freebsd-dev/sys/dev/hyperv/vmbus/hv_hv.c
Wei Hu da2f98a1cf Microsoft vmbus, storage and other related driver enhancements for HyperV.
- Vmbus multi channel support.
    - Vector interrupt support.
    - Signal optimization.
    - Storvsc driver performance improvement.
    - Scatter and gather support for storvsc driver.
    - Minor bug fix for KVP driver.
Thanks royger, jhb and delphij from FreeBSD community for the reviews
and comments. Also thanks Hovy Xu from NetApp for the contributions to
the storvsc driver.

PR:     195238
Submitted by:   whu
Reviewed by:    royger, jhb, delphij
Approved by:    royger
MFC after:      2 weeks
Relnotes:       yes
Sponsored by:   Microsoft OSTC
2015-04-29 10:12:34 +00:00

471 lines
12 KiB
C

/*-
* Copyright (c) 2009-2012 Microsoft Corp.
* Copyright (c) 2012 NetApp Inc.
* Copyright (c) 2012 Citrix Inc.
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice unmodified, this list of conditions, and the following
* disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
* IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
* IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/**
* Implements low-level interactions with Hypver-V/Azure
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/malloc.h>
#include <sys/pcpu.h>
#include <sys/timetc.h>
#include <machine/bus.h>
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/pmap.h>
#include "hv_vmbus_priv.h"
#define HV_X64_MSR_GUEST_OS_ID 0x40000000
#define HV_X64_CPUID_MIN 0x40000005
#define HV_X64_CPUID_MAX 0x4000ffff
#define HV_X64_MSR_TIME_REF_COUNT 0x40000020
#define HV_NANOSECONDS_PER_SEC 1000000000L
static u_int hv_get_timecount(struct timecounter *tc);
static inline void do_cpuid_inline(unsigned int op, unsigned int *eax,
unsigned int *ebx, unsigned int *ecx, unsigned int *edx) {
__asm__ __volatile__("cpuid" : "=a" (*eax), "=b" (*ebx), "=c" (*ecx),
"=d" (*edx) : "0" (op), "c" (ecx));
}
/**
* Globals
*/
hv_vmbus_context hv_vmbus_g_context = {
.syn_ic_initialized = FALSE,
.hypercall_page = NULL,
};
static struct timecounter hv_timecounter = {
hv_get_timecount, 0, ~0u, HV_NANOSECONDS_PER_SEC/100, "Hyper-V", HV_NANOSECONDS_PER_SEC/100
};
static u_int
hv_get_timecount(struct timecounter *tc)
{
u_int now = rdmsr(HV_X64_MSR_TIME_REF_COUNT);
return (now);
}
/**
* @brief Query the cpuid for presence of windows hypervisor
*/
int
hv_vmbus_query_hypervisor_presence(void)
{
u_int regs[4];
int hyper_v_detected = 0;
/*
* When Xen is detected and native Xen PV support is enabled,
* ignore Xen's HyperV emulation.
*/
if (vm_guest == VM_GUEST_XEN)
return (0);
do_cpuid(1, regs);
if (regs[2] & 0x80000000) { /* if(a hypervisor is detected) */
/* make sure this really is Hyper-V */
/* we look at the CPUID info */
do_cpuid(HV_X64_MSR_GUEST_OS_ID, regs);
hyper_v_detected =
regs[0] >= HV_X64_CPUID_MIN &&
regs[0] <= HV_X64_CPUID_MAX &&
!memcmp("Microsoft Hv", &regs[1], 12);
}
return (hyper_v_detected);
}
/**
* @brief Get version of the windows hypervisor
*/
static int
hv_vmbus_get_hypervisor_version(void)
{
unsigned int eax;
unsigned int ebx;
unsigned int ecx;
unsigned int edx;
unsigned int maxLeaf;
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 = HV_CPU_ID_FUNCTION_HV_VENDOR_AND_MAX_FUNCTION;
do_cpuid_inline(op, &eax, &ebx, &ecx, &edx);
maxLeaf = eax;
eax = 0;
ebx = 0;
ecx = 0;
edx = 0;
op = HV_CPU_ID_FUNCTION_HV_INTERFACE;
do_cpuid_inline(op, &eax, &ebx, &ecx, &edx);
if (maxLeaf >= HV_CPU_ID_FUNCTION_MS_HV_VERSION) {
eax = 0;
ebx = 0;
ecx = 0;
edx = 0;
op = HV_CPU_ID_FUNCTION_MS_HV_VERSION;
do_cpuid_inline(op, &eax, &ebx, &ecx, &edx);
}
return (maxLeaf);
}
/**
* @brief Invoke the specified hypercall
*/
static uint64_t
hv_vmbus_do_hypercall(uint64_t control, void* input, void* output)
{
#ifdef __x86_64__
uint64_t hv_status = 0;
uint64_t input_address = (input) ? hv_get_phys_addr(input) : 0;
uint64_t output_address = (output) ? hv_get_phys_addr(output) : 0;
volatile void* hypercall_page = hv_vmbus_g_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
uint32_t control_high = control >> 32;
uint32_t control_low = control & 0xFFFFFFFF;
uint32_t hv_status_high = 1;
uint32_t hv_status_low = 1;
uint64_t input_address = (input) ? hv_get_phys_addr(input) : 0;
uint32_t input_address_high = input_address >> 32;
uint32_t input_address_low = input_address & 0xFFFFFFFF;
uint64_t output_address = (output) ? hv_get_phys_addr(output) : 0;
uint32_t output_address_high = output_address >> 32;
uint32_t output_address_low = output_address & 0xFFFFFFFF;
volatile void* hypercall_page = hv_vmbus_g_context.hypercall_page;
__asm__ __volatile__ ("call *%8" : "=d"(hv_status_high),
"=a"(hv_status_low) : "d" (control_high),
"a" (control_low), "b" (input_address_high),
"c" (input_address_low),
"D"(output_address_high),
"S"(output_address_low), "m" (hypercall_page));
return (hv_status_low | ((uint64_t)hv_status_high << 32));
#endif /* __x86_64__ */
}
/**
* @brief Main initialization routine.
