdz; + } message_connection_info; + struct { + u64 rsvdz; + } event_connection_info; + struct { + u64 monitor_address; + } monitor_connection_info; + }; +}; + +/* Define synthetic interrupt controller message flags. */ +union hv_message_flags { + u8 asu8; + struct { + u8 msg_pending:1; + u8 reserved:7; + }; +}; + +/* Define synthetic interrupt controller message header. */ +struct hv_message_header { + enum hv_message_type message_type; + u8 payload_size; + union hv_message_flags message_flags; + u8 reserved[2]; + union { + u64 sender; + union hv_port_id port; + }; +}; + +/* + * 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 reserved_z0:12; + u64 sintx:4; + u64 reserved_z1:44; + }; +}; + + +/* Define timer message payload structure. */ +struct hv_timer_message_payload { + u32 timer_index; + u32 reserved; + u64 expiration_time; /* When the timer expired */ + u64 delivery_time; /* When the message was delivered */ +}; + +/* Define synthetic interrupt controller message format. */ +struct hv_message { + struct hv_message_header header; + union { + u64 payload[HV_MESSAGE_PAYLOAD_QWORD_COUNT]; + } u ; +}; + +/* Define the number of message buffers associated with each port. */ +#define HV_PORT_MESSAGE_BUFFER_COUNT (16) + +/* Define the synthetic interrupt message page layout. */ +struct hv_message_page { + struct hv_message sint_message[HV_SYNIC_SINT_COUNT]; +}; + +/* Define the synthetic interrupt controller event flags format. */ +union hv_synic_event_flags { + u8 flags8[HV_EVENT_FLAGS_BYTE_COUNT]; + u32 flags32[HV_EVENT_FLAGS_DWORD_COUNT]; +}; + +/* Define the synthetic interrupt flags page layout. */ +struct hv_synic_event_flags_page { + union hv_synic_event_flags sintevent_flags[HV_SYNIC_SINT_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]; +}; + +/* Declare the various hypercall operations. */ +enum hv_call_code { + HVCALL_POST_MESSAGE = 0x005c, + HVCALL_SIGNAL_EVENT = 0x005d, +}; + +/* Definition of the hv_post_message hypercall input structure. */ +struct hv_input_post_message { + union hv_connection_id connectionid; + u32 reserved; + enum hv_message_type message_type; + u32 payload_size; + u64 payload[HV_MESSAGE_PAYLOAD_QWORD_COUNT]; +}; + +/* + * Versioning definitions used for guests reporting themselves to the + * hypervisor, and visa versa. + */ + +/* Version info reported by guest OS's */ +enum hv_guest_os_vendor { + HVGUESTOS_VENDOR_MICROSOFT = 0x0001 +}; + +enum hv_guest_os_microsoft_ids { + HVGUESTOS_MICROSOFT_UNDEFINED = 0x00, + HVGUESTOS_MICROSOFT_MSDOS = 0x01, + HVGUESTOS_MICROSOFT_WINDOWS3X = 0x02, + HVGUESTOS_MICROSOFT_WINDOWS9X = 0x03, + HVGUESTOS_MICROSOFT_WINDOWSNT = 0x04, + HVGUESTOS_MICROSOFT_WINDOWSCE = 0x05 +}; + +/* + * Declare the MSR used to identify the guest OS. + */ +#define HV_X64_MSR_GUEST_OS_ID 0x40000000 + +union hv_x64_msr_guest_os_id_contents { + u64 as_uint64; + struct { + u64 build_number:16; + u64 service_version:8; /* Service Pack, etc. */ + u64 minor_version:8; + u64 major_version:8; + u64 os_id:8; /* enum hv_guest_os_microsoft_ids (if Vendor=MS) */ + u64 vendor_id:16; /* enum hv_guest_os_vendor */ + }; +}; + +/* + * Declare the MSR used to setup pages used to communicate with the hypervisor. + */ +#define HV_X64_MSR_HYPERCALL 0x40000001 + +union hv_x64_msr_hypercall_contents { + u64 as_uint64; + struct { + u64 enable:1; + u64 reserved:11; + u64 guest_physical_address:52; + }; +}; + + +enum { + VMBUS_MESSAGE_CONNECTION_ID = 1, + 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, +}; + +/* #defines */ + +#define HV_PRESENT_BIT 0x80000000 + +/* + * The guest OS needs to register the guest ID with the hypervisor. + * The guest ID is a 64 bit entity and the structure of this ID is + * specified in the Hyper-V specification: + * + * http://msdn.microsoft.com/en-us/library/windows/hardware/ff542653%28v=vs.85%29.aspx + * + * While