// SPDX-License-Identifier: GPL-2.0-or-later

/* PASST - Plug A Simple Socket Transport
 *  for qemu/UNIX domain socket mode
 *
 * PASTA - Pack A Subtle Tap Abstraction
 *  for network namespace/tap device mode
 *
 * migrate.c - Migration sections, layout, and routines
 *
 * Copyright (c) 2025 Red Hat GmbH
 * Author: Stefano Brivio <sbrivio@redhat.com>
 */

#include <errno.h>
#include <sys/uio.h>

#include "util.h"
#include "ip.h"
#include "passt.h"
#include "inany.h"
#include "flow.h"
#include "flow_table.h"
#include "repair.h"

#include "migrate.h"

/* Current version of migration data */
#define MIGRATE_VERSION		1

/* Magic as we see it and as seen with reverse endianness */
#define MIGRATE_MAGIC		0xB1BB1D1B0BB1D1B0
#define MIGRATE_MAGIC_SWAPPED	0xB0D1B1B01B1DBBB1

/* Migration header to send from source */
static union migrate_header header = {
	/* Immutable part of header structure: keep these two sections at the
	 * beginning, because they are enough to identify a version regardless
	 * of metadata.
	 */
	.magic		= MIGRATE_MAGIC,
	.version	= htonl_constant(MIGRATE_VERSION),
	/* End of immutable part of header structure */

	.time_t_size	= htonl_constant(sizeof(time_t)),
	.flow_size	= htonl_constant(sizeof(union flow)),
	.flow_sidx_size	= htonl_constant(sizeof(struct flow_sidx)),
	.voidp_size	= htonl_constant(sizeof(void *)),
};

/**
 * migrate_send_block() - Migration stage handler to send verbatim data
 * @c:		Execution context
 * @m:		Migration metadata
 * @stage:	Migration stage
 * @fd:		Migration fd
 *
 * Sends the buffer in @stage->iov over the migration channel.
 */
static int migrate_send_block(struct ctx *c, struct migrate_meta *m,
			      const struct migrate_stage *stage, int fd)
{
	(void)c;
	(void)m;

	if (write_remainder(fd, &stage->iov, 1, 0) < 0)
		return errno;

	return 0;
}

/**
 * migrate_recv_block() - Migration stage handler to receive verbatim data
 * @c:		Execution context
 * @m:		Migration metadata
 * @stage:	Migration stage
 * @fd:		Migration fd
 *
 * Reads the buffer in @stage->iov from the migration channel.
 *
 * #syscalls:vu readv
 */
static int migrate_recv_block(struct ctx *c, struct migrate_meta *m,
			      const struct migrate_stage *stage, int fd)
{
	(void)c;
	(void)m;

	if (read_remainder(fd, &stage->iov, 1, 0) < 0)
		return errno;

	return 0;
}

#define DATA_STAGE(v) \
	{					\
		.name = #v,			\
		.source = migrate_send_block,	\
		.target = migrate_recv_block,	\
		.iov = { &(v), sizeof(v) },	\
	}

/* Stages for version 1 */
static const struct migrate_stage stages_v1[] = {
	{
		.name = "flow pre",
		.target = NULL,
	},
	DATA_STAGE(flow_first_free),
	DATA_STAGE(flowtab),
	DATA_STAGE(flow_hashtab),
	{
		.name = "flow post",
		.source = NULL,
	},
};

/* Set of data versions */
static const struct migrate_version versions[] = {
	{
		1,	stages_v1,	ARRAY_SIZE(stages_v1),
	},
};

/**
 * migrate_source() - Migration as source, send state to hypervisor
 * @c:		Execution context
 * @fd:		File descriptor for state transfer
 *
 * Return: 0 on success, positive error code on failure
 */
static int migrate_source(struct ctx *c, int fd)
{
	struct migrate_meta m;
	unsigned i;
	int rc;

	rc = write_all_buf(fd, &header, sizeof(header));
	if (rc) {
		err("Failed to send migration header: %s, abort", strerror_(rc));
		return rc;
	}

	for (i = 0; i < ARRAY_SIZE(stages_v1); i++) {
		const struct migrate_stage *stage = &stages_v1[i];

		if (!stage->source)
			continue;

		debug("Source side migration: %s", stage->name);

		if ((rc = stage->source(c, &m, stage, fd))) {
			err("Source migration failed stage %s: %s, abort",
			    stage->name, strerror_(rc));
			return rc;
		}
	}

	return 0;
}

/**
 * migrate_target_read_header() - Set metadata in target from source header
 * @fd:		Descriptor for state transfer
 * @m:		Migration metadata, filled on return
 *
 * Return: 0 on success, error code on failure
 */
static int migrate_target_read_header(int fd, struct migrate_meta *m)
{
	union migrate_header h;
	unsigned i;