*
* This routine must be called
* before any other routines in here are called
*/
int
hv_vmbus_init(void)
{
int max_leaf;
hv_vmbus_x64_msr_hypercall_contents hypercall_msr;
void* virt_addr = 0;
memset(
hv_vmbus_g_context.syn_ic_event_page,
0,
sizeof(hv_vmbus_handle) * MAXCPU);
memset(
hv_vmbus_g_context.syn_ic_msg_page,
0,
sizeof(hv_vmbus_handle) * MAXCPU);
if (vm_guest != VM_GUEST_HV)
goto cleanup;
max_leaf = hv_vmbus_get_hypervisor_version();
/*
* Write our OS info
*/
uint64_t os_guest_info = HV_FREEBSD_GUEST_ID;
wrmsr(HV_X64_MSR_GUEST_OS_ID, os_guest_info);
hv_vmbus_g_context.guest_id = os_guest_info;
/*
* See if the hypercall page is already set
*/
hypercall_msr.as_uint64_t = rdmsr(HV_X64_MSR_HYPERCALL);
virt_addr = malloc(PAGE_SIZE, M_DEVBUF, M_NOWAIT | M_ZERO);
KASSERT(virt_addr != NULL,
("Error VMBUS: malloc failed to allocate page during init!"));
if (virt_addr == NULL)
goto cleanup;
hypercall_msr.u.enable = 1;
hypercall_msr.u.guest_physical_address =
(hv_get_phys_addr(virt_addr) >> PAGE_SHIFT);
wrmsr(HV_X64_MSR_HYPERCALL, hypercall_msr.as_uint64_t);
/*
* Confirm that hypercall page did get set up
*/
hypercall_msr.as_uint64_t = 0;
hypercall_msr.as_uint64_t = rdmsr(HV_X64_MSR_HYPERCALL);
if (!hypercall_msr.u.enable)
goto cleanup;
hv_vmbus_g_context.hypercall_page = virt_addr;
tc_init(&hv_timecounter); /* register virtual timecount */
return (0);
cleanup:
if (virt_addr != NULL) {
if (hypercall_msr.u.enable) {
hypercall_msr.as_uint64_t = 0;
wrmsr(HV_X64_MSR_HYPERCALL,
hypercall_msr.as_uint64_t);
}
free(virt_addr, M_DEVBUF);
}
return (ENOTSUP);
}
/**
* @brief Cleanup routine, called normally during driver unloading or exiting
*/
void
hv_vmbus_cleanup(void)
{
hv_vmbus_x64_msr_hypercall_contents hypercall_msr;
if (hv_vmbus_g_context.guest_id == HV_FREEBSD_GUEST_ID) {
if (hv_vmbus_g_context.hypercall_page != NULL) {
hypercall_msr.as_uint64_t = 0;
wrmsr(HV_X64_MSR_HYPERCALL,
hypercall_msr.as_uint64_t);
free(hv_vmbus_g_context.hypercall_page, M_DEVBUF);
hv_vmbus_g_context.hypercall_page = NULL;
}
}
}
/**
* @brief Post a message using the hypervisor message IPC.
* (This involves a hypercall.)
*/
hv_vmbus_status
hv_vmbus_post_msg_via_msg_ipc(
hv_vmbus_connection_id connection_id,
hv_vmbus_msg_type message_type,
void* payload,
size_t payload_size)
{
struct alignedinput {
uint64_t alignment8;
hv_vmbus_input_post_message msg;
};
hv_vmbus_input_post_message* aligned_msg;
hv_vmbus_status status;
size_t addr;
if (payload_size > HV_MESSAGE_PAYLOAD_BYTE_COUNT)
return (EMSGSIZE);
addr = (size_t) malloc(sizeof(struct alignedinput), M_DEVBUF,
M_ZERO | M_NOWAIT);
KASSERT(addr != 0,
("Error VMBUS: malloc failed to allocate message buffer!"));
if (addr == 0)
return (ENOMEM);
aligned_msg = (hv_vmbus_input_post_message*)
(HV_ALIGN_UP(addr, HV_HYPERCALL_PARAM_ALIGN));
aligned_msg->connection_id = connection_id;
aligned_msg->message_type = message_type;
aligned_msg->payload_size = payload_size;
memcpy((void*) aligned_msg->payload, payload, payload_size);
status = hv_vmbus_do_hypercall(
HV_CALL_POST_MESSAGE, aligned_msg, 0) & 0xFFFF;
free((void *) addr, M_DEVBUF);
return (status);
}
/**
* @brief Signal an event on the specified connection using the hypervisor
* event IPC. (This involves a hypercall.)
