/**
* qemu_co_queue_run_restart:
*
* Enter each coroutine that was previously marked for restart by
* qemu_co_queue_next() or qemu_co_queue_restart_all(). This function is
* invoked by the core coroutine code when the current coroutine yields or
* terminates.
*/
void qemu_co_queue_run_restart(Coroutine *co)
{
Coroutine *next;
trace_qemu_co_queue_run_restart(co);
while ((next = QSIMPLEQ_FIRST(&co->co_queue_wakeup))) {
QSIMPLEQ_REMOVE_HEAD(&co->co_queue_wakeup, co_queue_next);
qemu_coroutine_enter(next);
}
}
void nbd_export_close(NBDExport *exp)
{
while (!QSIMPLEQ_EMPTY(&exp->requests)) {
NBDRequest *first = QSIMPLEQ_FIRST(&exp->requests);
QSIMPLEQ_REMOVE_HEAD(&exp->requests, entry);
qemu_vfree(first->data);
g_free(first);
}
bdrv_close(exp->bs);
g_free(exp);
}
static void qed_unplug_allocating_write_reqs(BDRVQEDState *s)
{
QEDAIOCB *acb;
assert(s->allocating_write_reqs_plugged);
s->allocating_write_reqs_plugged = false;
acb = QSIMPLEQ_FIRST(&s->allocating_write_reqs);
if (acb) {
qed_aio_next_io(acb, 0);
}
}
bool qemu_co_enter_next(CoQueue *queue)
{
Coroutine *next;
next = QSIMPLEQ_FIRST(&queue->entries);
if (!next) {
return false;
}
QSIMPLEQ_REMOVE_HEAD(&queue->entries, co_queue_next);
qemu_coroutine_enter(next);
return true;
}
static void qed_need_check_timer_cb(void *opaque)
{
BDRVQEDState *s = opaque;
/* The timer should only fire when allocating writes have drained */
assert(!QSIMPLEQ_FIRST(&s->allocating_write_reqs));
trace_qed_need_check_timer_cb(s);
qed_plug_allocating_write_reqs(s);
/* Ensure writes are on disk before clearing flag */
bdrv_aio_flush(s->bs, qed_clear_need_check, s);
}
static void bitmap_list_free(Qcow2BitmapList *bm_list)
{
Qcow2Bitmap *bm;
if (bm_list == NULL) {
return;
}
while ((bm = QSIMPLEQ_FIRST(bm_list)) != NULL) {
QSIMPLEQ_REMOVE_HEAD(bm_list, entry);
bitmap_free(bm);
}
g_free(bm_list);
}
static NBDRequest *nbd_request_get(NBDClient *client)
{
NBDRequest *req;
NBDExport *exp = client->exp;
assert(client->nb_requests <= MAX_NBD_REQUESTS - 1);
client->nb_requests++;
if (QSIMPLEQ_EMPTY(&exp->requests)) {
req = g_malloc0(sizeof(NBDRequest));
req->data = qemu_blockalign(exp->bs, NBD_BUFFER_SIZE);
} else {
req = QSIMPLEQ_FIRST(&exp->requests);
QSIMPLEQ_REMOVE_HEAD(&exp->requests, entry);
}
nbd_client_get(client);
req->client = client;
return req;
}
static bool qemu_co_queue_do_restart(CoQueue *queue, bool single)
{
Coroutine *self = qemu_coroutine_self();
Coroutine *next;
if (QSIMPLEQ_EMPTY(&queue->entries)) {
return false;
}
while ((next = QSIMPLEQ_FIRST(&queue->entries)) != NULL) {
QSIMPLEQ_REMOVE_HEAD(&queue->entries, co_queue_next);
QSIMPLEQ_INSERT_TAIL(&self->co_queue_wakeup, next, co_queue_next);
trace_qemu_co_queue_next(next);
if (single) {
break;
}
}
return true;
}
static void entropy_available(void *opaque)
{
RndRandom *s = RNG_RANDOM(opaque);
while (!QSIMPLEQ_EMPTY(&s->parent.requests)) {
RngRequest *req = QSIMPLEQ_FIRST(&s->parent.requests);
ssize_t len;
len = read(s->fd, req->data, req->size);
if (len < 0 && errno == EAGAIN) {
return;
}
g_assert(len != -1);
req->receive_entropy(req->opaque, req->data, len);
rng_backend_finalize_request(&s->parent, req);
}
/* We've drained all requests, the fd handler can be reset. */
qemu_set_fd_handler(s->fd, NULL, NULL, NULL);
}
/* Submit async read while handling COW.
