2003-06-30 12:03:06 +02:00
|
|
|
/*
|
|
|
|
* QEMU System Emulator block driver
|
2007-09-16 23:08:06 +02:00
|
|
|
*
|
2003-06-30 12:03:06 +02:00
|
|
|
* Copyright (c) 2003 Fabrice Bellard
|
2007-09-16 23:08:06 +02:00
|
|
|
*
|
2003-06-30 12:03:06 +02:00
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
|
|
* of this software and associated documentation files (the "Software"), to deal
|
|
|
|
* in the Software without restriction, including without limitation the rights
|
|
|
|
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
|
|
* copies of the Software, and to permit persons to whom the Software is
|
|
|
|
* furnished to do so, subject to the following conditions:
|
|
|
|
*
|
|
|
|
* The above copyright notice and this permission notice shall be included in
|
|
|
|
* all copies or substantial portions of the Software.
|
|
|
|
*
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
|
|
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
|
|
* THE SOFTWARE.
|
|
|
|
*/
|
2008-12-05 18:53:21 +01:00
|
|
|
#include "config-host.h"
|
2007-11-11 03:51:17 +01:00
|
|
|
#include "qemu-common.h"
|
2010-05-22 19:15:08 +02:00
|
|
|
#include "trace.h"
|
2012-12-17 18:19:49 +01:00
|
|
|
#include "monitor/monitor.h"
|
2012-12-17 18:19:44 +01:00
|
|
|
#include "block/block_int.h"
|
|
|
|
#include "block/blockjob.h"
|
2012-12-17 18:20:00 +01:00
|
|
|
#include "qemu/module.h"
|
2012-12-17 18:19:43 +01:00
|
|
|
#include "qapi/qmp/qjson.h"
|
2012-12-17 18:20:04 +01:00
|
|
|
#include "sysemu/sysemu.h"
|
2012-12-17 18:20:00 +01:00
|
|
|
#include "qemu/notify.h"
|
2012-12-17 18:19:44 +01:00
|
|
|
#include "block/coroutine.h"
|
2011-09-21 22:16:47 +02:00
|
|
|
#include "qmp-commands.h"
|
2012-12-17 18:20:00 +01:00
|
|
|
#include "qemu/timer.h"
|
2003-06-30 12:03:06 +02:00
|
|
|
|
2009-07-27 16:12:56 +02:00
|
|
|
#ifdef CONFIG_BSD
|
2005-04-26 23:59:26 +02:00
|
|
|
#include <sys/types.h>
|
|
|
|
#include <sys/stat.h>
|
|
|
|
#include <sys/ioctl.h>
|
2009-09-12 09:36:22 +02:00
|
|
|
#include <sys/queue.h>
|
2009-03-07 21:06:23 +01:00
|
|
|
#ifndef __DragonFly__
|
2005-04-26 23:59:26 +02:00
|
|
|
#include <sys/disk.h>
|
|
|
|
#endif
|
2009-03-07 21:06:23 +01:00
|
|
|
#endif
|
2005-04-26 23:59:26 +02:00
|
|
|
|
2009-03-08 17:26:59 +01:00
|
|
|
#ifdef _WIN32
|
|
|
|
#include <windows.h>
|
|
|
|
#endif
|
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
#define NOT_DONE 0x7fffffff /* used while emulated sync operation in progress */
|
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
typedef enum {
|
|
|
|
BDRV_REQ_COPY_ON_READ = 0x1,
|
2012-02-07 14:27:25 +01:00
|
|
|
BDRV_REQ_ZERO_WRITE = 0x2,
|
2012-01-18 15:40:42 +01:00
|
|
|
} BdrvRequestFlags;
|
|
|
|
|
2011-09-06 18:58:59 +02:00
|
|
|
static void bdrv_dev_change_media_cb(BlockDriverState *bs, bool load);
|
2009-04-07 20:43:24 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
|
2009-04-07 20:43:20 +02:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque);
|
2009-04-07 20:43:24 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_aio_writev_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
|
2006-08-07 04:38:06 +02:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque);
|
2011-07-15 13:50:26 +02:00
|
|
|
