freebsd-dev/sys/amd64/vmm/io/vatpit.c
John Baldwin 483d953a86 Initial support for bhyve save and restore.
Save and restore (also known as suspend and resume) permits a snapshot
to be taken of a guest's state that can later be resumed.  In the
current implementation, bhyve(8) creates a UNIX domain socket that is
used by bhyvectl(8) to send a request to save a snapshot (and
optionally exit after the snapshot has been taken).  A snapshot
currently consists of two files: the first holds a copy of guest RAM,
and the second file holds other guest state such as vCPU register
values and device model state.

To resume a guest, bhyve(8) must be started with a matching pair of
command line arguments to instantiate the same set of device models as
well as a pointer to the saved snapshot.

While the current implementation is useful for several uses cases, it
has a few limitations.  The file format for saving the guest state is
tied to the ABI of internal bhyve structures and is not
self-describing (in that it does not communicate the set of device
models present in the system).  In addition, the state saved for some
device models closely matches the internal data structures which might
prove a challenge for compatibility of snapshot files across a range
of bhyve versions.  The file format also does not currently support
versioning of individual chunks of state.  As a result, the current
file format is not a fixed binary format and future revisions to save
and restore will break binary compatiblity of snapshot files.  The
goal is to move to a more flexible format that adds versioning,
etc. and at that point to commit to providing a reasonable level of
compatibility.  As a result, the current implementation is not enabled
by default.  It can be enabled via the WITH_BHYVE_SNAPSHOT=yes option
for userland builds, and the kernel option BHYVE_SHAPSHOT.

Submitted by:	Mihai Tiganus, Flavius Anton, Darius Mihai
Submitted by:	Elena Mihailescu, Mihai Carabas, Sergiu Weisz
Relnotes:	yes
Sponsored by:	University Politehnica of Bucharest
Sponsored by:	Matthew Grooms (student scholarships)
Sponsored by:	iXsystems
Differential Revision:	https://reviews.freebsd.org/D19495
2020-05-05 00:02:04 +00:00

517 lines
12 KiB
C

/*-
* Copyright (c) 2014 Tycho Nightingale <tycho.nightingale@pluribusnetworks.com>
* Copyright (c) 2011 NetApp, Inc.
* All rights reserved.
* Copyright (c) 2018 Joyent, Inc.
*
* 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, 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 NETAPP, INC ``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 NETAPP, INC OR CONTRIBUTORS 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_bhyve_snapshot.h"
#include <sys/param.h>
#include <sys/types.h>
#include <sys/queue.h>
#include <sys/kernel.h>
#include <sys/lock.h>
#include <sys/malloc.h>
#include <sys/mutex.h>
#include <sys/systm.h>
#include <machine/vmm.h>
#include <machine/vmm_snapshot.h>
#include "vmm_ktr.h"
#include "vatpic.h"
#include "vioapic.h"
#include "vatpit.h"
static MALLOC_DEFINE(M_VATPIT, "atpit", "bhyve virtual atpit (8254)");
#define VATPIT_LOCK(vatpit) mtx_lock_spin(&((vatpit)->mtx))
#define VATPIT_UNLOCK(vatpit) mtx_unlock_spin(&((vatpit)->mtx))
#define VATPIT_LOCKED(vatpit) mtx_owned(&((vatpit)->mtx))
#define TIMER_SEL_MASK 0xc0
#define TIMER_RW_MASK 0x30
#define TIMER_MODE_MASK 0x0f
