diff options
Diffstat (limited to 'drivers/hv')
| -rw-r--r-- | drivers/hv/Kconfig | 14 | ||||
| -rw-r--r-- | drivers/hv/Makefile | 7 | ||||
| -rw-r--r-- | drivers/hv/channel.c | 815 | ||||
| -rw-r--r-- | drivers/hv/channel_mgmt.c | 647 | ||||
| -rw-r--r-- | drivers/hv/connection.c | 318 | ||||
| -rw-r--r-- | drivers/hv/hv.c | 429 | ||||
| -rw-r--r-- | drivers/hv/hv_kvp.c | 339 | ||||
| -rw-r--r-- | drivers/hv/hv_kvp.h | 184 | ||||
| -rw-r--r-- | drivers/hv/hv_util.c | 354 | ||||
| -rw-r--r-- | drivers/hv/hyperv_vmbus.h | 628 | ||||
| -rw-r--r-- | drivers/hv/ring_buffer.c | 527 | ||||
| -rw-r--r-- | drivers/hv/vmbus_drv.c | 772 |
12 files changed, 5034 insertions, 0 deletions
diff --git a/drivers/hv/Kconfig b/drivers/hv/Kconfig new file mode 100644 index 000000000000..9fa09ac000ad --- /dev/null +++ b/drivers/hv/Kconfig @@ -0,0 +1,14 @@ +config HYPERV + tristate "Microsoft Hyper-V client drivers" + depends on X86 && ACPI && PCI + 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. + + diff --git a/drivers/hv/Makefile b/drivers/hv/Makefile new file mode 100644 index 000000000000..a23938b991c9 --- /dev/null +++ b/drivers/hv/Makefile @@ -0,0 +1,7 @@ +obj-$(CONFIG_HYPERV) += hv_vmbus.o +obj-$(CONFIG_HYPERV_UTILS) += hv_utils.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 diff --git a/drivers/hv/channel.c b/drivers/hv/channel.c new file mode 100644 index 000000000000..406537420fff --- /dev/null +++ b/drivers/hv/channel.c @@ -0,0 +1,815 @@ +/* + * 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 "hyperv_vmbus.h" + +#define NUM_PAGES_SPANNED(addr, len) \ +((PAGE_ALIGN(addr + len) >> PAGE_SHIFT) - (addr >> PAGE_SHIFT)) + +/* Internal routines */ +static int create_gpadl_header( + void *kbuffer, /* must be phys and virt contiguous */ + u32 size, /* page-size multiple */ + struct vmbus_channel_msginfo **msginfo, + u32 *messagecount); +static void vmbus_setevent(struct vmbus_channel *channel); + +/* + * 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)); + + monitorpage = vmbus_connection.monitor_pages; + monitorpage++; /* Get the child to parent monitor page */ + + sync_set_bit(channel->monitor_bit, + (unsigned long *)&monitorpage->trigger_group + [channel->monitor_grp].pending); + + } else { + vmbus_set_event(channel->offermsg.child_relid); + } +} + +/* + * vmbus_get_debug_info -Retrieve various channel debug info + */ +void vmbus_get_debug_info(struct vmbus_channel *channel, + struct vmbus_channel_debug_info *debuginfo) +{ + struct hv_monitor_page *monitorpage; + u8 monitor_group = (u8)channel->offermsg.monitorid / 32; + u8 monitor_offset = (u8)channel->offermsg.monitorid % 32; + + debuginfo->relid = channel->offermsg.child_relid; + debuginfo->state = channel->state; + memcpy(&debuginfo->interfacetype, + &channel->offermsg.offer.if_type, sizeof(uuid_le)); + memcpy(&debuginfo->interface_instance, + &channel->offermsg.offer.if_instance, + sizeof(uuid_le)); + + monitorpage = (struct hv_monitor_page *)vmbus_connection.monitor_pages; + + debuginfo->monitorid = channel->offermsg.monitorid; + + debuginfo->servermonitor_pending = + monitorpage->trigger_group[monitor_group].pending; + debuginfo->servermonitor_latency = + monitorpage->latency[monitor_group][monitor_offset]; + debuginfo->servermonitor_connectionid = + monitorpage->parameter[monitor_group] + [monitor_offset].connectionid.u.id; + + monitorpage++; + + debuginfo->clientmonitor_pending = + monitorpage->trigger_group[monitor_group].pending; + debuginfo->clientmonitor_latency = + monitorpage->latency[monitor_group][monitor_offset]; + debuginfo->clientmonitor_connectionid = + monitorpage->parameter[monitor_group] + [monitor_offset].connectionid.u.id; + + hv_ringbuffer_get_debuginfo(&channel->inbound, &debuginfo->inbound); + hv_ringbuffer_get_debuginfo(&channel->outbound, &debuginfo->outbound); +} + +/* + * 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, t, err = 0; + + 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) + return -ENOMEM; + + + 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 errorout; + } + + ret = hv_ringbuffer_init( + &newchannel->inbound, in, recv_ringbuffer_size); + if (ret != 0) { + err = ret; + goto errorout; + } + + + /* 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 errorout; + } + + /* 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 errorout; + } + + 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->server_contextarea_gpadlhandle = 0; + + if (userdatalen > MAX_USER_DEFINED_BYTES) { + err = -EINVAL; + goto errorout; + } + + 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) + goto cleanup; + + t = wait_for_completion_timeout(&open_info->waitevent, 5*HZ); + if (t == 0) { + err = -ETIMEDOUT; + goto errorout; + } + + + if (open_info->response.open_result.status) + err = open_info->response.open_result.status; + +cleanup: + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, flags); + list_del(&open_info->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, flags); + + kfree(open_info); + return err; + +errorout: + hv_ringbuffer_cleanup(&newchannel->outbound); + hv_ringbuffer_cleanup(&newchannel->inbound); + free_pages((unsigned long)out, + get_order(send_ringbuffer_size + recv_ringbuffer_size)); + kfree(open_info); + 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; + unsigned long long pfn; + 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; + pfn = virt_to_phys(kbuffer) >> 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] = pfn+i; + *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] = pfn + pfnsum + i; + + /* add to msg header */ + list_add_tail(&msgbody->msglistentry, + &msgheader->submsglist); + pfnsum += pfncurr; + pfnleft -= pfncurr; + } + } else { + /* everything fits in a header */ + msgsize = sizeof(struct vmbus_channel_msginfo) + + sizeof(struct vmbus_channel_gpadl_header) + + sizeof(struct gpa_range) + pagecount * sizeof(u64); + msgheader = kzalloc(msgsize, GFP_KERNEL); + if (msgheader == NULL) + goto nomem; + 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] = pfn+i; + + *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 + * @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; + int t; + + next_gpadl_handle = atomic_read(&vmbus_connection.next_gpadl_handle); + atomic_inc(&vmbus_connection.next_gpadl_handle); + + 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; + + } + } + t = wait_for_completion_timeout(&msginfo->waitevent, 5*HZ); + BUG_ON(t == 0); + + + /* 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, t; + + 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)); + + BUG_ON(ret != 0); + t = wait_for_completion_timeout(&info->waitevent, 5*HZ); + BUG_ON(t == 0); + + /* Received a torndown response */ + 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); + +/* + * vmbus_close - Close the specified channel + */ +void vmbus_close(struct vmbus_channel *channel) +{ + struct vmbus_channel_close_channel *msg; + int ret; + unsigned long flags; + + /* Stop callback and cancel the timer asap */ + spin_lock_irqsave(&channel->inbound_lock, flags); + channel->onchannel_callback = NULL; + spin_unlock_irqrestore(&channel->inbound_lock, flags); + + /* 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)); + + BUG_ON(ret != 0); + /* Tear down the gpadl for the channel's ring buffer */ + if (channel->ringbuffer_gpadlhandle) + vmbus_teardown_gpadl(channel, + channel->ringbuffer_gpadlhandle); + + /* 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)); + + +} +EXPORT_SYMBOL_GPL(vmbus_close); + +/** + * vmbus_sendpacket() - Send the specified buffer on the given channel + * @channel: Pointer to vmbus_channel structure. + * @buffer: Pointer to the buffer you want to receive the data into. + * @bufferlen: Maximum size of what the the buffer