mirror of
https://gitlab.com/qemu-project/qemu
synced 2024-11-05 20:35:44 +00:00
migration: move rate limiting to QEMUFile
Rate limiting is now simply a byte counter; client call qemu_file_rate_limit() manually to determine if they have to exit. So it is possible and simple to move the functionality to QEMUFile. This makes the remaining functionality of s->file redundant; in the next patch we can remove it and write directly to s->migration_file. Reviewed-by: Orit Wasserman <owasserm@redhat.com> Reviewed-by: Juan Quintela <quintela@redhat.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Juan Quintela <quintela@redhat.com>
This commit is contained in:
parent
442773cef1
commit
1964a39706
4 changed files with 21 additions and 99 deletions
|
@ -55,10 +55,7 @@ QEMUFile with:
|
||||||
QEMUFile *qemu_fopen_ops(void *opaque,
|
QEMUFile *qemu_fopen_ops(void *opaque,
|
||||||
QEMUFilePutBufferFunc *put_buffer,
|
QEMUFilePutBufferFunc *put_buffer,
|
||||||
QEMUFileGetBufferFunc *get_buffer,
|
QEMUFileGetBufferFunc *get_buffer,
|
||||||
QEMUFileCloseFunc *close,
|
QEMUFileCloseFunc *close);
|
||||||
QEMUFileRateLimit *rate_limit,
|
|
||||||
QEMUFileSetRateLimit *set_rate_limit,
|
|
||||||
QEMUFileGetRateLimit *get_rate_limit);
|
|
||||||
|
|
||||||
The functions have the following functionality:
|
The functions have the following functionality:
|
||||||
|
|
||||||
|
@ -80,24 +77,9 @@ Close a file and return an error code.
|
||||||
|
|
||||||
typedef int (QEMUFileCloseFunc)(void *opaque);
|
typedef int (QEMUFileCloseFunc)(void *opaque);
|
||||||
|
|
||||||
Called to determine if the file has exceeded its bandwidth allocation. The
|
|
||||||
bandwidth capping is a soft limit, not a hard limit.
|
|
||||||
|
|
||||||
typedef int (QEMUFileRateLimit)(void *opaque);
|
|
||||||
|
|
||||||
Called to change the current bandwidth allocation. This function must return
|
|
||||||
the new actual bandwidth. It should be new_rate if everything goes OK, and
|
|
||||||
the old rate otherwise.
|
|
||||||
|
|
||||||
typedef size_t (QEMUFileSetRateLimit)(void *opaque, size_t new_rate);
|
|
||||||
typedef size_t (QEMUFileGetRateLimit)(void *opaque);
|
|
||||||
|
|
||||||
You can use any internal state that you need using the opaque void *
|
You can use any internal state that you need using the opaque void *
|
||||||
pointer that is passed to all functions.
|
pointer that is passed to all functions.
|
||||||
|
|
||||||
The rate limiting functions are used to limit the bandwidth used by
|
|
||||||
QEMU migration.
|
|
||||||
|
|
||||||
The important functions for us are put_buffer()/get_buffer() that
|
The important functions for us are put_buffer()/get_buffer() that
|
||||||
allow to write/read a buffer into the QEMUFile.
|
allow to write/read a buffer into the QEMUFile.
|
||||||
|
|
||||||
|
|
|
@ -51,26 +51,11 @@ typedef int (QEMUFileCloseFunc)(void *opaque);
|
||||||
*/
|
*/
|
||||||
typedef int (QEMUFileGetFD)(void *opaque);
|
typedef int (QEMUFileGetFD)(void *opaque);
|
||||||
|
|
||||||
/* Called to determine if the file has exceeded its bandwidth allocation. The
|
|
||||||
* bandwidth capping is a soft limit, not a hard limit.