#define TIMER_SEL_READBACK 0xc0
#define TIMER_STS_OUT 0x80
#define TIMER_STS_NULLCNT 0x40
#define TIMER_RB_LCTR 0x20
#define TIMER_RB_LSTATUS 0x10
#define TIMER_RB_CTR_2 0x08
#define TIMER_RB_CTR_1 0x04
#define TIMER_RB_CTR_0 0x02
#define TMR2_OUT_STS 0x20
#define PIT_8254_FREQ 1193182
#define TIMER_DIV(freq, hz) (((freq) + (hz) / 2) / (hz))
struct vatpit_callout_arg {
struct vatpit *vatpit;
int channel_num;
};
struct channel {
int mode;
uint16_t initial; /* initial counter value */
struct bintime now_bt; /* uptime when counter was loaded */
uint8_t cr[2];
uint8_t ol[2];
bool slatched; /* status latched */
uint8_t status;
int crbyte;
int olbyte;
int frbyte;
struct callout callout;
struct bintime callout_bt; /* target time */
struct vatpit_callout_arg callout_arg;
};
struct vatpit {
struct vm *vm;
struct mtx mtx;
struct bintime freq_bt;
struct channel channel[3];
};
static void pit_timer_start_cntr0(struct vatpit *vatpit);
static uint64_t
vatpit_delta_ticks(struct vatpit *vatpit, struct channel *c)
{
struct bintime delta;
uint64_t result;
binuptime(&delta);
bintime_sub(&delta, &c->now_bt);
result = delta.sec * PIT_8254_FREQ;
result += delta.frac / vatpit->freq_bt.frac;
return (result);
}
static int
vatpit_get_out(struct vatpit *vatpit, int channel)
{
struct channel *c;
uint64_t delta_ticks;
int out;
c = &vatpit->channel[channel];
switch (c->mode) {
case TIMER_INTTC:
delta_ticks = vatpit_delta_ticks(vatpit, c);
out = (delta_ticks >= c->initial);
break;
default:
out = 0;
break;
}
return (out);
}
static void
vatpit_callout_handler(void *a)
{
struct vatpit_callout_arg *arg = a;
struct vatpit *vatpit;
struct callout *callout;
struct channel *c;
vatpit = arg->vatpit;
c = &vatpit->channel[arg->channel_num];
callout = &c->callout;
VM_CTR1(vatpit->vm, "atpit t%d fired", arg->channel_num);
VATPIT_LOCK(vatpit);
if (callout_pending(callout)) /* callout was reset */
goto done;
if (!callout_active(callout)) /* callout was stopped */
goto done;
callout_deactivate(callout);
if (c->mode == TIMER_RATEGEN) {
pit_timer_start_cntr0(vatpit);
}
vatpic_pulse_irq(vatpit->vm, 0);
vioapic_pulse_irq(vatpit->vm, 2);
done:
VATPIT_UNLOCK(vatpit);
return;
}
static void
pit_timer_start_cntr0(struct vatpit *vatpit)
{
struct channel *c;
struct bintime now, delta;
sbintime_t precision;
c = &vatpit->channel[0];
if (c->initial != 0) {
delta.sec = 0;
delta.frac = vatpit->freq_bt.frac * c->initial;
bintime_add(&c->callout_bt, &delta);
precision = bttosbt(delta) >> tc_precexp;
/*
* Reset 'callout_bt' if the time that the callout
* was supposed to fire is more than 'c->initial'
* ticks in the past.
*/
binuptime(&now);
if (bintime_cmp(&c->callout_bt, &now, <)) {
c->callout_bt = now;
bintime_add(&c->callout_bt, &delta);
}
callout_reset_sbt(&c->callout, bttosbt(c->callout_bt),
precision, vatpit_callout_handler, &c->callout_arg,
C_ABSOLUTE);
}
}
static uint16_t
pit_update_counter(struct vatpit *vatpit, struct channel *c, bool latch)
{
uint16_t lval;
uint64_t delta_ticks;
/* cannot latch a new value until the old one has been consumed */
if (latch && c->olbyte != 0)
return (0);
if (c->initial == 0) {
/*
* This is possibly an o/s bug - reading the value of
* the timer without having set up the initial value.
*
* The original user-space version of this code set
* the timer to 100hz in this condition; do the same
* here.