	if (read_all_buf(fd, &h, sizeof(h)))
		return errno;

	debug("Source magic: 0x%016" PRIx64 ", sizeof(void *): %u, version: %u",
	      h.magic, ntohl(h.voidp_size), ntohl(h.version));

	m->version = ntohl(h.version);
	m->v = NULL;
	for (i = 0; i < ARRAY_SIZE(versions); i++) {
		if (versions[i].version == m->version)
			m->v = &versions[i];
	}
	if (!m->v)
		return ENOTSUP;

	if (h.magic == MIGRATE_MAGIC)
		m->bswap = false;
	else if (h.magic == MIGRATE_MAGIC_SWAPPED)
		m->bswap = true;
	else
		return ENOTSUP;

	if (ntohl(h.voidp_size) == 4)
		m->source_64b = false;
	else if (ntohl(h.voidp_size) == 8)
		m->source_64b = true;
	else
		return ENOTSUP;

	if (ntohl(h.time_t_size) == 4)
		m->time_64b = false;
	else if (ntohl(h.time_t_size) == 8)
		m->time_64b = true;
	else
		return ENOTSUP;

	m->flow_size = ntohl(h.flow_size);
	m->flow_sidx_size = ntohl(h.flow_sidx_size);

	return 0;
}

/**
 * migrate_target() - Migration as target, receive state from hypervisor
 * @c:		Execution context
 * @fd:		File descriptor for state transfer
 *
 * Return: 0 on success, positive error code on failure
 */
static int migrate_target(struct ctx *c, int fd)
{
	struct migrate_meta m;
	unsigned i;
	int rc;

	rc = migrate_target_read_header(fd, &m);
	if (rc) {
		err("Migration header check failed: %s, abort", strerror_(rc));
		return rc;
	}

	for (i = 0; i < m.v->nstages; i++) {
		const struct migrate_stage *stage = &m.v->stages[i];

		if (!stage->target)
			continue;

		debug("Target side migration: %s", stage->name);
		
		if ((rc = stage->target(c, &m, stage, fd))) {
			err("Target migration failed stage %s: %s, abort",
			    stage->name, strerror_(rc));
			return rc;
		}
	}
	
	return 0;
}

/**
 * set_migration_watch() - Add the migration file descriptor to epoll
 * @c:		Execution context
 * @fd:		File descriptor to add
 * @target:	Are we the target of the migration?
 */
static void set_migration_watch(const struct ctx *c, int fd, bool target)
{
	union epoll_ref ref = {
		.type = EPOLL_TYPE_DEVICE_STATE,
		.fd = fd,
	 };
	struct epoll_event ev = { 0 };

	ev.data.u64 = ref.u64;
	ev.events = target ? EPOLLIN : EPOLLOUT;

	epoll_ctl(c->epollfd, EPOLL_CTL_ADD, ref.fd, &ev);
}

/**
 * migrate_init() - Set up things necessary for migration
 * @c:		Execution context
 */
void migrate_init(struct ctx *c)
{
	c->device_state_fd = -1;
	c->device_state_result = -1;
}

/**
 * migrate_close() - Close migration channel
 * @c:		Execution context
 */
void migrate_close(struct ctx *c)
{
	if (c->device_state_fd != -1) {
		debug("Closing migration channel, fd: %d", c->device_state_fd);
		epoll_del(c, c->device_state_fd);
		close(c->device_state_fd);
		c->device_state_fd = -1;
		c->device_state_result = -1;
	}
}

/**
 * migrate_request() - Request a migration of device state
 * @c:		Execution context
 * @fd:		fd to transfer state
 * @target:	Are we the target of the migration?
 */
void migrate_request(struct ctx *c, int fd, bool target)
{
	debug("Migration requested, fd: %d (was %d)",
	      fd, c->device_state_fd);

	if (c->device_state_fd != -1)
		migrate_close(c);

	c->device_state_fd = fd;
	set_migration_watch(c, c->device_state_fd, target);

}

/**
 * migrate_handler() - Send/receive passt internal state to/from QEMU
 * @c:		Execution context
 * @events:	epoll events
 */
void migrate_handler(struct ctx *c, uint32_t events)
{
	int rc = EIO;

	debug("migrate_handler fd %d events %x", c->device_state_fd, events);

	if (events & EPOLLOUT)
		rc = migrate_source(c, c->device_state_fd);
	else if (events & EPOLLIN)
		rc = migrate_target(c, c->device_state_fd);

	/* EPOLLHUP without EPOLLIN/EPOLLOUT, or EPOLLERR? Migration failed */

	migrate_close(c);

	c->device_state_result = rc;
}