*/
hv_vmbus_status
hv_vmbus_signal_event(void *con_id)
{
hv_vmbus_status status;
status = hv_vmbus_do_hypercall(
HV_CALL_SIGNAL_EVENT,
con_id,
0) & 0xFFFF;
return (status);
}
/**
* @brief hv_vmbus_synic_init
*/
void
hv_vmbus_synic_init(void *arg)
{
int cpu;
uint64_t hv_vcpu_index;
hv_vmbus_synic_simp simp;
hv_vmbus_synic_siefp siefp;
hv_vmbus_synic_scontrol sctrl;
hv_vmbus_synic_sint shared_sint;
uint64_t version;
hv_setup_args* setup_args = (hv_setup_args *)arg;
cpu = PCPU_GET(cpuid);
if (hv_vmbus_g_context.hypercall_page == NULL)
return;
/*
* TODO: Check the version
*/
version = rdmsr(HV_X64_MSR_SVERSION);
hv_vmbus_g_context.syn_ic_msg_page[cpu] =
setup_args->page_buffers[2 * cpu];
hv_vmbus_g_context.syn_ic_event_page[cpu] =
setup_args->page_buffers[2 * cpu + 1];
/*
* Setup the Synic's message page
*/
simp.as_uint64_t = rdmsr(HV_X64_MSR_SIMP);
simp.u.simp_enabled = 1;
simp.u.base_simp_gpa = ((hv_get_phys_addr(
hv_vmbus_g_context.syn_ic_msg_page[cpu])) >> PAGE_SHIFT);
wrmsr(HV_X64_MSR_SIMP, simp.as_uint64_t);
/*
* Setup the Synic's event page
*/
siefp.as_uint64_t = rdmsr(HV_X64_MSR_SIEFP);
siefp.u.siefp_enabled = 1;
siefp.u.base_siefp_gpa = ((hv_get_phys_addr(
hv_vmbus_g_context.syn_ic_event_page[cpu])) >> PAGE_SHIFT);
wrmsr(HV_X64_MSR_SIEFP, siefp.as_uint64_t);
/*HV_SHARED_SINT_IDT_VECTOR + 0x20; */
shared_sint.as_uint64_t = 0;
shared_sint.u.vector = setup_args->vector;
shared_sint.u.masked = FALSE;
shared_sint.u.auto_eoi = TRUE;
wrmsr(HV_X64_MSR_SINT0 + HV_VMBUS_MESSAGE_SINT,
shared_sint.as_uint64_t);
/* Enable the global synic bit */
sctrl.as_uint64_t = rdmsr(HV_X64_MSR_SCONTROL);
sctrl.u.enable = 1;
wrmsr(HV_X64_MSR_SCONTROL, sctrl.as_uint64_t);
hv_vmbus_g_context.syn_ic_initialized = TRUE;
/*
* Set up the cpuid mapping from Hyper-V to FreeBSD.
* The array is indexed using FreeBSD cpuid.
*/
hv_vcpu_index = rdmsr(HV_X64_MSR_VP_INDEX);
hv_vmbus_g_context.hv_vcpu_index[cpu] = (uint32_t)hv_vcpu_index;
return;
}
/**
* @brief Cleanup routine for hv_vmbus_synic_init()
*/
void hv_vmbus_synic_cleanup(void *arg)
{
hv_vmbus_synic_sint shared_sint;
hv_vmbus_synic_simp simp;
hv_vmbus_synic_siefp siefp;
if (!hv_vmbus_g_context.syn_ic_initialized)
return;
shared_sint.as_uint64_t = rdmsr(
HV_X64_MSR_SINT0 + HV_VMBUS_MESSAGE_SINT);
shared_sint.u.masked = 1;
/*
* Disable the interrupt
*/
wrmsr(
HV_X64_MSR_SINT0 + HV_VMBUS_MESSAGE_SINT,
shared_sint.as_uint64_t);
simp.as_uint64_t = rdmsr(HV_X64_MSR_SIMP);
simp.u.simp_enabled = 0;
simp.u.base_simp_gpa = 0;
wrmsr(HV_X64_MSR_SIMP, simp.as_uint64_t);
siefp.as_uint64_t = rdmsr(HV_X64_MSR_SIEFP);
siefp.u.siefp_enabled = 0;
siefp.u.base_siefp_gpa = 0;
wrmsr(HV_X64_MSR_SIEFP, siefp.as_uint64_t);
}