*/
c->initial = TIMER_DIV(PIT_8254_FREQ, 100);
binuptime(&c->now_bt);
c->status &= ~TIMER_STS_NULLCNT;
}
delta_ticks = vatpit_delta_ticks(vatpit, c);
lval = c->initial - delta_ticks % c->initial;
if (latch) {
c->olbyte = 2;
c->ol[1] = lval; /* LSB */
c->ol[0] = lval >> 8; /* MSB */
}
return (lval);
}
static int
pit_readback1(struct vatpit *vatpit, int channel, uint8_t cmd)
{
struct channel *c;
c = &vatpit->channel[channel];
/*
* Latch the count/status of the timer if not already latched.
* N.B. that the count/status latch-select bits are active-low.
*/
if (!(cmd & TIMER_RB_LCTR) && !c->olbyte) {
(void) pit_update_counter(vatpit, c, true);
}
if (!(cmd & TIMER_RB_LSTATUS) && !c->slatched) {
c->slatched = true;
/*
* For mode 0, see if the elapsed time is greater
* than the initial value - this results in the
* output pin being set to 1 in the status byte.
*/
if (c->mode == TIMER_INTTC && vatpit_get_out(vatpit, channel))
c->status |= TIMER_STS_OUT;
else
c->status &= ~TIMER_STS_OUT;
}
return (0);
}
static int
pit_readback(struct vatpit *vatpit, uint8_t cmd)
{
int error;
/*
* The readback command can apply to all timers.
*/
error = 0;
if (cmd & TIMER_RB_CTR_0)
error = pit_readback1(vatpit, 0, cmd);
if (!error && cmd & TIMER_RB_CTR_1)
error = pit_readback1(vatpit, 1, cmd);
if (!error && cmd & TIMER_RB_CTR_2)
error = pit_readback1(vatpit, 2, cmd);
return (error);
}
static int
vatpit_update_mode(struct vatpit *vatpit, uint8_t val)
{
struct channel *c;
int sel, rw, mode;
sel = val & TIMER_SEL_MASK;
rw = val & TIMER_RW_MASK;
mode = val & TIMER_MODE_MASK;
if (sel == TIMER_SEL_READBACK)
return (pit_readback(vatpit, val));
if (rw != TIMER_LATCH && rw != TIMER_16BIT)
return (-1);
if (rw != TIMER_LATCH) {
/*
* Counter mode is not affected when issuing a
* latch command.
*/
if (mode != TIMER_INTTC &&
mode != TIMER_RATEGEN &&
mode != TIMER_SQWAVE &&
mode != TIMER_SWSTROBE)
return (-1);
}
c = &vatpit->channel[sel >> 6];
if (rw == TIMER_LATCH)
pit_update_counter(vatpit, c, true);
else {
c->mode = mode;
c->olbyte = 0; /* reset latch after reprogramming */
c->status |= TIMER_STS_NULLCNT;
}
return (0);
}
int
vatpit_handler(struct vm *vm, int vcpuid, bool in, int port, int bytes,
uint32_t *eax)
{
struct vatpit *vatpit;
struct channel *c;
uint8_t val;
int error;
vatpit = vm_atpit(vm);
if (bytes != 1)
return (-1);
val = *eax;
if (port == TIMER_MODE) {
if (in) {
VM_CTR0(vatpit->vm, "vatpit attempt to read mode");
return (-1);
}
VATPIT_LOCK(vatpit);
error = vatpit_update_mode(vatpit, val);
VATPIT_UNLOCK(vatpit);
return (error);
}
/* counter ports */
KASSERT(port >= TIMER_CNTR0 && port <= TIMER_CNTR2,
("invalid port 0x%x", port));
c = &vatpit->channel[port - TIMER_CNTR0];
VATPIT_LOCK(vatpit);
if (in && c->slatched) {
/*
* Return the status byte if latched
*/
*eax = c->status;
c->slatched = false;
c->status = 0;
} else if (in) {
/*
* The spec says that once the output latch is completely
* read it should revert to "following" the counter. Use
* the free running counter for this case (i.e. Linux
* TSC calibration). Assuming the access mode is 16-bit,
* toggle the MSB/LSB bit on each read.