will hold + * @requestid: Identifier of the request + * @type: Type of packet that is being send e.g. negotiate, time + * packet etc. + * + * Sends data in @buffer directly to hyper-v via the vmbus + * This will send the data unparsed to hyper-v. + * + * Mainly used by Hyper-V drivers. + */ +int vmbus_sendpacket(struct vmbus_channel *channel, const void *buffer, + u32 bufferlen, u64 requestid, + enum vmbus_packet_type type, u32 flags) +{ + struct vmpacket_descriptor desc; + u32 packetlen = sizeof(struct vmpacket_descriptor) + bufferlen; + u32 packetlen_aligned = ALIGN(packetlen, sizeof(u64)); + struct scatterlist bufferlist[3]; + u64 aligned_data = 0; + int ret; + + + /* 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; + + sg_init_table(bufferlist, 3); + sg_set_buf(&bufferlist[0], &desc, sizeof(struct vmpacket_descriptor)); + sg_set_buf(&bufferlist[1], buffer, bufferlen); + sg_set_buf(&bufferlist[2], &aligned_data, + packetlen_aligned - packetlen); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3); + + if (ret == 0 && !hv_get_ringbuffer_interrupt_mask(&channel->outbound)) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL(vmbus_sendpacket); + +/* + * 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) +{ + int ret; + int i; + struct vmbus_channel_packet_page_buffer desc; + u32 descsize; + u32 packetlen; + u32 packetlen_aligned; + struct scatterlist bufferlist[3]; + u64 aligned_data = 0; + + if (pagecount > MAX_PAGE_BUFFER_COUNT) + return -EINVAL; + + + /* + * Adjust the size down since vmbus_channel_packet_page_buffer is the + * largest size we support + */ + descsize = sizeof(struct vmbus_channel_packet_page_buffer) - + ((MAX_PAGE_BUFFER_COUNT - pagecount) * + sizeof(struct hv_page_buffer)); + packetlen = descsize + bufferlen; + packetlen_aligned = ALIGN(packetlen, sizeof(u64)); + + /* Setup the descriptor */ + desc.type = VM_PKT_DATA_USING_GPA_DIRECT; + desc.flags = VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED; + desc.dataoffset8 = descsize >> 3; /* in 8-bytes 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; + } + + sg_init_table(bufferlist, 3); + sg_set_buf(&bufferlist[0], &desc, descsize); + sg_set_buf(&bufferlist[1], buffer, bufferlen); + sg_set_buf(&bufferlist[2], &aligned_data, + packetlen_aligned - packetlen); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3); + + if (ret == 0 && !hv_get_ringbuffer_interrupt_mask(&channel->outbound)) + vmbus_setevent(channel); + + return ret; +} +EXPORT_SYMBOL_GPL(vmbus_sendpacket_pagebuffer); + +/* + * 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 scatterlist bufferlist[3]; + u64 aligned_data = 0; + u32 pfncount = NUM_PAGES_SPANNED(multi_pagebuffer->offset, + multi_pagebuffer->len); + + + if ((pfncount < 0) || (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)); + + sg_init_table(bufferlist, 3); + sg_set_buf(&bufferlist[0], &desc, descsize); + sg_set_buf(&bufferlist[1], buffer, bufferlen); + sg_set_buf(&bufferlist[2], &aligned_data, + packetlen_aligned - packetlen); + + ret = hv_ringbuffer_write(&channel->outbound, bufferlist, 3); + + if (ret == 0 && !hv_get_ringbuffer_interrupt_mask(&channel->outbound)) + 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; + + *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)); + + + 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; + u32 userlen; + int ret; + + *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 = packetlen; + + if (packetlen > bufferlen) { + pr_err("Buffer too small - needed %d bytes but " + "got space for only %d bytes\n", + 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); + + return 0; +} +EXPORT_SYMBOL_GPL(vmbus_recvpacket_raw); diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c new file mode 100644 index 000000000000..41bf287baa1c --- /dev/null +++ b/drivers/hv/channel_mgmt.c @@ -0,0 +1,647 @@ +/* + * 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" + +struct vmbus_channel_message_table_entry { + enum vmbus_channel_message_type message_type; + void (*message_handler)(struct vmbus_channel_message_header *msg); +}; + +#define MAX_MSG_TYPES 4 +#define MAX_NUM_DEVICE_CLASSES_SUPPORTED 8 + +static const uuid_le + supported_device_classes[MAX_NUM_DEVICE_CLASSES_SUPPORTED] = { + /* {ba6163d9-04a1-4d29-b605-72e2ffb1dc7f} */ + /* Storage - SCSI */ + { + .b = { + 0xd9, 0x63, 0x61, 0xba, 0xa1, 0x04, 0x29, 0x4d, + 0xb6, 0x05, 0x72, 0xe2, 0xff, 0xb1, 0xdc, 0x7f + } + }, + + /* {F8615163-DF3E-46c5-913F-F2D2F965ED0E} */ + /* Network */ + { + .b = { + 0x63, 0x51, 0x61, 0xF8, 0x3E, 0xDF, 0xc5, 0x46, + 0x91, 0x3F, 0xF2, 0xD2, 0xF9, 0x65, 0xED, 0x0E + } + }, + + /* {CFA8B69E-5B4A-4cc0-B98B-8BA1A1F3F95A} */ + /* Input */ + { + .b = { + 0x9E, 0xB6, 0xA8, 0xCF, 0x4A, 0x5B, 0xc0, 0x4c, + 0xB9, 0x8B, 0x8B, 0xA1, 0xA1, 0xF3, 0xF9, 0x5A + } + }, + + /* {32412632-86cb-44a2-9b5c-50d1417354f5} */ + /* IDE */ + { + .b = { + 0x32, 0x26, 0x41, 0x32, 0xcb, 0x86, 0xa2, 0x44, + 0x9b, 0x5c, 0x50, 0xd1, 0x41, 0x73, 0x54, 0xf5 + } + }, + /* 0E0B6031-5213-4934-818B-38D90CED39DB */ + /* Shutdown */ + { + .b = { + 0x31, 0x60, 0x0B, 0X0E, 0x13, 0x52, 0x34, 0x49, + 0x81, 0x8B, 0x38, 0XD9, 0x0C, 0xED, 0x39, 0xDB + } + }, + /* {9527E630-D0AE-497b-ADCE-E80AB0175CAF} */ + /* TimeSync */ + { + .b = { + 0x30, 0xe6, 0x27, 0x95, 0xae, 0xd0, 0x7b, 0x49, + 0xad, 0xce, 0xe8, 0x0a, 0xb0, 0x17, 0x5c, 0xaf + } + }, + /* {57164f39-9115-4e78-ab55-382f3bd5422d} */ + /* Heartbeat */ + { + .b = { + 0x39, 0x4f, 0x16, 0x57, 0x15, 0x91, 0x78, 0x4e, + 0xab, 0x55, 0x38, 0x2f, 0x3b, 0xd5, 0x42, 0x2d + } + }, + /* {A9A0F4E7-5A45-4d96-B827-8A841E8C03E6} */ + /* KVP */ + { + .b = { + 0xe7, 0xf4, 0xa0, 0xa9, 0x45, 0x5a, 0x96, 0x4d, + 0xb8, 0x27, 0x8a, 0x84, 0x1e, 0x8c, 0x3, 0xe6 + } + }, + +}; + + +/** + * 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. This response can + * come from both the vmbus driver and the hv_utils driver. The current api + * will respond properly to both Windows 2008 and Windows 2008-R2 operating + * systems. + * + * Mainly used by Hyper-V drivers. + */ +void prep_negotiate_resp(struct icmsg_hdr *icmsghdrp, + struct icmsg_negotiate *negop, + u8 *buf) +{ + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + icmsghdrp->icmsgsize = 0x10; + + negop = (struct icmsg_negotiate *)&buf[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + if (negop->icframe_vercnt == 2 && + negop->icversion_data[1].major == 3) { + negop->icversion_data[0].major = 3; + negop->icversion_data[0].minor = 0; + negop->icversion_data[1].major = 3; + negop->icversion_data[1].minor = 0; + } else { + negop->icversion_data[0].major = 1; + negop->icversion_data[0].minor = 0; + negop->icversion_data[1].major = 1; + negop->icversion_data[1].minor = 0; + } + + negop->icframe_vercnt = 1; + negop->icmsg_vercnt = 1; + } +} +EXPORT_SYMBOL(prep_negotiate_resp); + +/* + * alloc_channel - Allocate and initialize a vmbus channel object + */ +static struct vmbus_channel *alloc_channel(void) +{ + struct vmbus_channel *channel; + + channel = kzalloc(sizeof(*channel), GFP_ATOMIC); + if (!channel) + return NULL; + + spin_lock_init(&channel->inbound_lock); + + channel->controlwq = create_workqueue("hv_vmbus_ctl"); + if (!channel->controlwq) { + kfree(channel); + return NULL; + } + + return channel; +} + +/* + * release_hannel - Release the vmbus channel object itself + */ +static void release_channel(struct work_struct *work) +{ + struct vmbus_channel *channel = container_of(work, + struct vmbus_channel, + work); + + destroy_workqueue(channel->controlwq); + + kfree(channel); +} + +/* + * free_channel - Release the resources used by the vmbus channel object + */ +void free_channel(struct vmbus_channel *channel) +{ + + /* + * We have to release the channel's workqueue/thread in the vmbus's + * workqueue/thread