* Returns: The number of bytes copied after and including offset,
* excluding any bytes copied prior to offset due to alignment.
* This will be @bytes if no alignment is necessary, or
* (new_end - offset) if tail is rounded up or down due to
* alignment or buffer limit.
*/
static uint64_t mirror_do_read(MirrorBlockJob *s, int64_t offset,
uint64_t bytes)
{
BlockBackend *source = s->common.blk;
int nb_chunks;
uint64_t ret;
MirrorOp *op;
uint64_t max_bytes;
max_bytes = s->granularity * s->max_iov;
/* We can only handle as much as buf_size at a time. */
bytes = MIN(s->buf_size, MIN(max_bytes, bytes));
assert(bytes);
assert(bytes < BDRV_REQUEST_MAX_BYTES);
ret = bytes;
if (s->cow_bitmap) {
ret += mirror_cow_align(s, &offset, &bytes);
}
assert(bytes <= s->buf_size);
/* The offset is granularity-aligned because:
* 1) Caller passes in aligned values;
* 2) mirror_cow_align is used only when target cluster is larger. */
assert(QEMU_IS_ALIGNED(offset, s->granularity));
/* The range is sector-aligned, since bdrv_getlength() rounds up. */
assert(QEMU_IS_ALIGNED(bytes, BDRV_SECTOR_SIZE));
nb_chunks = DIV_ROUND_UP(bytes, s->granularity);
while (s->buf_free_count < nb_chunks) {
trace_mirror_yield_in_flight(s, offset, s->in_flight);
mirror_wait_for_io(s);
}
/* Allocate a MirrorOp that is used as an AIO callback. */
op = g_new(MirrorOp, 1);
op->s = s;
op->offset = offset;
op->bytes = bytes;
/* Now make a QEMUIOVector taking enough granularity-sized chunks
* from s->buf_free.
*/
qemu_iovec_init(&op->qiov, nb_chunks);
while (nb_chunks-- > 0) {
MirrorBuffer *buf = QSIMPLEQ_FIRST(&s->buf_free);
size_t remaining = bytes - op->qiov.size;
QSIMPLEQ_REMOVE_HEAD(&s->buf_free, next);
s->buf_free_count--;
qemu_iovec_add(&op->qiov, buf, MIN(s->granularity, remaining));
}
/* Copy the dirty cluster. */
s->in_flight++;
s->bytes_in_flight += bytes;
trace_mirror_one_iteration(s, offset, bytes);
blk_aio_preadv(source, offset, &op->qiov, 0, mirror_read_complete, op);
return ret;
}
static uint64_t coroutine_fn mirror_iteration(MirrorBlockJob *s)
{
BlockDriverState *source = s->common.bs;
int nb_sectors, sectors_per_chunk, nb_chunks;
int64_t end, sector_num, next_chunk, next_sector, hbitmap_next_sector;
uint64_t delay_ns = 0;
MirrorOp *op;
int pnum;
int64_t ret;
s->sector_num = hbitmap_iter_next(&s->hbi);
if (s->sector_num < 0) {
bdrv_dirty_iter_init(s->dirty_bitmap, &s->hbi);
s->sector_num = hbitmap_iter_next(&s->hbi);
trace_mirror_restart_iter(s, bdrv_get_dirty_count(s->dirty_bitmap));
assert(s->sector_num >= 0);
}
hbitmap_next_sector = s->sector_num;
sector_num = s->sector_num;
sectors_per_chunk = s->granularity >> BDRV_SECTOR_BITS;
end = s->bdev_length / BDRV_SECTOR_SIZE;
/* Extend the QEMUIOVector to include all adjacent blocks that will
* be copied in this operation.
*
* We have to do this if we have no backing file yet in the destination,
* and the cluster size is very large. Then we need to do COW ourselves.
* The first time a cluster is copied, copy it entirely. Note that,
* because both the granularity and the cluster size are powers of two,
* the number of sectors to copy cannot exceed one cluster.
*
* We also want to extend the QEMUIOVector to include more adjacent
* dirty blocks if possible, to limit the number of I/O operations and
* run efficiently even with a small granularity.
*/
nb_chunks = 0;
nb_sectors = 0;
next_sector = sector_num;
next_chunk = sector_num / sectors_per_chunk;
/* Wait for I/O to this cluster (from a previous iteration) to be done. */
while (test_bit(next_chunk, s->in_flight_bitmap)) {
trace_mirror_yield_in_flight(s, sector_num, s->in_flight);
s->waiting_for_io = true;
qemu_coroutine_yield();
s->waiting_for_io = false;
}
do {
int added_sectors, added_chunks;
if (!bdrv_get_dirty(source, s->dirty_bitmap, next_sector) ||
test_bit(next_chunk, s->in_flight_bitmap)) {
assert(nb_sectors > 0);
break;
}
added_sectors = sectors_per_chunk;
if (s->cow_bitmap && !test_bit(next_chunk, s->cow_bitmap)) {
bdrv_round_to_clusters(s->target,
next_sector, added_sectors,
&next_sector, &added_sectors);
/* On the first iteration, the rounding may make us copy
* sectors before the first dirty one.