the current guideline does not specify how Linux guest ID(s) + * need to be generated, our plan is to publish the guidelines for + * Linux and other guest operating systems that currently are hosted + * on Hyper-V. The implementation here conforms to this yet + * unpublished guidelines. + * + * + * Bit(s) + * 63 - Indicates if the OS is Open Source or not; 1 is Open Source + * 62:56 - Os Type; Linux is 0x100 + * 55:48 - Distro specific identification + * 47:16 - Linux kernel version number + * 15:0 - Distro specific identification + * + * + */ + +#define HV_LINUX_VENDOR_ID 0x8100 + +/* + * Generate the guest ID based on the guideline described above. + */ + +static inline __u64 generate_guest_id(__u8 d_info1, __u32 kernel_version, + __u16 d_info2) +{ + __u64 guest_id = 0; + + guest_id = (((__u64)HV_LINUX_VENDOR_ID) << 48); + guest_id |= (((__u64)(d_info1)) << 48); + guest_id |= (((__u64)(kernel_version)) << 16); + guest_id |= ((__u64)(d_info2)); + + return guest_id; +} + + +#define HV_CPU_POWER_MANAGEMENT (1 << 0) +#define HV_RECOMMENDATIONS_MAX 4 + +#define HV_X64_MAX 5 +#define HV_CAPS_MAX 8 + + +#define HV_HYPERCALL_PARAM_ALIGN sizeof(u64) + + +/* Service definitions */ + +#define HV_SERVICE_PARENT_PORT (0) +#define HV_SERVICE_PARENT_CONNECTION (0) + +#define HV_SERVICE_CONNECT_RESPONSE_SUCCESS (0) +#define HV_SERVICE_CONNECT_RESPONSE_INVALID_PARAMETER (1) +#define HV_SERVICE_CONNECT_RESPONSE_UNKNOWN_SERVICE (2) +#define HV_SERVICE_CONNECT_RESPONSE_CONNECTION_REJECTED (3) + +#define HV_SERVICE_CONNECT_REQUEST_MESSAGE_ID (1) +#define HV_SERVICE_CONNECT_RESPONSE_MESSAGE_ID (2) +#define HV_SERVICE_DISCONNECT_REQUEST_MESSAGE_ID (3) +#define HV_SERVICE_DISCONNECT_RESPONSE_MESSAGE_ID (4) +#define HV_SERVICE_MAX_MESSAGE_ID (4) + +#define HV_SERVICE_PROTOCOL_VERSION (0x0010) +#define HV_CONNECT_PAYLOAD_BYTE_COUNT 64 + +/* #define VMBUS_REVISION_NUMBER 6 */ + +/* Our local vmbus's port and connection id. Anything >0 is fine */ +/* #define VMBUS_PORT_ID 11 */ + +/* 628180B8-308D-4c5e-B7DB-1BEB62E62EF4 */ +static const uuid_le VMBUS_SERVICE_ID = { + .b = { + 0xb8, 0x80, 0x81, 0x62, 0x8d, 0x30, 0x5e, 0x4c, + 0xb7, 0xdb, 0x1b, 0xeb, 0x62, 0xe6, 0x2e, 0xf4 + }, +}; + + + +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 *hypercall_page; + + bool synic_initialized; + + void *synic_message_page[NR_CPUS]; + void *synic_event_page[NR_CPUS]; + /* + * Hypervisor's notion of virtual processor ID is different from + * Linux' notion of CPU ID. This information can only be retrieved + * in the context of the calling CPU. Setup a map for easy access + * to this information: + * + * vp_index[a] is the Hyper-V's processor ID corresponding to + * Linux cpuid 'a'. + */ + u32 vp_index[NR_CPUS]; + /* + * Starting with win8, we can take channel interrupts on any CPU; + * we will manage the tasklet that handles events on a per CPU + * basis. + */ + struct tasklet_struct *event_dpc[NR_CPUS]; + /* + * To optimize the mapping of relid to channel, maintain + * per-cpu list of the channels based on their CPU affinity. + */ + struct list_head percpu_list[NR_CPUS]; + /* + * buffer to post messages to the host. + */ + void *post_msg_page[NR_CPUS]; + /* + * Support PV clockevent device. + */ + struct clock_event_device *clk_evt[NR_CPUS]; +}; + +extern struct hv_context hv_context; + +struct hv_ring_buffer_debug_info { + u32 current_interrupt_mask; + u32 current_read_index; + u32 current_write_index; + u32 bytes_avail_toread; + u32 bytes_avail_towrite; +}; + +/* Hv Interface */ + +extern int hv_init(void); + +extern void hv_cleanup(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 u16 hv_signal_event(void *con_id); + +extern int hv_synic_alloc(void); + +extern void hv_synic_free(void); + +extern void hv_synic_init(void *irqarg); + +extern void hv_synic_cleanup(void *arg); + +extern void