context + * ie we can't destroy ourselves. + */ + INIT_WORK(&channel->work, release_channel); + queue_work(vmbus_connection.work_queue, &channel->work); +} + + + +/* + * vmbus_process_rescind_offer - + * Rescind the offer by initiating a device removal + */ +static void vmbus_process_rescind_offer(struct work_struct *work) +{ + struct vmbus_channel *channel = container_of(work, + struct vmbus_channel, + work); + + vmbus_device_unregister(channel->device_obj); +} + +/* + * vmbus_process_offer - Process the offer by creating a channel/device + * associated with this offer + */ +static void vmbus_process_offer(struct work_struct *work) +{ + struct vmbus_channel *newchannel = container_of(work, + struct vmbus_channel, + work); + struct vmbus_channel *channel; + bool fnew = true; + int ret; + unsigned long flags; + + /* The next possible work is rescind handling */ + INIT_WORK(&newchannel->work, vmbus_process_rescind_offer); + + /* 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); + + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + + if (!fnew) { + free_channel(newchannel); + return; + } + + /* + * 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); + + /* + * Add the new device to the bus. This will kick off device-driver + * binding which eventually invokes the device driver's AddDevice() + * method. + */ + ret = vmbus_device_register(newchannel->device_obj); + if (ret != 0) { + pr_err("unable to add child device object (relid %d)\n", + newchannel->offermsg.child_relid); + + spin_lock_irqsave(&vmbus_connection.channel_lock, flags); + list_del(&newchannel->listentry); + spin_unlock_irqrestore(&vmbus_connection.channel_lock, flags); + + free_channel(newchannel); + } else { + /* + * 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; + } +} + +/* + * 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; + uuid_le *guidtype; + uuid_le *guidinstance; + int i; + int fsupported = 0; + + offer = (struct vmbus_channel_offer_channel *)hdr; + for (i = 0; i < MAX_NUM_DEVICE_CLASSES_SUPPORTED; i++) { + if (!uuid_le_cmp(offer->offer.if_type, + supported_device_classes[i])) { + fsupported = 1; + break; + } + } + + if (!fsupported) + return; + + guidtype = &offer->offer.if_type; + guidinstance = &offer->offer.if_instance; + + /* Allocate the channel object and save this offer. */ + newchannel = alloc_channel(); + if (!newchannel) { + pr_err("Unable to allocate channel object\n"); + return; + } + + memcpy(&newchannel->offermsg, offer, + sizeof(struct vmbus_channel_offer_channel)); + newchannel->monitor_grp = (u8)offer->monitorid / 32; + newchannel->monitor_bit = (u8)offer->monitorid % 32; + + INIT_WORK(&newchannel->work, vmbus_process_offer); + queue_work(newchannel->controlwq, &newchannel->work); +} + +/* + * 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; + + rescind = (struct vmbus_channel_rescind_offer *)hdr; + channel = relid2channel(rescind->child_relid); + + if (channel == NULL) + /* Just return here, no channel found */ + return; + + /* work is initialized for vmbus_process_rescind_offer() from + * vmbus_process_offer() where the channel got created */ + queue_work(channel->controlwq, &channel->work); +} + +/* + * 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_initiate_contact *initiate; + 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) { + initiate = + (struct vmbus_channel_initiate_contact *)requestheader; + 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 */ +static struct vmbus_channel_message_table_entry + channel_message_table[CHANNELMSG_COUNT] = { + {CHANNELMSG_INVALID, NULL}, + {CHANNELMSG_OFFERCHANNEL, vmbus_onoffer}, + {CHANNELMSG_RESCIND_CHANNELOFFER, vmbus_onoffer_rescind}, + {CHANNELMSG_REQUESTOFFERS, NULL}, + {CHANNELMSG_ALLOFFERS_DELIVERED, vmbus_onoffers_delivered}, + {CHANNELMSG_OPENCHANNEL, NULL}, + {CHANNELMSG_OPENCHANNEL_RESULT, vmbus_onopen_result}, + {CHANNELMSG_CLOSECHANNEL, NULL}, + {CHANNELMSG_GPADL_HEADER, NULL}, + {CHANNELMSG_GPADL_BODY, NULL}, + {CHANNELMSG_GPADL_CREATED, vmbus_ongpadl_created}, + {CHANNELMSG_GPADL_TEARDOWN, NULL}, + {CHANNELMSG_GPADL_TORNDOWN, vmbus_ongpadl_torndown}, + {CHANNELMSG_RELID_RELEASED, NULL}, + {CHANNELMSG_INITIATE_CONTACT, NULL}, + {CHANNELMSG_VERSION_RESPONSE, vmbus_onversion_response}, + {CHANNELMSG_UNLOAD, 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, t; + + msginfo = kmalloc(sizeof(*msginfo) + + sizeof(struct vmbus_channel_message_header), + GFP_KERNEL); + if (!msginfo) + return -ENOMEM; + + init_completion(&msginfo->waitevent); + + 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; + } + + t = wait_for_completion_timeout(&msginfo->waitevent, 5*HZ); + if (t == 0) { + ret = -ETIMEDOUT; + goto cleanup; + } + + + +cleanup: + kfree(msginfo); + + return ret; +} + +/* eof */ diff --git a/drivers/hv/connection.c b/drivers/hv/connection.c new file mode 100644 index 000000000000..5f438b650068 --- /dev/null +++ b/drivers/hv/connection.c @@ -0,0 +1,318 @@ +/* + * + * 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 "hyperv_vmbus.h" + + +struct vmbus_connection vmbus_connection = { + .conn_state = DISCONNECTED, + .next_gpadl_handle = ATOMIC_INIT(0xE1E10), +}; + +/* + * vmbus_connect - Sends a connect request on the partition service connection + */ +int vmbus_connect(void) +{ + int ret = 0; + int t; + struct vmbus_channel_msginfo *msginfo = NULL; + struct vmbus_channel_initiate_contact *msg; + unsigned long flags; + + /* 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 = + (void *)__get_free_pages((GFP_KERNEL|__GFP_ZERO), 1); + if (vmbus_connection.monitor_pages == NULL) { + ret = -ENOMEM; + goto cleanup; + } + + msginfo = kzalloc(sizeof(*msginfo) + + sizeof(struct vmbus_channel_initiate_contact), + GFP_KERNEL); + if (msginfo == NULL) { + ret = -ENOMEM; + goto cleanup; + } + + init_completion(&msginfo->waitevent); + + msg = (struct vmbus_channel_initiate_contact *)msginfo->msg; + + msg->header.msgtype = CHANNELMSG_INITIATE_CONTACT; + msg->vmbus_version_requested = VMBUS_REVISION_NUMBER; + msg->interrupt_page = virt_to_phys(vmbus_connection.int_page); + msg->monitor_page1 = virt_to_phys(vmbus_connection.monitor_pages); + msg->monitor_page2 = virt_to_phys( + (void *)((unsigned long)vmbus_connection.monitor_pages + + PAGE_SIZE)); + + /* + * 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); + goto cleanup; + } + + /* Wait for the connection response */ + t = wait_for_completion_timeout(&msginfo->waitevent, 5*HZ); + if (t == 0) { + spin_lock_irqsave(&vmbus_connection.channelmsg_lock, + flags); + list_del(&msginfo->msglistentry); + spin_unlock_irqrestore(&vmbus_connection.channelmsg_lock, + flags); + ret = -ETIMEDOUT; + goto cleanup; + } + + 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 { + pr_err("Unable to connect, " + "Version %d not supported by Hyper-V\n", + VMBUS_REVISION_NUMBER); + ret = -ECONNREFUSED; + goto cleanup; + } + + kfree(msginfo); + return 0; + +cleanup: + vmbus_connection.conn_state = DISCONNECTED; + + if (vmbus_connection.work_queue) + destroy_workqueue(vmbus_connection.work_queue); + + if (vmbus_connection.int_page) { + free_pages((unsigned long)vmbus_connection.int_page, 0); + vmbus_connection.int_page = NULL; + } + + if (vmbus_connection.monitor_pages) { + free_pages((unsigned long)vmbus_connection.monitor_pages, 1); + vmbus_connection.monitor_pages = NULL; + } + + kfree(msginfo); + + return ret; +} + + +/* + * 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; + + 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; + } + } + 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; + unsigned long flags; + + /* + * Find the channel based on this relid and invokes the + * channel callback to process the event + */ + channel = relid2channel(relid); + + if (!channel) { + pr_err("channel not found for relid - %u\n", relid); + 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 under the + * protection of the channel inbound_lock. Thus, checking + * and invoking the driver specific callback takes care of + * orderly unloading of the driver. + */ + + spin_lock_irqsave(&channel->inbound_lock, flags); + if (channel->onchannel_callback != NULL) + channel->onchannel_callback(channel->channel_callback_context); + else + pr_err("no channel callback for relid - %u\n", relid); + + spin_unlock_irqrestore(&channel->inbound_lock, flags); +} + +/* + * vmbus_on_event - Handler for events + */ +void vmbus_on_event(unsigned long data) +{ + u32 dword; + u32 maxdword = MAX_NUM_CHANNELS_SUPPORTED >> 5; + int bit; + u32 relid; + u32 *recv_int_page = vmbus_connection.recv_int_page; + + /* 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; + + 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 < 3) { + ret = hv_post_message(conn_id, 1, buffer, buflen); + if (ret != HV_STATUS_INSUFFICIENT_BUFFERS) + return ret; + retries++; + msleep(100); + } + return ret; +} + +/* + * vmbus_set_event - Send an event notification to the parent + */ +int vmbus_set_event(u32 child_relid) +{ + /* 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(); +} diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c new file mode 100644 index 000000000000..931b7b030784 --- /dev/null +++ b/drivers/hv/hv.c @@ -0,0 +1,429 @@ +/* + * 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 "hyperv_vmbus.h" + +/* The one and only */ +struct hv_context hv_context = { + .synic_initialized = false, + .hypercall_page = NULL, + .signal_event_param = NULL, + .signal_event_buffer = NULL, +}; + +/* + * query_hypervisor_presence + * - Query the cpuid for presence of windows hypervisor + */ +static int query_hypervisor_presence(void) +{ + unsigned int eax; + unsigned int ebx; + unsigned int ecx; + unsigned int edx; + unsigned int op; + + eax = 0; + ebx = 0; + ecx = 0; + edx = 0; + op = HVCPUID_VERSION_FEATURES; + cpuid(op, &eax, &ebx, &ecx, &edx); + + return ecx & HV_PRESENT_BIT; +} + +/* + * query_hypervisor_info - Get version info of the windows hypervisor + */ +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); + pr_info("Hyper-V Host OS Build:%d-%d.%d-%d-%d.%d\n", + eax, + ebx >> 16, + ebx & 0xFFFF, + ecx, + edx >> 24, + edx & 0xFFFFFF); + } + 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 *) * MAX_NUM_CPUS); + memset(hv_context.synic_message_page, 0, + sizeof(void *) * MAX_NUM_CPUS); + + if (!query_hypervisor_presence()) + goto cleanup; + + max_leaf = query_hypervisor_info(); + + rdmsrl(HV_X64_MSR_GUEST_OS_ID, hv_context.guestid); + + if (hv_context.guestid != 0) + goto cleanup; + + /* Write our OS info */ + wrmsrl(HV_X64_MSR_GUEST_OS_ID, HV_LINUX_GUEST_ID); + hv_context.guestid = HV_LINUX_GUEST_ID; + + /* 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; + + /* Setup the global signal event param for the signal event hypercall */ + hv_context.signal_event_buffer = + kmalloc(sizeof(struct hv_input_signal_event_buffer), + GFP_KERNEL); + if (!hv_context.signal_event_buffer) + goto cleanup; + + hv_context.signal_event_param = + (struct hv_input_signal_event *) + (ALIGN((unsigned long) + hv_context.signal_event_buffer, + HV_HYPERCALL_PARAM_ALIGN)); + hv_context.signal_event_param->connectionid.asu32 = 0; + hv_context.signal_event_param->connectionid.u.id = + VMBUS_EVENT_CONNECTION_ID; + hv_context.signal_event_param->flag_number = 0; + hv_context.signal_event_param->rsvdz = 0; + + 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; + + kfree(hv_context.signal_event_buffer); + hv_context.signal_event_buffer = NULL; + hv_context.signal_event_param = NULL; + + 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. + */ +u16 hv_post_message(union hv_connection_id connection_id, + enum hv_message_type message_type, + void *payload, size_t payload_size) +{ + struct aligned_input { + u64 alignment8; + struct hv_input_post_message msg; + }; + + struct hv_input_post_message *aligned_msg; + u16 status; + unsigned long addr; + + if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT) + return -EMSGSIZE; + + addr = (unsigned long)kmalloc(sizeof(struct aligned_input), GFP_ATOMIC); + if (!addr) + return -ENOMEM; + + aligned_msg = (struct hv_input_post_message *) + (ALIGN(addr, HV_HYPERCALL_PARAM_ALIGN)); + + aligned_msg->connectionid = connection_id; + 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; + + kfree((void *)addr); + + 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) +{ + u16 status; + + status = do_hypercall(HVCALL_SIGNAL_EVENT, + hv_context.signal_event_param, + NULL) & 0xFFFF; + return status; +} + +/* + * 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 *irqarg) +{ + u64 version; + union hv_synic_simp simp; + union hv_synic_siefp siefp; + union hv_synic_sint shared_sint; + union hv_synic_scontrol sctrl; + + u32 irq_vector = *((u32 *)(irqarg)); + int cpu = smp_processor_id(); + + if (!hv_context.hypercall_page) + return; + + /* Check the version */ + rdmsrl(HV_X64_MSR_SVERSION, version); + + 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 cleanup; + } + + 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 cleanup; + } + + /* 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 = irq_vector; /* HV_SHARED_SINT_IDT_VECTOR + 0x20; */ + shared_sint.masked = false; + shared_sint.auto_eoi = false; + + 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; + return; + +cleanup: + 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]); + return; +} + +/* + * 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; + int cpu = smp_processor_id(); + + if (!hv_context.synic_initialized) + return; + + 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); + + free_page((unsigned long)hv_context.synic_message_page[cpu]); + free_page((unsigned long)hv_context.synic_event_page[cpu]); +} diff --git a/drivers/hv/hv_kvp.c b/drivers/hv/hv_kvp.c new file mode 100644 index 000000000000..69c4c985daeb --- /dev/null +++ b/drivers/hv/hv_kvp.c @@ -0,0 +1,339 @@ +/* + * An implementation of key value pair (KVP) functionality for Linux. + * + * + * Copyright (C) 2010, Novell, Inc. + * Author : K. Y. Srinivasan <ksrinivasan@novell.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or + * NON INFRINGEMENT. See the GNU General Public License for more + * details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. + * + */ +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include <linux/net.h> +#include <linux/nls.h> +#include <linux/connector.h> +#include <linux/workqueue.h> +#include <linux/hyperv.h> + +#include "hv_kvp.h" + + + +/* + * 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. */ + int index; /* current index */ + struct vmbus_channel *recv_channel; /* chn we got the request */ + u64 recv_req_id; /* request ID. */ +} kvp_transaction; + +static void kvp_send_key(struct work_struct *dummy); + +#define TIMEOUT_FIRED 1 + +static void kvp_respond_to_host(char *key, char *value, int error); +static void kvp_work_func(struct work_struct *dummy); +static void kvp_register(void); + +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. + */ + +static void +kvp_register(void) +{ + + struct cn_msg *msg; + + msg = kzalloc(sizeof(*msg) + strlen(HV_DRV_VERSION) + 1 , GFP_ATOMIC); + + if (msg) { + msg->id.idx = CN_KVP_IDX; + msg->id.val = CN_KVP_VAL; + msg->seq = KVP_REGISTER; + strcpy(msg->data, HV_DRV_VERSION); + msg->len = strlen(HV_DRV_VERSION) + 1; + cn_netlink_send(msg, 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("Unknown key", "Guest timed out", TIMEOUT_FIRED); +} + +/* + * Callback when data is received from user mode. + */ + +static void +kvp_cn_callback(struct cn_msg *msg, struct netlink_skb_parms *nsp) +{ + struct hv_ku_msg *message; + + message = (struct hv_ku_msg *)msg->data; + if (msg->seq == KVP_REGISTER) { + pr_info("KVP: user-mode registering done.