*/
if (next_sector < sector_num) {
assert(nb_sectors == 0);
sector_num = next_sector;
next_chunk = next_sector / sectors_per_chunk;
}
}
added_sectors = MIN(added_sectors, end - (sector_num + nb_sectors));
added_chunks = (added_sectors + sectors_per_chunk - 1) / sectors_per_chunk;
/* When doing COW, it may happen that there is not enough space for
* a full cluster. Wait if that is the case.
*/
while (nb_chunks == 0 && s->buf_free_count < added_chunks) {
trace_mirror_yield_buf_busy(s, nb_chunks, s->in_flight);
s->waiting_for_io = true;
qemu_coroutine_yield();
s->waiting_for_io = false;
}
if (s->buf_free_count < nb_chunks + added_chunks) {
trace_mirror_break_buf_busy(s, nb_chunks, s->in_flight);
break;
}
if (IOV_MAX < nb_chunks + added_chunks) {
trace_mirror_break_iov_max(s, nb_chunks, added_chunks);
break;
}
/* We have enough free space to copy these sectors. */
bitmap_set(s->in_flight_bitmap, next_chunk, added_chunks);
nb_sectors += added_sectors;
nb_chunks += added_chunks;
next_sector += added_sectors;
next_chunk += added_chunks;
if (!s->synced && s->common.speed) {
delay_ns = ratelimit_calculate_delay(&s->limit, added_sectors);
}
} while (delay_ns == 0 && next_sector < end);
/* Allocate a MirrorOp that is used as an AIO callback. */
op = g_new(MirrorOp, 1);
op->s = s;
op->sector_num = sector_num;
op->nb_sectors = nb_sectors;
/* Now make a QEMUIOVector taking enough granularity-sized chunks
* from s->buf_free.
*/
qemu_iovec_init(&op->qiov, nb_chunks);
next_sector = sector_num;
while (nb_chunks-- > 0) {
MirrorBuffer *buf = QSIMPLEQ_FIRST(&s->buf_free);
size_t remaining = (nb_sectors * BDRV_SECTOR_SIZE) - op->qiov.size;
QSIMPLEQ_REMOVE_HEAD(&s->buf_free, next);
s->buf_free_count--;
qemu_iovec_add(&op->qiov, buf, MIN(s->granularity, remaining));
/* Advance the HBitmapIter in parallel, so that we do not examine
* the same sector twice.
*/
if (next_sector > hbitmap_next_sector
&& bdrv_get_dirty(source, s->dirty_bitmap, next_sector)) {
hbitmap_next_sector = hbitmap_iter_next(&s->hbi);
}
next_sector += sectors_per_chunk;
}
bdrv_reset_dirty_bitmap(s->dirty_bitmap, sector_num, nb_sectors);
/* Copy the dirty cluster. */
s->in_flight++;
s->sectors_in_flight += nb_sectors;
trace_mirror_one_iteration(s, sector_num, nb_sectors);
ret = bdrv_get_block_status_above(source, NULL, sector_num,
nb_sectors, &pnum);
if (ret < 0 || pnum < nb_sectors ||
(ret & BDRV_BLOCK_DATA && !(ret & BDRV_BLOCK_ZERO))) {
bdrv_aio_readv(source, sector_num, &op->qiov, nb_sectors,
mirror_read_complete, op);
} else if (ret & BDRV_BLOCK_ZERO) {
bdrv_aio_write_zeroes(s->target, sector_num, op->nb_sectors,
s->unmap ? BDRV_REQ_MAY_UNMAP : 0,
mirror_write_complete, op);
} else {
assert(!(ret & BDRV_BLOCK_DATA));
bdrv_aio_discard(s->target, sector_num, op->nb_sectors,
mirror_write_complete, op);
}
return delay_ns;
}
static void coroutine_fn mirror_iteration(MirrorBlockJob *s)
{
BlockDriverState *source = s->common.bs;
int nb_sectors, sectors_per_chunk, nb_chunks;
int64_t end, sector_num, next_chunk, next_sector, hbitmap_next_sector;
MirrorOp *op;
s->sector_num = hbitmap_iter_next(&s->hbi);
if (s->sector_num < 0) {
bdrv_dirty_iter_init(source, &s->hbi);
s->sector_num = hbitmap_iter_next(&s->hbi);
trace_mirror_restart_iter(s, bdrv_get_dirty_count(source));
assert(s->sector_num >= 0);
}
hbitmap_next_sector = s->sector_num;
sector_num = s->sector_num;
sectors_per_chunk = s->granularity >> BDRV_SECTOR_BITS;
end = s->common.len >> BDRV_SECTOR_BITS;
/* Extend the QEMUIOVector to include all adjacent blocks that will
* be copied in this operation.