hv_synic_clockevents_cleanup(void); + +/* + * Host version information. + */ +extern unsigned int host_info_eax; +extern unsigned int host_info_ebx; +extern unsigned int host_info_ecx; +extern unsigned int host_info_edx; + +/* Interface */ + + +int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info, void *buffer, + u32 buflen); + +void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info); + +int hv_ringbuffer_write(struct hv_ring_buffer_info *ring_info, + struct kvec *kv_list, + u32 kv_count, bool *signal); + +int hv_ringbuffer_peek(struct hv_ring_buffer_info *ring_info, void *buffer, + u32 buflen); + +int hv_ringbuffer_read(struct hv_ring_buffer_info *ring_info, + void *buffer, + u32 buflen, + u32 offset, bool *signal); + + +void hv_ringbuffer_get_debuginfo(struct hv_ring_buffer_info *ring_info, + struct hv_ring_buffer_debug_info *debug_info); + +void hv_begin_read(struct hv_ring_buffer_info *rbi); + +u32 hv_end_read(struct hv_ring_buffer_info *rbi); + +/* + * 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 { + enum vmbus_connect_state conn_state; + + atomic_t next_gpadl_handle; + + /* + * 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; + spinlock_t channel_lock; + + struct workqueue_struct *work_queue; +}; + + +struct vmbus_msginfo { + /* Bookkeeping stuff */ + struct list_head msglist_entry; + + /* The message itself */ + unsigned char msg[0]; +}; + + +extern struct vmbus_connection vmbus_connection; + +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 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); + +/* static void */ +/* VmbusChildDeviceDestroy( */ +/* struct hv_device *); */ + +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); + +int vmbus_set_event(struct vmbus_channel *channel); + +void vmbus_on_event(unsigned long data); + +int hv_fcopy_init(struct hv_util_service *); +void hv_fcopy_deinit(void); +void hv_fcopy_onchannelcallback(void *); + + +#endif /* _HYPERV_VMBUS_H */ diff --git a/kernel/drivers/hv/ring_buffer.c b/kernel/drivers/hv/ring_buffer.c new file mode 100644 index 000000000..6361d124f --- /dev/null +++ b/kernel/drivers/hv/ring_buffer.c @@ -0,0 +1,566 @@ +/* + * + * 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 "hyperv_vmbus.h" + +void hv_begin_read(struct hv_ring_buffer_info *rbi) +{ + rbi->ring_buffer->interrupt_mask = 1; + mb(); +} + +u32 hv_end_read(struct hv_ring_buffer_info *rbi) +{ + u32 read; + u32 write; + + rbi->ring_buffer->interrupt_mask = 0; + mb(); + + /* + * Now check to see if the ring buffer is still empty. + * If it is not, we raced and we need to process new + * incoming messages. + */ + hv_get_ringbuffer_availbytes(rbi, &read, &write); + + return read; +} + +/* + * 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. + */ + +static bool hv_need_to_signal(u32 old_write, struct hv_ring_buffer_info *rbi) +{ + mb(); + if (rbi->ring_buffer->interrupt_mask) + return false; + + /* check interrupt_mask before read_index */ + rmb(); + /* + * This is the only case we need to signal when the + * ring transitions from being empty to non-empty. + */ + if (old_write == rbi->ring_buffer->read_index) + return true; + + return false; +} + +/* + * To optimize the flow management on the send-side, + * when the sender is blocked because of lack of + * sufficient space in the ring buffer, potential the + * consumer of the ring buffer can signal the producer. + * This is controlled by the following parameters: + * + * 1. pending_send_sz: This is the size in bytes that the + * producer is trying to send. + * 2. The feature bit feat_pending_send_sz set to indicate if + * the consumer of the ring will signal when the ring + * state transitions from being full to a state where + * there is room for the producer to send the pending packet. + */ + +static bool hv_need_to_signal_on_read(u32 old_rd, + struct hv_ring_buffer_info *rbi) +{ + u32 prev_write_sz; + u32 cur_write_sz; + u32 r_size; + u32 write_loc = rbi->ring_buffer->write_index; + u32 read_loc = rbi->ring_buffer->read_index; + u32 pending_sz = rbi->ring_buffer->pending_send_sz; + + /* + * If the other end is not blocked on write don't bother. + */ + if (pending_sz == 0) + return false; + + r_size = rbi->ring_datasize; + cur_write_sz = write_loc >= read_loc ? r_size - (write_loc - read_loc) : + read_loc - write_loc; + + prev_write_sz = write_loc >= old_rd ? r_size - (write_loc - old_rd) : + old_rd - write_loc; + + + if ((prev_write_sz < pending_sz) && (cur_write_sz >= pending_sz)) + return true; + + return false; +} + +/* + * hv_get_next_write_location() + * + * 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; +} + +/* + * hv_set_next_write_location() + * + * 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; +} + +/* + * hv_get_next_read_location() + * + * Get the next read location for the specified ring buffer + */ +static inline u32 +hv_get_next_read_location(struct hv_ring_buffer_info *ring_info) +{ + u32 next = ring_info->ring_buffer->read_index; + + return next; +} + +/* + * hv_get_next_readlocation_withoffset() + * + * Get the next read location + offset for the specified ring buffer. + * This allows the caller to skip + */ +static inline u32 +hv_get_next_readlocation_withoffset(struct hv_ring_buffer_info *ring_info, + u32 offset) +{ + u32 next = ring_info->ring_buffer->read_index; + + next += offset; + next %= ring_info->ring_datasize; + + return next; +} + +/* + * + * hv_set_next_read_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; +} + + +/* + * + * hv_get_ring_buffer() + * + * Get the start of the ring buffer + */ +static inline void * +hv_get_ring_buffer(struct hv_ring_buffer_info *ring_info) +{ + return (void *)ring_info->ring_buffer->buffer; +} + + +/* + * + * hv_get_ring_buffersize() + * + * Get the size of the ring buffer + */ +static inline u32 +hv_get_ring_buffersize(struct hv_ring_buffer_info *ring_info) +{ + return ring_info->ring_datasize; +} + +/* + * + * hv_get_ring_bufferindices() + * + * 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; +} + +/* + * + * hv_copyfrom_ringbuffer() + * + * Helper routine to copy to source from ring buffer. + * Assume there is enough room. Handles wrap-around in src case only!! + * + */ +static u32 hv_copyfrom_ringbuffer( + struct hv_ring_buffer_info *ring_info, + void *dest, + u32 destlen, + u32 start_read_offset) +{ + void *ring_buffer = hv_get_ring_buffer(ring_info); + u32 ring_buffer_size = hv_get_ring_buffersize(ring_info); + + u32 frag_len; + + /* wrap-around detected at the src */ + if (destlen > ring_buffer_size - start_read_offset) { + frag_len = ring_buffer_size - start_read_offset; + + memcpy(dest, ring_buffer + start_read_offset, frag_len); + memcpy(dest + frag_len, ring_buffer, destlen - frag_len); + } else + + memcpy(dest, ring_buffer + start_read_offset, destlen); + + + start_read_offset += destlen; + start_read_offset %= ring_buffer_size; + + return start_read_offset; +} + + +/* + * + * hv_copyto_ringbuffer() + * + * 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, + void *src, + u32 srclen) +{ + void *ring_buffer = hv_get_ring_buffer(ring_info); + u32 ring_buffer_size = hv_get_ring_buffersize(ring_info); + u32 frag_len; + + /* wrap-around detected! */ + if (srclen > ring_buffer_size - start_write_offset) { + frag_len = ring_buffer_size - start_write_offset; + memcpy(ring_buffer + start_write_offset, src, frag_len); + memcpy(ring_buffer, src + frag_len, srclen - frag_len); + } else + memcpy(ring_buffer + start_write_offset, src, srclen); + + start_write_offset += srclen; + start_write_offset %= ring_buffer_size; + + return start_write_offset; +} + +/* + * + * hv_ringbuffer_get_debuginfo() + * + * Get various debug metrics for the specified ring buffer + * + */ +void hv_ringbuffer_get_debuginfo(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) { + 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; + } +} + +/* + * + * hv_ringbuffer_init() + * + *Initialize the ring buffer + * + */ +int hv_ringbuffer_init(struct hv_ring_buffer_info *ring_info, + void *buffer, u32 buflen) +{ + if (sizeof(struct hv_ring_buffer) != PAGE_SIZE) + return -EINVAL; + + memset(ring_info, 0, sizeof(struct hv_ring_buffer_info)); + + ring_info->ring_buffer = (struct hv_ring_buffer *)buffer; + 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 = buflen; + ring_info->ring_datasize = buflen - sizeof(struct hv_ring_buffer); + + spin_lock_init(&ring_info->ring_lock); + + return 0; +} + +/* + * + * hv_ringbuffer_cleanup() + * + * Cleanup the ring buffer + * + */ +void hv_ringbuffer_cleanup(struct hv_ring_buffer_info *ring_info) +{ +} + +/* + * + * hv_ringbuffer_write() + * + * Write to the ring buffer + * + */ +int hv_ringbuffer_write(struct hv_ring_buffer_info *outring_info, + struct kvec *kv_list, u32 kv_count, bool *signal) +{ + int i = 0; + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + u32 totalbytes_towrite = 0; + + u32 next_write_location; + u32 old_write; + u64 prev_indices = 0; + unsigned long flags; + + for (i = 0; i < kv_count; i++) + totalbytes_towrite += kv_list[i].iov_len; + + totalbytes_towrite += sizeof(u64); + + spin_lock_irqsave(&outring_info->ring_lock, flags); + + hv_get_ringbuffer_availbytes(outring_info, + &bytes_avail_toread, + &bytes_avail_towrite); + + + /* 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 */ + mb(); + + /* Now, update the write location */ + hv_set_next_write_location(outring_info, next_write_location); + + + spin_unlock_irqrestore(&outring_info->ring_lock, flags); + + *signal = hv_need_to_signal(old_write, outring_info); + return 0; +} + + +/* + * + * hv_ringbuffer_peek() + * + * Read without advancing the read index + * + */ +int hv_ringbuffer_peek(struct hv_ring_buffer_info *Inring_info, + void *Buffer, u32 buflen) +{ + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + u32 next_read_location = 0; + unsigned long flags; + + spin_lock_irqsave(&Inring_info->ring_lock, flags); + + hv_get_ringbuffer_availbytes(Inring_info, + &bytes_avail_toread, + &bytes_avail_towrite); + + /* Make sure there is something to read */ + if (bytes_avail_toread < buflen) { + + spin_unlock_irqrestore(&Inring_info->ring_lock, flags); + + return -EAGAIN; + } + + /* Convert to byte offset */ + next_read_location = hv_get_next_read_location(Inring_info); + + next_read_location = hv_copyfrom_ringbuffer(Inring_info, + Buffer, + buflen, + next_read_location); + + spin_unlock_irqrestore(&Inring_info->ring_lock, flags); + + return 0; +} + + +/* + * + * hv_ringbuffer_read() + * + * Read and advance the read index + * + */ +int hv_ringbuffer_read(struct hv_ring_buffer_info *inring_info, void *buffer, + u32 buflen, u32 offset, bool *signal) +{ + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + u32 next_read_location = 0; + u64 prev_indices = 0; + unsigned long flags; + u32 old_read; + + if (buflen <= 0) + return -EINVAL; + + spin_lock_irqsave(&inring_info->ring_lock, flags); + + hv_get_ringbuffer_availbytes(inring_info, + &bytes_avail_toread, + &bytes_avail_towrite); + + old_read = bytes_avail_toread; + + /* Make sure there is something to read */ + if (bytes_avail_toread < buflen) { + spin_unlock_irqrestore(&inring_info->ring_lock, flags); + + return -EAGAIN; + } + + next_read_location = + hv_get_next_readlocation_withoffset(inring_info, offset); + + next_read_location = hv_copyfrom_ringbuffer(inring_info, + buffer, + buflen, + next_read_location); + + next_read_location = hv_copyfrom_ringbuffer(inring_info, + &prev_indices, + sizeof(u64), + next_read_location); + + /* 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 */ + mb(); + + /* Update the read index */ + hv_set_next_read_location(inring_info, next_read_location); + + spin_unlock_irqrestore(&inring_info->ring_lock, flags); + + *signal = hv_need_to_signal_on_read(old_read, inring_info); + + return 0; +} diff --git a/kernel/drivers/hv/vmbus_drv.c b/kernel/drivers/hv/vmbus_drv.c new file mode 100644 index 000000000..c85235e9f --- /dev/null +++ b/kernel/drivers/hv/vmbus_drv.c @@ -0,0 +1,1114 @@ +/* + * 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 <asm/hyperv.h> +#include <asm/hypervisor.h> +#include <asm/mshyperv.h> +#include <linux/notifier.h> +#include <linux/ptrace.h> +#include "hyperv_vmbus.h" + +static struct acpi_device *hv_acpi_dev; + +static struct tasklet_struct msg_dpc; +static struct completion probe_event; +static int irq; + + +static int hyperv_panic_event(struct notifier_block *nb, + unsigned long event, void *ptr) +{ + struct pt_regs *regs; + + regs = current_pt_regs(); + + wrmsrl(HV_X64_MSR_CRASH_P0, regs->ip); + wrmsrl(HV_X64_MSR_CRASH_P1, regs->ax); + wrmsrl(HV_X64_MSR_CRASH_P2, regs->bx); + wrmsrl(HV_X64_MSR_CRASH_P3, regs->cx); + wrmsrl(HV_X64_MSR_CRASH_P4, regs->dx); + + /* + * Let Hyper-V know there is crash data available + */ + wrmsrl(HV_X64_MSR_CRASH_CTL, HV_CRASH_CTL_CRASH_NOTIFY); + return NOTIFY_DONE; +} + +static struct notifier_block hyperv_panic_block = { + .notifier_call = hyperv_panic_event, +}; + +struct resource hyperv_mmio = { + .name = "hyperv mmio", + .flags = IORESOURCE_MEM, +}; +EXPORT_SYMBOL_GPL(hyperv_mmio); + +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(struct vmbus_channel *channel) +{ + return (u8)channel->offermsg.monitorid / 32; +} + +static u8 channel_monitor_offset(struct vmbus_channel *channel) +{ + return (u8)channel->offermsg.monitorid % 32; +} + +static u32 channel_pending(struct vmbus_channel *channel, + 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(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->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); + +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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->outbound, &outbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + 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; + + if (!hv_dev->channel) + return -ENODEV; + hv_ringbuffer_get_debuginfo(&hv_dev->channel->inbound, &inbound); + return sprintf(buf, "%d\n", inbound.bytes_avail_towrite); +} +static DEVICE_ATTR_RO(in_write_bytes_avail); + +/* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */ +static struct attribute *vmbus_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, + &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, + NULL, +}; +ATTRIBUTE_GROUPS(vmbus); + +/* + * 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 __u8 *guid) +{ + if (memcmp(guid, &null_guid, sizeof(uuid_le))) + 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( + const struct hv_vmbus_device_id *id, + const __u8 *guid) +{ + for (; !is_null_guid(id->guid); id++) + if (!memcmp(&id->guid, guid, sizeof(uuid_le))) + return id; + + return NULL; +} + + + +/* + * 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); + + if (hv_vmbus_get_id(drv->id_table, hv_dev->dev_type.b)) + 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->id_table, dev->dev_type.b); + 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); + u32 relid = dev->channel->offermsg.child_relid; + + if (child_device->driver) { + drv = drv_to_hv_drv(child_device->driver); + if (drv->remove) + drv->remove(dev); + else { + hv_process_channel_removal(dev->channel, relid); + pr_err("remove not set for driver %s\n", + dev_name(child_device)); + } + } else { + /* + * We don't have a driver for this device; deal with the + * rescind message by removing the channel. + */ + hv_process_channel_removal(dev->channel, relid); + } + + 