\n"); + kvp_register(); + } + + if (msg->seq == KVP_USER_SET) { + /* + * 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->kvp_key, + message->kvp_value, + !strlen(message->kvp_key)); + } +} + +static void +kvp_send_key(struct work_struct *dummy) +{ + struct cn_msg *msg; + int index = kvp_transaction.index; + + msg = kzalloc(sizeof(*msg) + sizeof(struct hv_kvp_msg) , GFP_ATOMIC); + + if (msg) { + msg->id.idx = CN_KVP_IDX; + msg->id.val = CN_KVP_VAL; + msg->seq = KVP_KERNEL_GET; + ((struct hv_ku_msg *)msg->data)->kvp_index = index; + msg->len = sizeof(struct hv_ku_msg); + cn_netlink_send(msg, 0, GFP_ATOMIC); + kfree(msg); + } + return; +} + +/* + * Send a response back to the host. + */ + +static void +kvp_respond_to_host(char *key, char *value, int error) +{ + struct hv_kvp_msg *kvp_msg; + struct hv_kvp_msg_enumerate *kvp_data; + char *key_name; + struct icmsg_hdr *icmsghdrp; + int keylen, valuelen; + u32 buf_len; + struct vmbus_channel *channel; + u64 req_id; + + /* + * 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; + + if (channel->onchannel_callback == NULL) + /* + * We have raced with util driver being unloaded; + * silently return. + */ + return; + + icmsghdrp = (struct icmsg_hdr *) + &recv_buffer[sizeof(struct vmbuspipe_hdr)]; + kvp_msg = (struct hv_kvp_msg *) + &recv_buffer[sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + kvp_data = &kvp_msg->kvp_data; + key_name = key; + + /* + * If the error parameter is set, terminate the host's enumeration. + */ + if (error) { + /* + * We don't support this index or the we have timedout; + * terminate the host-side iteration by returning an error. + */ + icmsghdrp->status = HV_E_FAIL; + goto response_done; + } + + /* + * The windows host expects the key/value pair to be encoded + * in utf16. + */ + keylen = utf8s_to_utf16s(key_name, strlen(key_name), + (wchar_t *)kvp_data->data.key); + kvp_data->data.key_size = 2*(keylen + 1); /* utf16 encoding */ + valuelen = utf8s_to_utf16s(value, strlen(value), + (wchar_t *)kvp_data->data.value); + kvp_data->data.value_size = 2*(valuelen + 1); /* utf16 encoding */ + + kvp_data->data.value_type = REG_SZ; /* all our values are strings */ + icmsghdrp->status = HV_S_OK; + +response_done: + icmsghdrp->icflags = ICMSGHDRFLAG_TRANSACTION | ICMSGHDRFLAG_RESPONSE; + + vmbus_sendpacket(channel, recv_buffer, buf_len, req_id, + VM_PKT_DATA_INBAND, 0); + +} + +/* + * This callback is invoked when we get a KVP message from the host. + * The host ensures that only one KVP transaction can be active at a time. + * KVP implementation in Linux needs to forward the key to a user-mde + * component to 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 hv_kvp_msg_enumerate *kvp_data; + + struct icmsg_hdr *icmsghdrp; + struct icmsg_negotiate *negop = NULL; + + + vmbus_recvpacket(channel, recv_buffer, PAGE_SIZE, &recvlen, &requestid); + + if (recvlen > 0) { + icmsghdrp = (struct icmsg_hdr *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr)]; + + if (icmsghdrp->icmsgtype == ICMSGTYPE_NEGOTIATE) { + prep_negotiate_resp(icmsghdrp, negop, recv_buffer); + } else { + kvp_msg = (struct hv_kvp_msg *)&recv_buffer[ + sizeof(struct vmbuspipe_hdr) + + sizeof(struct icmsg_hdr)]; + + kvp_data = &kvp_msg->kvp_data; + + /* + * We only support the "get" operation on + * "KVP_POOL_AUTO" pool. + */ + + if ((kvp_msg->kvp_hdr.pool != KVP_POOL_AUTO) || + (kvp_msg->kvp_hdr.operation != + KVP_OP_ENUMERATE)) { + icmsghdrp->status = HV_E_FAIL; + goto callback_done; + } + + /* + * 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.index = kvp_data->index; + + /* + * 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; + + } + +callback_done: + + 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; + + 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/drivers/hv/hv_kvp.h b/drivers/hv/hv_kvp.h new file mode 100644 index 000000000000..9b765d7df838 --- /dev/null +++ b/drivers/hv/hv_kvp.h @@ -0,0 +1,184 @@ +/* + * An implementation of HyperV 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. + * + */ +#ifndef _KVP_H +#define _KVP_H_ + +/* + * Maximum value size - used for both key names and value data, and includes + * any applicable NULL terminators. + * + * Note: This limit is somewhat arbitrary, but falls easily within what is + * supported for all native guests (back to Win 2000) and what is reasonable + * for the IC KVP exchange functionality. Note that Windows Me/98/95 are + * limited to 255 character key names. + * + * MSDN recommends not storing data values larger than 2048 bytes in the + * registry. + * + * Note: This value is used in defining the KVP exchange message - this value + * cannot be modified without affecting the message size and compatibility. + */ + +/* + * bytes, including any null terminators + */ +#define HV_KVP_EXCHANGE_MAX_VALUE_SIZE (2048) + + +/* + * Maximum key size - the registry limit for the length of an entry name + * is 256 characters, including the null terminator + */ + +#define HV_KVP_EXCHANGE_MAX_KEY_SIZE (512) + +/* + * In Linux, we implement the KVP functionality in two components: + * 1) The kernel component which is packaged as part of the hv_utils driver + * is responsible for communicating with the host and responsible for + * implementing the host/guest protocol. 2) A user level daemon that is + * responsible for data gathering. + * + * Host/Guest Protocol: The host iterates over an index and expects the guest + * to assign a key name to the index and also return the value corresponding to + * the key. The host will have atmost one KVP transaction outstanding at any + * given point in time. The host side iteration stops when the guest returns + * an error. Microsoft has specified the following mapping of key names to + * host specified index: + * + * Index Key Name + * 0 FullyQualifiedDomainName + * 1 IntegrationServicesVersion + * 2 NetworkAddressIPv4 + * 3 NetworkAddressIPv6 + * 4 OSBuildNumber + * 5 OSName + * 6 OSMajorVersion + * 7 OSMinorVersion + * 8 OSVersion + * 9 ProcessorArchitecture + * + * The Windows host expects the Key Name and Key Value to be encoded in utf16. + * + * Guest Kernel/KVP Daemon Protocol: As noted earlier, we implement all of the + * data gathering functionality in a user mode daemon. The user level daemon + * is also responsible for binding the key name to the index as well. The + * kernel and user-level daemon communicate using a connector channel. + * + * The user mode component first registers with the + * the kernel component. Subsequently, the kernel component requests, data + * for the specified keys. In response to this message the user mode component + * fills in the value corresponding to the specified key. We overload the + * sequence field in the cn_msg header to define our KVP message types. + * + * + * The kernel component simply acts as a conduit for communication between the + * Windows host and the user-level daemon. The kernel component passes up the + * index received from the Host to the user-level daemon. If the index is + * valid (supported), the corresponding key as well as its + * value (both are strings) is returned. If the index is invalid + * (not supported), a NULL key string is returned. + */ + +/* + * + * The following definitions are shared with the user-mode component; do not + * change any of this without making the corresponding changes in + * the KVP user-mode component. + */ + +#define CN_KVP_VAL 0x1 /* This supports queries from the kernel */ +#define CN_KVP_USER_VAL 0x2 /* This supports queries from the user */ + +enum hv_ku_op { + KVP_REGISTER = 0, /* Register