|
|
||||||
*/
|
|
||||||
typedef int (QEMUFileRateLimit)(void *opaque);
|
|
||||||
|
|
||||||
/* Called to change the current bandwidth allocation. This function must return
|
|
||||||
* the new actual bandwidth. It should be new_rate if everything goes ok, and
|
|
||||||
* the old rate otherwise
|
|
||||||
*/
|
|
||||||
typedef int64_t (QEMUFileSetRateLimit)(void *opaque, int64_t new_rate);
|
|
||||||
typedef int64_t (QEMUFileGetRateLimit)(void *opaque);
|
|
||||||
|
|
||||||
typedef struct QEMUFileOps {
|
typedef struct QEMUFileOps {
|
||||||
QEMUFilePutBufferFunc *put_buffer;
|
QEMUFilePutBufferFunc *put_buffer;
|
||||||
QEMUFileGetBufferFunc *get_buffer;
|
QEMUFileGetBufferFunc *get_buffer;
|
||||||
QEMUFileCloseFunc *close;
|
QEMUFileCloseFunc *close;
|
||||||
QEMUFileGetFD *get_fd;
|
QEMUFileGetFD *get_fd;
|
||||||
QEMUFileRateLimit *rate_limit;
|
|
||||||
QEMUFileSetRateLimit *set_rate_limit;
|
|
||||||
QEMUFileGetRateLimit *get_rate_limit;
|
|
||||||
} QEMUFileOps;
|
} QEMUFileOps;
|
||||||
|
|
||||||
QEMUFile *qemu_fopen_ops(void *opaque, const QEMUFileOps *ops);
|
QEMUFile *qemu_fopen_ops(void *opaque, const QEMUFileOps *ops);
|
||||||
|
@ -109,7 +94,8 @@ unsigned int qemu_get_be32(QEMUFile *f);
|
||||||
uint64_t qemu_get_be64(QEMUFile *f);
|
uint64_t qemu_get_be64(QEMUFile *f);
|
||||||
|
|
||||||
int qemu_file_rate_limit(QEMUFile *f);
|
int qemu_file_rate_limit(QEMUFile *f);
|
||||||
int64_t qemu_file_set_rate_limit(QEMUFile *f, int64_t new_rate);
|
void qemu_file_reset_rate_limit(QEMUFile *f);
|
||||||
|
void qemu_file_set_rate_limit(QEMUFile *f, int64_t new_rate);
|
||||||
int64_t qemu_file_get_rate_limit(QEMUFile *f);
|
int64_t qemu_file_get_rate_limit(QEMUFile *f);
|
||||||
int qemu_file_get_error(QEMUFile *f);
|
int qemu_file_get_error(QEMUFile *f);
|
||||||
|
|
||||||
|
|
51
migration.c
51
migration.c
|
@ -518,7 +518,6 @@ static int migration_put_buffer(void *opaque, const uint8_t *buf,
|
||||||
return ret;
|
return ret;
|
||||||
}
|
}
|
||||||
|
|
||||||
s->bytes_xfer += size;
|
|
||||||
return size;
|
return size;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -543,49 +542,6 @@ static int migration_get_fd(void *opaque)
|
||||||
return qemu_get_fd(s->migration_file);
|
return qemu_get_fd(s->migration_file);
|
||||||
}
|
}
|
||||||
|
|
||||||
/*
|
|
||||||
* The meaning of the return values is:
|
|
||||||
* 0: We can continue sending
|
|
||||||
* 1: Time to stop
|
|
||||||
* negative: There has been an error
|
|
||||||
*/
|
|
||||||
static int migration_rate_limit(void *opaque)
|
|
||||||
{
|
|
||||||
MigrationState *s = opaque;
|
|
||||||
int ret;
|
|
||||||
|
|
||||||
ret = qemu_file_get_error(s->file);
|
|
||||||
if (ret) {
|
|
||||||
return ret;
|
|
||||||
}
|
|
||||||
|
|
||||||
if (s->bytes_xfer >= s->xfer_limit) {
|
|
||||||
return 1;
|
|
||||||
}
|
|
||||||
|
|
||||||
return 0;
|
|
||||||
}
|
|
||||||
|
|
||||||
static int64_t migration_set_rate_limit(void *opaque, int64_t new_rate)
|
|
||||||
{
|
|
||||||
MigrationState *s = opaque;
|
|
||||||
if (qemu_file_get_error(s->file)) {
|
|
||||||
goto out;
|
|
||||||
}
|
|
||||||
|
|
||||||
s->xfer_limit = new_rate;
|
|
||||||
|
|
||||||
out:
|
|
||||||
return s->xfer_limit;
|
|
||||||
}
|
|
||||||
|
|
||||||
static int64_t migration_get_rate_limit(void *opaque)
|
|
||||||
{
|
|
||||||
MigrationState *s = opaque;
|
|
||||||
|
|
||||||
return s->xfer_limit;
|
|
||||||
}
|
|
||||||
|
|
||||||
static void *migration_thread(void *opaque)
|
static void *migration_thread(void *opaque)
|
||||||
{
|
{
|
||||||
MigrationState *s = opaque;
|
MigrationState *s = opaque;
|
||||||
|
@ -646,7 +602,7 @@ static void *migration_thread(void *opaque)
|
||||||
s->expected_downtime = s->dirty_bytes_rate / bandwidth;
|
s->expected_downtime = s->dirty_bytes_rate / bandwidth;
|
||||||
}
|
}
|
||||||
|
|
||||||
s->bytes_xfer = 0;
|
qemu_file_reset_rate_limit(s->file);
|
||||||
sleep_time = 0;
|
sleep_time = 0;
|
||||||
initial_time = current_time;
|
initial_time = current_time;
|
||||||
initial_bytes = qemu_ftell(s->file);
|