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); + + return; +} + + +/* + * 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); + + 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_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, int cpu) +{ + struct clock_event_device *dev = hv_context.clk_evt[cpu]; + + if (dev->event_handler) + dev->event_handler(dev); + + msg->header.message_type = HVMSG_NONE; + + /* + * Make sure the write to MessageType (ie set to + * HVMSG_NONE) happens before we read the + * MessagePending and EOMing. Otherwise, the EOMing + * will not deliver any more messages since there is + * no empty slot + */ + mb(); + + if (msg->header.message_flags.msg_pending) { + /* + * This will cause message queue rescan to + * possibly deliver another msg from the + * hypervisor + */ + wrmsrl(HV_X64_MSR_EOM, 0); + } +} + +static void vmbus_on_msg_dpc(unsigned long data) +{ + int cpu = smp_processor_id(); + void *page_addr = hv_context.synic_message_page[cpu]; + struct hv_message *msg = (struct hv_message *)page_addr + + VMBUS_MESSAGE_SINT; + struct vmbus_channel_message_header *hdr; + struct vmbus_channel_message_table_entry *entry; + struct onmessage_work_context *ctx; + + while (1) { + if (msg->header.message_type == HVMSG_NONE) + /* no msg */ + break; + + hdr = (struct vmbus_channel_message_header *)msg->u.payload; + + 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->handler_type == VMHT_BLOCKING) { + ctx = kmalloc(sizeof(*ctx), GFP_ATOMIC); + if (ctx == NULL) + continue; + + INIT_WORK(&ctx->work, vmbus_onmessage_work); + memcpy(&ctx->msg, msg, sizeof(*msg)); + + queue_work(vmbus_connection.work_queue, &ctx->work); + } else + entry->message_handler(hdr); + +msg_handled: + msg->header.message_type = HVMSG_NONE; + + /* + * Make sure the write to MessageType (ie set to + * HVMSG_NONE) happens before we read the + * MessagePending and EOMing. Otherwise, the EOMing + * will not deliver any more messages since there is + * no empty slot + */ + mb(); + + if (msg->header.message_flags.msg_pending) { + /* + * This will cause message queue rescan to + * possibly deliver another msg from the + * hypervisor + */ + wrmsrl(HV_X64_MSR_EOM, 0); + } + } +} + +static void vmbus_isr(void) +{ + int cpu = smp_processor_id(); + void *page_addr; + struct hv_message *msg; + union hv_synic_event_flags *event; + bool handled = false; + + page_addr = hv_context.synic_event_page[cpu]; + if (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, + (unsigned long *) &event->flags32[0])) { + 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) + tasklet_schedule(hv_context.event_dpc[cpu]); + + + page_addr = hv_context.synic_message_page[cpu]; + 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, cpu); + else + tasklet_schedule(&msg_dpc); + } +} + +#ifdef CONFIG_HOTPLUG_CPU +static int hyperv_cpu_disable(void) +{ + return -ENOSYS; +} + +static void hv_cpu_hotplug_quirk(bool vmbus_loaded) +{ + static void *previous_cpu_disable; + + /* + * Offlining a CPU when running on newer hypervisors (WS2012R2, Win8, + * ...) is not supported at this moment as channel interrupts are + * distributed across all of them. + */ + + if ((vmbus_proto_version == VERSION_WS2008) || + (vmbus_proto_version == VERSION_WIN7)) + return; + + if (vmbus_loaded) { + previous_cpu_disable = smp_ops.cpu_disable; + smp_ops.cpu_disable = hyperv_cpu_disable; + pr_notice("CPU offlining is not supported by hypervisor\n"); + } else if (previous_cpu_disable) + smp_ops.cpu_disable = previous_cpu_disable; +} +#else +static void hv_cpu_hotplug_quirk(bool vmbus_loaded) +{ +} +#endif + +/* + * vmbus_bus_init -Main vmbus driver initialization routine. + * + * Here, we + * - initialize the vmbus driver context + * - invoke the vmbus hv main init routine + * - get the irq resource + * - retrieve the channel offers + */ +static int