the user mode component */ + KVP_KERNEL_GET, /* Kernel is requesting the value */ + KVP_KERNEL_SET, /* Kernel is providing the value */ + KVP_USER_GET, /* User is requesting the value */ + KVP_USER_SET /* User is providing the value */ +}; + +struct hv_ku_msg { + __u32 kvp_index; /* Key index */ + __u8 kvp_key[HV_KVP_EXCHANGE_MAX_KEY_SIZE]; /* Key name */ + __u8 kvp_value[HV_KVP_EXCHANGE_MAX_VALUE_SIZE]; /* Key value */ +}; + + + + +#ifdef __KERNEL__ + +/* + * Registry value types. + */ + +#define REG_SZ 1 + +enum hv_kvp_exchg_op { + KVP_OP_GET = 0, + KVP_OP_SET, + KVP_OP_DELETE, + KVP_OP_ENUMERATE, + KVP_OP_COUNT /* Number of operations, must be last. */ +}; + +enum hv_kvp_exchg_pool { + KVP_POOL_EXTERNAL = 0, + KVP_POOL_GUEST, + KVP_POOL_AUTO, + KVP_POOL_AUTO_EXTERNAL, + KVP_POOL_AUTO_INTERNAL, + KVP_POOL_COUNT /* Number of pools, must be last. */ +}; + +struct hv_kvp_hdr { + u8 operation; + u8 pool; +}; + +struct hv_kvp_exchg_msg_value { + u32 value_type; + u32 key_size; + u32 value_size; + u8 key[HV_KVP_EXCHANGE_MAX_KEY_SIZE]; + u8 value[HV_KVP_EXCHANGE_MAX_VALUE_SIZE]; +}; + +struct hv_kvp_msg_enumerate { + u32 index; + struct hv_kvp_exchg_msg_value data; +}; + +struct hv_kvp_msg { + struct hv_kvp_hdr kvp_hdr; + struct hv_kvp_msg_enumerate kvp_data; +}; + +int hv_kvp_init(struct hv_util_service *); +void hv_kvp_deinit(void); +void hv_kvp_onchannelcallback(void *); + +#endif /* __KERNEL__ */ +#endif /* _KVP_H */ + diff --git a/drivers/hv/hv_util.c b/drivers/hv/hv_util.c new file mode 100644 index 000000000000..e0e3a6d0244d --- /dev/null +++ b/drivers/hv/hv_util.c @@ -0,0 +1,354 @@ +/* + * 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 "hv_kvp.h" + + +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 void shutdown_onchannelcallback(void *context) +{ + struct vmbus_channel *channel = context; + u32 recvlen; + u64 requestid; + u8 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) { + prep_negotiate_resp(icmsghdrp, negop, shut_txf_buf); + } 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) + orderly_poweroff(true); +} + +/* + * 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; + + 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) { + prep_negotiate_resp(icmsghdrp, NULL, time_txf_buf); + } 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; + + 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) { + prep_negotiate_resp(icmsghdrp, NULL, hbeat_txf_buf); + } 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, GFP_KERNEL); + if (!srv->recv_buffer) + return -ENOMEM; + if (srv->util_init) { + ret = srv->util_init(srv); + if (ret) { + ret = -ENODEV; + goto error1; + } + } + + ret = vmbus_open(dev->channel, 2 * PAGE_SIZE, 2 * PAGE_SIZE, NULL, 0, + srv->util_cb, dev->channel); + if (ret) + goto error; + + hv_set_drvdata(dev, srv); + 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); + + vmbus_close(dev->channel); + if (srv->util_deinit) + srv->util_deinit(); + kfree(srv->recv_buffer); + + return 0; +} + +static const struct hv_vmbus_device_id id_table[] = { + /* Shutdown guid */ + { VMBUS_DEVICE(0x31, 0x60, 0x0B, 0X0E, 0x13, 0x52, 0x34, 0x49, + 0x81, 0x8B, 0x38, 0XD9, 0x0C, 0xED, 0x39, 0xDB) + .driver_data = (unsigned long)&util_shutdown }, + /* Time synch guid */ + { VMBUS_DEVICE(0x30, 0xe6, 0x27, 0x95, 0xae, 0xd0, 0x7b, 0x49, + 0xad, 0xce, 0xe8, 0x0a, 0xb0, 0x17, 0x5c, 0xaf) + .driver_data = (unsigned long)&util_timesynch }, + /* Heartbeat guid */ + { VMBUS_DEVICE(0x39, 0x4f, 0x16, 0x57, 0x15, 0x91, 0x78, 0x4e, + 0xab, 0x55, 0x38, 0x2f, 0x3b, 0xd5, 0x42, 0x2d) + .driver_data = (unsigned long)&util_heartbeat }, + /* KVP guid */ + { VMBUS_DEVICE(0xe7, 0xf4, 0xa0, 0xa9, 0x45, 0x5a, 0x96, 0x4d, + 0xb8, 0x27, 0x8a, 0x84, 0x1e, 0x8c, 0x3, 0xe6) + .driver_data = (unsigned long)&util_kvp }, + { }, +}; + +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_VERSION(HV_DRV_VERSION); +MODULE_LICENSE("GPL"); diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h new file mode 100644 index 000000000000..8261cb64931b --- /dev/null +++ b/drivers/hv/hyperv_vmbus.h @@ -0,0 +1,628 @@ +/* + * + * 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 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 connection identifier type. */ +union hv_connection_id { + u32 asu32; + struct { + u32 id:24; + u32 reserved:8; + } u; +}; + +/* 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; + }; +}; + +/* 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]; +}; + +/* Definition of the hv_signal_event hypercall input structure. */ +struct hv_input_signal_event { + union hv_connection_id connectionid; + u16 flag_number; + u16 rsvdz; +}; + +/* + * 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 + +#define HV_LINUX_GUEST_ID_LO 0x00000000 +#define HV_LINUX_GUEST_ID_HI 0xB16B00B5 +#define HV_LINUX_GUEST_ID (((u64)HV_LINUX_GUEST_ID_HI << 32) | \ + HV_LINUX_GUEST_ID_LO) + +#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 + }, +}; + +#define MAX_NUM_CPUS 32 + + +struct hv_input_signal_event_buffer { + u64 align8; + struct hv_input_signal_event event; +}; + +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; + + /* + * This is used as an input param to HvCallSignalEvent hypercall. The + * input param is immutable in our usage and must be dynamic mem (vs + * stack or global). */ + struct hv_input_signal_event_buffer *signal_event_buffer; + /* 8-bytes aligned of the buffer above */ + struct hv_input_signal_event *signal_event_param; + + void *synic_message_page[MAX_NUM_CPUS]; + void *synic_event_page[MAX_NUM_CPUS]; +}; + +extern struct hv_context hv_context; + + +/* Hv Interface */ + +extern int hv_init(void); + +extern void hv_cleanup(void); + +extern u16 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); + +extern void hv_synic_init(void *irqarg); + +extern void hv_synic_cleanup(void *arg); + + +/* 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 scatterlist *sglist, + u32 sgcount); + +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); + +u32 hv_get_ringbuffer_interrupt_mask(struct hv_ring_buffer_info *ring_info); + +void hv_dump_ring_info(struct hv_ring_buffer_info *ring_info, char *prefix); + +void hv_ringbuffer_get_debuginfo(struct hv_ring_buffer_info *ring_info, + struct hv_ring_buffer_debug_info *debug_info); + +/* + * 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 + */ + void *monitor_pages; + 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; + +/* General vmbus interface */ + +struct hv_device *vmbus_device_create(uuid_le *type, + 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); + + +/* Connection interface */ + +int vmbus_connect(void); + +int vmbus_post_msg(void *buffer, size_t buflen); + +int vmbus_set_event(u32 child_relid); + +void vmbus_on_event(unsigned long data); + + +#endif /* _HYPERV_VMBUS_H */ diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c new file mode 100644 index 000000000000..f594ed09d7e0 --- /dev/null +++ b/drivers/hv/ring_buffer.c @@ -0,0 +1,527 @@ +/* + * + * 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 "hyperv_vmbus.h" + + +/* #defines */ + + +/* Amount of space to write to */ +#define BYTES_AVAIL_TO_WRITE(r, w, z) \ + ((w) >= (r)) ? ((z) - ((w) - (r))) : ((r) - (w)) + + +/* + * + * hv_get_ringbuffer_availbytes() + * + * Get number of bytes available to read and to write to + * for the specified ring buffer + */ +static inline void +hv_get_ringbuffer_availbytes(struct hv_ring_buffer_info *rbi, + u32 *read, u32 *write) +{ + u32 read_loc, write_loc; + + smp_read_barrier_depends(); + + /* Capture the read/write indices before they changed */ + read_loc = rbi->ring_buffer->read_index; + write_loc = rbi->ring_buffer->write_index; + + *write = BYTES_AVAIL_TO_WRITE(read_loc, write_loc, rbi->ring_datasize); + *read = rbi->ring_datasize - *write; +} + +/* + * 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_dump_ring_info() + * + * Dump out to console the ring buffer info + * + */ +void hv_dump_ring_info(struct hv_ring_buffer_info *ring_info, char *prefix) +{ + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + + hv_get_ringbuffer_availbytes(ring_info, + &bytes_avail_toread, + &bytes_avail_towrite); + + DPRINT(VMBUS, + DEBUG_RING_LVL, + "%s <<ringinfo %p buffer %p avail write %u " + "avail read %u read idx %u write idx %u>>", + prefix, + ring_info, + ring_info->ring_buffer->buffer, + bytes_avail_towrite, + bytes_avail_toread, + ring_info->ring_buffer->read_index, + ring_info->ring_buffer->write_index); +} + + +/* + * + * 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_get_ringbuffer_interrupt_mask() + * + * Get the interrupt mask for the specified ring buffer + * + */ +u32 hv_get_ringbuffer_interrupt_mask(struct hv_ring_buffer_info *rbi) +{ + return rbi->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; + + 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 scatterlist *sglist, u32 sgcount) +{ + int i = 0; + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + u32 totalbytes_towrite = 0; + + struct scatterlist *sg; + u32 next_write_location; + u64 prev_indices = 0; + unsigned long flags; + + for_each_sg(sglist, sg, sgcount, i) + { + totalbytes_towrite += sg->length; + } + + 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); + + for_each_sg(sglist, sg, sgcount, i) + { + next_write_location = hv_copyto_ringbuffer(outring_info, + next_write_location, + sg_virt(sg), + sg->length); + } + + /* 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)); + + /* Make sure we flush all writes before updating the writeIndex */ + smp_wmb(); + + /* Now, update the write location */ + hv_set_next_write_location(outring_info, next_write_location); + + + spin_unlock_irqrestore(&outring_info->ring_lock, flags); + 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) +{ + u32 bytes_avail_towrite; + u32 bytes_avail_toread; + u32 next_read_location = 0; + u64 prev_indices = 0; + unsigned long flags; + + 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); + + /* 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 */ + smp_mb(); + + /* Update the read index */ + hv_set_next_read_location(inring_info, next_read_location); + + spin_unlock_irqrestore(&inring_info->ring_lock, flags); + + return 0; +} diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c new file mode 100644 index 000000000000..b0d08f980de1 --- /dev/null +++ b/drivers/hv/vmbus_drv.c @@ -0,0 +1,772 @@ +/* + * 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/irq.h> +#include <linux/interrupt.h> +#include <linux/sysctl.h> +#include <linux/slab.h> +#include <linux/acpi.h> +#include <acpi/acpi_bus.h> +#include <linux/completion.h> +#include <linux/hyperv.h> + +#include "hyperv_vmbus.h" + + +static struct acpi_device *hv_acpi_dev; + +static struct tasklet_struct msg_dpc; +static struct tasklet_struct event_dpc; + +unsigned int vmbus_loglevel = (ALL_MODULES << 16 | INFO_LVL); +EXPORT_SYMBOL(vmbus_loglevel); + +static struct completion probe_event; +static int irq; + +static void get_channel_info(struct hv_device *device, + struct hv_device_info *info) +{ + struct vmbus_channel_debug_info debug_info; + + if (!device->channel) + return; + + vmbus_get_debug_info(device->channel, &debug_info); + + info->chn_id = debug_info.relid; + info->chn_state = debug_info.state; + memcpy(&info->chn_type, &debug_info.interfacetype, + sizeof(uuid_le)); + memcpy(&info->chn_instance, &debug_info.interface_instance, + sizeof(uuid_le)); + + info->monitor_id = debug_info.monitorid; + + info->server_monitor_pending = debug_info.servermonitor_pending; + info->server_monitor_latency = debug_info.servermonitor_latency; + info->server_monitor_conn_id = debug_info.servermonitor_connectionid; + + info->client_monitor_pending = debug_info.clientmonitor_pending; + info->client_monitor_latency = debug_info.clientmonitor_latency; + info->client_monitor_conn_id = debug_info.clientmonitor_connectionid; + + info->inbound.int_mask = debug_info.inbound.current_interrupt_mask; + info->inbound.read_idx = debug_info.inbound.current_read_index; + info->inbound.write_idx = debug_info.inbound.current_write_index; + info->inbound.bytes_avail_toread = + debug_info.inbound.bytes_avail_toread; + info->inbound.bytes_avail_towrite = + debug_info.inbound.bytes_avail_towrite; + + info->outbound.int_mask = + debug_info.outbound.current_interrupt_mask; + info->outbound.read_idx = debug_info.outbound.current_read_index; + info->outbound.write_idx = debug_info.outbound.current_write_index; + info->outbound.bytes_avail_toread = + debug_info.outbound.bytes_avail_toread; + info->outbound.bytes_avail_towrite = + debug_info.outbound.bytes_avail_towrite; +} + +#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]); +} + +/* + * vmbus_show_device_attr - Show the device attribute in sysfs. + * + * This is invoked when user does a + * "cat /sys/bus/vmbus/devices/<busdevice>/<attr name>" + */ +static ssize_t vmbus_show_device_attr(struct device *dev, + struct device_attribute *dev_attr, + char *buf) +{ + struct hv_device *hv_dev = device_to_hv_device(dev); + struct hv_device_info *device_info; + char alias_name[VMBUS_ALIAS_LEN + 1]; + int ret = 0; + + device_info = kzalloc(sizeof(struct hv_device_info), GFP_KERNEL); + if (!device_info) + return ret; + + get_channel_info(hv_dev, device_info); + + if (!strcmp(dev_attr->attr.name, "class_id")) { + ret = sprintf(buf, "{%02x%02x%02x%02x-%02x%02x-%02x%02x-" + "%02x%02x%02x%02x%02x%02x%02x%02x}\n", + device_info->chn_type.b[3], + device_info->chn_type.b[2], + device_info->chn_type.b[1], + device_info->chn_type.b[0], + device_info->chn_type.b[5], + device_info->chn_type.b[4], + device_info->chn_type.b[7], + device_info->chn_type.b[6], + device_info->chn_type.b[8], + device_info->chn_type.b[9], + device_info->chn_type.b[10], + device_info->chn_type.b[11], + device_info->chn_type.b[12], + device_info->chn_type.b[13], + device_info->chn_type.b[14], + device_info->chn_type.b[15]); + } else if (!strcmp(dev_attr->attr.name, "device_id")) { + ret = sprintf(buf, "{%02x%02x%02x%02x-%02x%02x-%02x%02x-" + "%02x%02x%02x%02x%02x%02x%02x%02x}\n", + device_info->chn_instance.b[3], + device_info->chn_instance.b[2], + device_info->chn_instance.b[1], + device_info->chn_instance.b[0], + device_info->chn_instance.b[5], + device_info->chn_instance.b[4], + device_info->chn_instance.b[7], + device_info->chn_instance.b[6], + device_info->chn_instance.b[8], + device_info->chn_instance.b[9], + device_info->chn_instance.b[10], + device_info->chn_instance.b[11], + device_info->chn_instance.b[12], + device_info->chn_instance.b[13], + device_info->chn_instance.b[14], + device_info->chn_instance.b[15]); + } else if (!strcmp(dev_attr->attr.name, "modalias")) { + print_alias_name(hv_dev, alias_name); + ret = sprintf(buf, "vmbus:%s\n", alias_name); + } else if (!strcmp(dev_attr->attr.name, "state")) { + ret = sprintf(buf, "%d\n", device_info->chn_state); + } else if (!strcmp(dev_attr->attr.name, "id")) { + ret = sprintf(buf, "%d\n", device_info->chn_id); + } else if (!strcmp(dev_attr->attr.name, "out_intr_mask")) { + ret = sprintf(buf, "%d\n", device_info->outbound.int_mask); + } else if (!strcmp(dev_attr->attr.name, "out_read_index")) { + ret = sprintf(buf, "%d\n", device_info->outbound.read_idx); + } else if (!strcmp(dev_attr->attr.name, "out_write_index")) { + ret = sprintf(buf, "%d\n", device_info->outbound.write_idx); + } else if (!strcmp(dev_attr->attr.name, "out_read_bytes_avail")) { + ret = sprintf(buf, "%d\n", + device_info->outbound.bytes_avail_toread); + } else if (!strcmp(dev_attr->attr.name, "out_write_bytes_avail")) { + ret = sprintf(buf, "%d\n", + device_info->outbound.bytes_avail_towrite); + } else if (!strcmp(dev_attr->attr.name, "in_intr_mask")) { + ret = sprintf(buf, "%d\n", device_info->inbound.int_mask); + } else if (!strcmp(dev_attr->attr.name, "in_read_index")) { + ret = sprintf(buf, "%d\n", device_info->inbound.read_idx); + } else if (!strcmp(dev_attr->attr.name, "in_write_index")) { + ret = sprintf(buf, "%d\n", device_info->inbound.write_idx); + } else if (!strcmp(dev_attr->attr.name, "in_read_bytes_avail")) { + ret = sprintf(buf, "%d\n", + device_info->inbound.bytes_avail_toread); + } else if (!strcmp(dev_attr->attr.name, "in_write_bytes_avail")) { + ret = sprintf(buf, "%d\n", + device_info->inbound.bytes_avail_towrite); + } else if (!strcmp(dev_attr->attr.name, "monitor_id")) { + ret = sprintf(buf, "%d\n", device_info->monitor_id); + } else if (!strcmp(dev_attr->attr.name, "server_monitor_pending")) { + ret = sprintf(buf, "%d\n", device_info->server_monitor_pending); + } else if (!strcmp(dev_attr->attr.name, "server_monitor_latency")) { + ret = sprintf(buf, "%d\n", device_info->server_monitor_latency); + } else if (!strcmp(dev_attr->attr.name, "server_monitor_conn_id")) { + ret = sprintf(buf, "%d\n", + device_info->server_monitor_conn_id); + } else if (!strcmp(dev_attr->attr.name, "client_monitor_pending")) { + ret = sprintf(buf, "%d\n", device_info->client_monitor_pending); + } else if (!strcmp(dev_attr->attr.name, "client_monitor_latency")) { + ret = sprintf(buf, "%d\n", device_info->client_monitor_latency); + } else if (!strcmp(dev_attr->attr.name, "client_monitor_conn_id")) { + ret = sprintf(buf, "%d\n", + device_info->client_monitor_conn_id); + } + + kfree(device_info); + return ret; +} + +/* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */ +static struct device_attribute vmbus_device_attrs[] = { + __ATTR(id, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(state, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(class_id, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(device_id, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(monitor_id, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(modalias, S_IRUGO, vmbus_show_device_attr, NULL), + + __ATTR(server_monitor_pending, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(server_monitor_latency, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(server_monitor_conn_id, S_IRUGO, vmbus_show_device_attr, NULL), + + __ATTR(client_monitor_pending, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(client_monitor_latency, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(client_monitor_conn_id, S_IRUGO, vmbus_show_device_attr, NULL), + + __ATTR(out_intr_mask, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(out_read_index, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(out_write_index, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(out_read_bytes_avail, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(out_write_bytes_avail, S_IRUGO, vmbus_show_device_attr, NULL), + + __ATTR(in_intr_mask, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(in_read_index, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(in_write_index, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(in_read_bytes_avail, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR(in_write_bytes_avail, S_IRUGO, vmbus_show_device_attr, NULL), + __ATTR_NULL +}; + + +/* + * 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 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, + __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 = drv_to_hv_drv(child_device->driver); + struct hv_device *dev = device_to_hv_device(child_device); + + if (drv->remove) + drv->remove(dev); + else + pr_err("remove not set for driver %s\n", + dev_name(child_device)); + + 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_attrs = vmbus_device_attrs, +}; + +static const char *driver_name = "hyperv"; + + +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; + + ctx = container_of(work, struct onmessage_work_context, + work); + vmbus_onmessage(&ctx->msg); + kfree(ctx); +} + +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 onmessage_work_context *ctx; + + while (1) { + if (msg->header.message_type == HVMSG_NONE) { + /* no msg */ + break; + } else { + 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); + } + + 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 + */ + smp_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 irqreturn_t vmbus_isr(int irq, void *dev_id) +{ + int cpu = smp_processor_id(); + void *page_addr; + struct hv_message *msg; + union hv_synic_event_flags *event; + bool handled = false; + + /* + * 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. + */ + + page_addr = hv_context.synic_event_page[cpu]; + event = (union hv_synic_event_flags *)page_addr + VMBUS_MESSAGE_SINT; + + /* 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; + tasklet_schedule(&event_dpc); + } + + 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) { + handled = true; + tasklet_schedule(&msg_dpc); + } + + if (handled) + return IRQ_HANDLED; + else + return IRQ_NONE; +} + +/* + * 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; + unsigned int vector; + + /* 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); + tasklet_init(&event_dpc, vmbus_on_event, 0); + + ret = bus_register(&hv_bus); + if (ret) + goto err_cleanup; + + ret = request_irq(irq, vmbus_isr, IRQF_SAMPLE_RANDOM, + driver_name, hv_acpi_dev); + + if (ret != 0) { + pr_err("Unable to request IRQ %d\n", + irq); + goto err_unregister; + } + + vector = IRQ0_VECTOR + irq; + + /* + * Notify the hypervisor of our irq and + * connect to the host. + */ + on_each_cpu(hv_synic_init, (void *)&vector, 1); + ret = vmbus_connect(); + if (ret) + goto err_irq; + + vmbus_request_offers(); + + return 0; + +err_irq: + free_irq(irq, hv_acpi_dev); + +err_unregister: + 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); + + 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); + + vmbus_request_offers(); + + 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); + + 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(uuid_le *type, + 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; + + static atomic_t device_num = ATOMIC_INIT(0); + + dev_set_name(&child_device_obj->device, "vmbus_0_%d", + atomic_inc_return(&device_num)); + + 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_info("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) +{ + /* + * Kick off the process of unregistering the device. + * This will call vmbus_remove() and eventually vmbus_device_release() + */ + device_unregister(&device_obj->device); + + pr_info("child device %s unregistered\n", + dev_name(&device_obj->device)); +} + + +/* + * VMBUS is an acpi enumerated device. Get the the IRQ information + * from DSDT. + */ + +static acpi_status vmbus_walk_resources(struct acpi_resource *res, void *irq) +{ + + if (res->type == ACPI_RESOURCE_TYPE_IRQ) { + struct acpi_resource_irq *irqp; + irqp = &res->data.irq; + + *((unsigned int *)irq) = irqp->interrupts[0]; + } + + return AE_OK; +} + +static int vmbus_acpi_add(struct acpi_device *device) +{ + acpi_status result; + + hv_acpi_dev = device; + + result = acpi_walk_resources(device->handle, METHOD_NAME__CRS, + vmbus_walk_resources, &irq); + + if (ACPI_FAILURE(result)) { + complete(&probe_event); + return -ENODEV; + } + complete(&probe_event); + return 0; +} + +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; + + 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); + return ret; +} + + +MODULE_LICENSE("GPL"); +MODULE_VERSION(HV_DRV_VERSION); +module_param(vmbus_loglevel, int, S_IRUGO|S_IWUSR); + +module_init(hv_acpi_init); |