initial_bytes = qemu_ftell(s->file);
|
||||||
|
@ -679,9 +635,6 @@ static const QEMUFileOps migration_file_ops = {
|
||||||
.get_fd = migration_get_fd,
|
.get_fd = migration_get_fd,
|
||||||
.put_buffer = migration_put_buffer,
|
.put_buffer = migration_put_buffer,
|
||||||
.close = migration_close,
|
.close = migration_close,
|
||||||
.rate_limit = migration_rate_limit,
|
|
||||||
.get_rate_limit = migration_get_rate_limit,
|
|
||||||
.set_rate_limit = migration_set_rate_limit,
|
|
||||||
};
|
};
|
||||||
|
|
||||||
void migrate_fd_connect(MigrationState *s)
|
void migrate_fd_connect(MigrationState *s)
|
||||||
|
@ -689,10 +642,8 @@ void migrate_fd_connect(MigrationState *s)
|
||||||
s->state = MIG_STATE_ACTIVE;
|
s->state = MIG_STATE_ACTIVE;
|
||||||
trace_migrate_set_state(MIG_STATE_ACTIVE);
|
trace_migrate_set_state(MIG_STATE_ACTIVE);
|
||||||
|
|
||||||
s->bytes_xfer = 0;
|
|
||||||
/* This is a best 1st approximation. ns to ms */
|
/* This is a best 1st approximation. ns to ms */
|
||||||
s->expected_downtime = max_downtime/1000000;
|
s->expected_downtime = max_downtime/1000000;
|
||||||
|
|
||||||
s->cleanup_bh = qemu_bh_new(migrate_fd_cleanup, s);
|
s->cleanup_bh = qemu_bh_new(migrate_fd_cleanup, s);
|
||||||
s->file = qemu_fopen_ops(s, &migration_file_ops);
|
s->file = qemu_fopen_ops(s, &migration_file_ops);
|
||||||
|
|
||||||
|
|
31
savevm.c
31
savevm.c
|
@ -119,6 +119,9 @@ struct QEMUFile {
|
||||||
void *opaque;
|
void *opaque;
|
||||||
int is_write;
|
int is_write;
|
||||||
|
|
||||||
|
int64_t bytes_xfer;
|
||||||
|
int64_t xfer_limit;
|
||||||
|
|
||||||
int64_t pos; /* start of buffer when writing, end of buffer
|
int64_t pos; /* start of buffer when writing, end of buffer
|
||||||
when reading */
|
when reading */
|
||||||
int buf_index;
|
int buf_index;
|
||||||
|
@ -479,7 +482,6 @@ QEMUFile *qemu_fopen_ops(void *opaque, const QEMUFileOps *ops)
|
||||||
f->opaque = opaque;
|
f->opaque = opaque;
|
||||||
f->ops = ops;
|
f->ops = ops;
|
||||||
f->is_write = 0;
|
f->is_write = 0;
|
||||||
|
|
||||||
return f;
|
return f;
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -605,6 +607,7 @@ void qemu_put_buffer(QEMUFile *f, const uint8_t *buf, int size)
|
||||||
memcpy(f->buf + f->buf_index, buf, l);
|
memcpy(f->buf + f->buf_index, buf, l);
|
||||||
f->is_write = 1;
|
f->is_write = 1;
|
||||||
f->buf_index += l;
|
f->buf_index += l;
|
||||||
|
f->bytes_xfer += l;
|
||||||
buf += l;
|
buf += l;
|
||||||
size -= l;
|
size -= l;
|
||||||
if (f->buf_index >= IO_BUF_SIZE) {
|
if (f->buf_index >= IO_BUF_SIZE) {
|
||||||
|
@ -725,28 +728,28 @@ int64_t qemu_ftell(QEMUFile *f)
|
||||||
|
|
||||||
int qemu_file_rate_limit(QEMUFile *f)
|
int qemu_file_rate_limit(QEMUFile *f)
|
||||||
{
|
{
|
||||||
if (f->ops->rate_limit)
|
if (qemu_file_get_error(f)) {
|
||||||
return f->ops->rate_limit(f->opaque);
|
return 1;
|
||||||
|
}
|
||||||
|
if (f->xfer_limit > 0 && f->bytes_xfer > f->xfer_limit) {
|
||||||
|
return 1;
|
||||||
|
}
|
||||||
return 0;
|
return 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
int64_t qemu_file_get_rate_limit(QEMUFile *f)
|
int64_t qemu_file_get_rate_limit(QEMUFile *f)
|
||||||
{
|
{
|
||||||
if (f->ops->get_rate_limit)
|
return f->xfer_limit;
|
||||||
return f->ops->get_rate_limit(f->opaque);
|
|
||||||
|
|
||||||
return 0;
|
|
||||||
}
|
}
|
||||||
|
|
||||||
int64_t qemu_file_set_rate_limit(QEMUFile *f, int64_t new_rate)
|
void qemu_file_set_rate_limit(QEMUFile *f, int64_t limit)
|
||||||
{
|
{
|
||||||
/* any failed or completed migration keeps its state to allow probing of
|
f->xfer_limit = limit;
|
||||||
* migration data, but has no associated file anymore */
|
}
|
||||||
if (f && f->ops->set_rate_limit)
|
|
||||||
return f->ops->set_rate_limit(f->opaque, new_rate);
|
|
||||||
|
|
||||||
return 0;
|
void qemu_file_reset_rate_limit(QEMUFile *f)
|
||||||
|
{
|
||||||
|
f->bytes_xfer = 0;
|
||||||
}
|
}
|
||||||
|
|
||||||
void qemu_put_be16(QEMUFile *f, unsigned int v)
|
void qemu_put_be16(QEMUFile *f, unsigned int v)
|
||||||
|
|
Loading…
Reference in a new issue