vmbus_bus_init(int irq) +{ + 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; + } + + tasklet_init(&msg_dpc, vmbus_on_msg_dpc, 0); + + ret = bus_register(&hv_bus); + if (ret) + goto err_cleanup; + + 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. + */ + on_each_cpu(hv_synic_init, NULL, 1); + ret = vmbus_connect(); + if (ret) + goto err_alloc; + + hv_cpu_hotplug_quirk(true); + + /* + * Only register if the crash MSRs are available + */ + if (ms_hyperv.features & HV_FEATURE_GUEST_CRASH_MSR_AVAILABLE) { + atomic_notifier_chain_register(&panic_notifier_list, + &hyperv_panic_block); + } + + vmbus_request_offers(); + + return 0; + +err_alloc: + hv_synic_free(); + hv_remove_vmbus_irq(); + + bus_unregister(&hv_bus); + +err_cleanup: + hv_cleanup(); + + return ret; +} + +/** + * __vmbus_child_driver_register - Register a vmbus's driver + * @drv: 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; + + ret = driver_register(&hv_driver->driver); + + return ret; +} +EXPORT_SYMBOL_GPL(__vmbus_driver_register); + +/** + * vmbus_driver_unregister() - Unregister a vmbus's driver + * @drv: 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); +} +EXPORT_SYMBOL_GPL(vmbus_driver_unregister); + +/* + * 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)); + + + return child_device_obj; +} + +/* + * vmbus_device_register - Register the child device + */ +int vmbus_device_register(struct hv_device *child_device_obj) +{ + int ret = 0; + + dev_set_name(&child_device_obj->device, "vmbus_%d", + child_device_obj->channel->id); + + 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"); + else + pr_debug("child device %s registered\n", + dev_name(&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)); + + /* + * 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 the information we + * need from DSDT. + */ + +static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *ctx) +{ + switch (res->type) { + case ACPI_RESOURCE_TYPE_IRQ: + irq = res->data.irq.interrupts[0]; + break; + + case ACPI_RESOURCE_TYPE_ADDRESS64: + hyperv_mmio.start = res->data.address64.address.minimum; + hyperv_mmio.end = res->data.address64.address.maximum; + break; + } + + return AE_OK; +} + +static int vmbus_acpi_add(struct acpi_device *device) +{ + acpi_status result; + int ret_val = -ENODEV; + + 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; + /* + * The parent of the vmbus acpi device (Gen2 firmware) is the VMOD that + * has the mmio ranges. Get that. + */ + if (device->parent) { + result = acpi_walk_resources(device->parent->handle, + METHOD_NAME__CRS, + vmbus_walk_resources, NULL); + + if (ACPI_FAILURE(result)) + goto acpi_walk_err; + if (hyperv_mmio.start && hyperv_mmio.end) + request_resource(&iomem_resource, &hyperv_mmio); + } + ret_val = 0; + +acpi_walk_err: + complete(&probe_event); + 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, + }, +}; + +static int __init hv_acpi_init(void) +{ + int ret, t; + + if (x86_hyper != &x86_hyper_ms_hyperv) + return -ENODEV; + + init_completion(&probe_event); + + /* + * Get irq 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; + } + + if (irq <= 0) { + ret = -ENODEV; + goto cleanup; + } + + ret = vmbus_bus_init(irq); + if (ret) + goto cleanup; + + return 0; + +cleanup: + acpi_bus_unregister_driver(&vmbus_acpi_driver); + hv_acpi_dev = NULL; + return ret; +} + +static void __exit vmbus_exit(void) +{ + int cpu; + + vmbus_connection.conn_state = DISCONNECTED; + hv_synic_clockevents_cleanup(); + hv_remove_vmbus_irq(); + vmbus_free_channels(); + bus_unregister(&hv_bus); + hv_cleanup(); + for_each_online_cpu(cpu) + smp_call_function_single(cpu, hv_synic_cleanup, NULL, 1); + acpi_bus_unregister_driver(&vmbus_acpi_driver); + hv_cpu_hotplug_quirk(false); + vmbus_disconnect(); +} + + +MODULE_LICENSE("GPL"); + +subsys_initcall(hv_acpi_init); +module_exit(vmbus_exit); |