*
* We have to do this if we have no backing file yet in the destination,
* and the cluster size is very large. Then we need to do COW ourselves.
* The first time a cluster is copied, copy it entirely. Note that,
* because both the granularity and the cluster size are powers of two,
* the number of sectors to copy cannot exceed one cluster.
*
* We also want to extend the QEMUIOVector to include more adjacent
* dirty blocks if possible, to limit the number of I/O operations and
* run efficiently even with a small granularity.
*/
nb_chunks = 0;
nb_sectors = 0;
next_sector = sector_num;
next_chunk = sector_num / sectors_per_chunk;
/* Wait for I/O to this cluster (from a previous iteration) to be done. */
while (test_bit(next_chunk, s->in_flight_bitmap)) {
trace_mirror_yield_in_flight(s, sector_num, s->in_flight);
qemu_coroutine_yield();
}
do {
int added_sectors, added_chunks;
if (!bdrv_get_dirty(source, next_sector) ||
test_bit(next_chunk, s->in_flight_bitmap)) {
assert(nb_sectors > 0);
break;
}
added_sectors = sectors_per_chunk;
if (s->cow_bitmap && !test_bit(next_chunk, s->cow_bitmap)) {
bdrv_round_to_clusters(s->target,
next_sector, added_sectors,
&next_sector, &added_sectors);
/* On the first iteration, the rounding may make us copy
* sectors before the first dirty one.
*/
if (next_sector < sector_num) {
assert(nb_sectors == 0);
sector_num = next_sector;
next_chunk = next_sector / sectors_per_chunk;
}
}
added_sectors = MIN(added_sectors, end - (sector_num + nb_sectors));
added_chunks = (added_sectors + sectors_per_chunk - 1) / sectors_per_chunk;
/* When doing COW, it may happen that there is not enough space for
* a full cluster. Wait if that is the case.
*/
while (nb_chunks == 0 && s->buf_free_count < added_chunks) {
trace_mirror_yield_buf_busy(s, nb_chunks, s->in_flight);
qemu_coroutine_yield();
}
if (s->buf_free_count < nb_chunks + added_chunks) {
trace_mirror_break_buf_busy(s, nb_chunks, s->in_flight);
break;
}
/* We have enough free space to copy these sectors. */
bitmap_set(s->in_flight_bitmap, next_chunk, added_chunks);
nb_sectors += added_sectors;
nb_chunks += added_chunks;
next_sector += added_sectors;
next_chunk += added_chunks;
} while (next_sector < end);
/* Allocate a MirrorOp that is used as an AIO callback. */
op = g_slice_new(MirrorOp);
op->s = s;
op->sector_num = sector_num;
op->nb_sectors = nb_sectors;
/* Now make a QEMUIOVector taking enough granularity-sized chunks
* from s->buf_free.
*/
qemu_iovec_init(&op->qiov, nb_chunks);
next_sector = sector_num;
while (nb_chunks-- > 0) {
MirrorBuffer *buf = QSIMPLEQ_FIRST(&s->buf_free);
QSIMPLEQ_REMOVE_HEAD(&s->buf_free, next);
s->buf_free_count--;
qemu_iovec_add(&op->qiov, buf, s->granularity);
/* Advance the HBitmapIter in parallel, so that we do not examine
* the same sector twice.
*/
if (next_sector > hbitmap_next_sector && bdrv_get_dirty(source, next_sector)) {
hbitmap_next_sector = hbitmap_iter_next(&s->hbi);
}
next_sector += sectors_per_chunk;
}
bdrv_reset_dirty(source, sector_num, nb_sectors);
/* Copy the dirty cluster. */
s->in_flight++;
trace_mirror_one_iteration(s, sector_num, nb_sectors);
bdrv_aio_readv(source, sector_num, &op->qiov, nb_sectors,
mirror_read_complete, op);
}
bool qemu_co_queue_empty(CoQueue *queue)
{
return QSIMPLEQ_FIRST(&queue->entries) == NULL;
}