*/
if (c->olbyte == 0) {
uint16_t tmp;
tmp = pit_update_counter(vatpit, c, false);
if (c->frbyte)
tmp >>= 8;
tmp &= 0xff;
*eax = tmp;
c->frbyte ^= 1;
} else
*eax = c->ol[--c->olbyte];
} else {
c->cr[c->crbyte++] = *eax;
if (c->crbyte == 2) {
c->status &= ~TIMER_STS_NULLCNT;
c->frbyte = 0;
c->crbyte = 0;
c->initial = c->cr[0] | (uint16_t)c->cr[1] << 8;
binuptime(&c->now_bt);
/* Start an interval timer for channel 0 */
if (port == TIMER_CNTR0) {
c->callout_bt = c->now_bt;
pit_timer_start_cntr0(vatpit);
}
if (c->initial == 0)
c->initial = 0xffff;
}
}
VATPIT_UNLOCK(vatpit);
return (0);
}
int
vatpit_nmisc_handler(struct vm *vm, int vcpuid, bool in, int port, int bytes,
uint32_t *eax)
{
struct vatpit *vatpit;
vatpit = vm_atpit(vm);
if (in) {
VATPIT_LOCK(vatpit);
if (vatpit_get_out(vatpit, 2))
*eax = TMR2_OUT_STS;
else
*eax = 0;
VATPIT_UNLOCK(vatpit);
}
return (0);
}
struct vatpit *
vatpit_init(struct vm *vm)
{
struct vatpit *vatpit;
struct vatpit_callout_arg *arg;
int i;
vatpit = malloc(sizeof(struct vatpit), M_VATPIT, M_WAITOK | M_ZERO);
vatpit->vm = vm;
mtx_init(&vatpit->mtx, "vatpit lock", NULL, MTX_SPIN);
FREQ2BT(PIT_8254_FREQ, &vatpit->freq_bt);
for (i = 0; i < 3; i++) {
callout_init(&vatpit->channel[i].callout, 1);
arg = &vatpit->channel[i].callout_arg;
arg->vatpit = vatpit;
arg->channel_num = i;
}
return (vatpit);
}
void
vatpit_cleanup(struct vatpit *vatpit)
{
int i;
for (i = 0; i < 3; i++)
callout_drain(&vatpit->channel[i].callout);
free(vatpit, M_VATPIT);
}
#ifdef BHYVE_SNAPSHOT
int
vatpit_snapshot(struct vatpit *vatpit, struct vm_snapshot_meta *meta)
{
int ret;
int i;
struct channel *channel;
SNAPSHOT_VAR_OR_LEAVE(vatpit->freq_bt.sec, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(vatpit->freq_bt.frac, meta, ret, done);
/* properly restore timers; they will NOT work currently */
printf("%s: snapshot restore does not reset timers!\r\n", __func__);
for (i = 0; i < nitems(vatpit->channel); i++) {
channel = &vatpit->channel[i];
SNAPSHOT_VAR_OR_LEAVE(channel->mode, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->initial, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->now_bt.sec, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->now_bt.frac, meta, ret, done);
SNAPSHOT_BUF_OR_LEAVE(channel->cr, sizeof(channel->cr),
meta, ret, done);
SNAPSHOT_BUF_OR_LEAVE(channel->ol, sizeof(channel->ol),
meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->slatched, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->status, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->crbyte, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->frbyte, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->callout_bt.sec, meta, ret, done);
SNAPSHOT_VAR_OR_LEAVE(channel->callout_bt.frac, meta, ret,
done);
}
done:
return (ret);
}
#endif