static int coroutine_fn bdrv_co_readv_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
QEMUIOVector *iov);
|
|
|
|
static int coroutine_fn bdrv_co_writev_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
QEMUIOVector *iov);
|
2011-10-05 18:17:03 +02:00
|
|
|
static int coroutine_fn bdrv_co_do_readv(BlockDriverState *bs,
|
2012-01-18 15:40:42 +01:00
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov,
|
|
|
|
BdrvRequestFlags flags);
|
2011-10-13 14:08:22 +02:00
|
|
|
static int coroutine_fn bdrv_co_do_writev(BlockDriverState *bs,
|
2012-02-07 14:27:25 +01:00
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov,
|
|
|
|
BdrvRequestFlags flags);
|
2011-10-13 14:08:23 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_co_aio_rw_vector(BlockDriverState *bs,
|
|
|
|
int64_t sector_num,
|
|
|
|
QEMUIOVector *qiov,
|
|
|
|
int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb,
|
|
|
|
void *opaque,
|
2011-10-13 22:09:28 +02:00
|
|
|
bool is_write);
|
2011-10-13 14:08:23 +02:00
|
|
|
static void coroutine_fn bdrv_co_do_rw(void *opaque);
|
2012-03-20 15:12:58 +01:00
|
|
|
static int coroutine_fn bdrv_co_do_write_zeroes(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors);
|
2006-04-26 00:36:06 +02:00
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
static bool bdrv_exceed_bps_limits(BlockDriverState *bs, int nb_sectors,
|
|
|
|
bool is_write, double elapsed_time, uint64_t *wait);
|
|
|
|
static bool bdrv_exceed_iops_limits(BlockDriverState *bs, bool is_write,
|
|
|
|
double elapsed_time, uint64_t *wait);
|
|
|
|
static bool bdrv_exceed_io_limits(BlockDriverState *bs, int nb_sectors,
|
|
|
|
bool is_write, int64_t *wait);
|
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
static QTAILQ_HEAD(, BlockDriverState) bdrv_states =
|
|
|
|
QTAILQ_HEAD_INITIALIZER(bdrv_states);
|
2008-09-17 21:04:14 +02:00
|
|
|
|
2010-04-13 11:29:33 +02:00
|
|
|
static QLIST_HEAD(, BlockDriver) bdrv_drivers =
|
|
|
|
QLIST_HEAD_INITIALIZER(bdrv_drivers);
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2010-06-25 10:33:39 +02:00
|
|
|
/* The device to use for VM snapshots */
|
|
|
|
static BlockDriverState *bs_snapshots;
|
|
|
|
|
2009-10-27 18:41:44 +01:00
|
|
|
/* If non-zero, use only whitelisted block drivers */
|
|
|
|
static int use_bdrv_whitelist;
|
|
|
|
|
2010-12-09 12:53:00 +01:00
|
|
|
#ifdef _WIN32
|
|
|
|
static int is_windows_drive_prefix(const char *filename)
|
|
|
|
{
|
|
|
|
return (((filename[0] >= 'a' && filename[0] <= 'z') ||
|
|
|
|
(filename[0] >= 'A' && filename[0] <= 'Z')) &&
|
|
|
|
filename[1] == ':');
|
|
|
|
}
|
|
|
|
|
|
|
|
int is_windows_drive(const char *filename)
|
|
|
|
{
|
|
|
|
if (is_windows_drive_prefix(filename) &&
|
|
|
|
filename[2] == '\0')
|
|
|
|
return 1;
|
|
|
|
if (strstart(filename, "\\\\.\\", NULL) ||
|
|
|
|
strstart(filename, "//./", NULL))
|
|
|
|
return 1;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
2011-11-03 09:57:25 +01:00
|
|
|
/* throttling disk I/O limits */
|
2011-11-08 06:00:14 +01:00
|
|
|
void bdrv_io_limits_disable(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
bs->io_limits_enabled = false;
|
|
|
|
|
|
|
|
while (qemu_co_queue_next(&bs->throttled_reqs));
|
|
|
|
|
|
|
|
if (bs->block_timer) {
|
|
|
|
qemu_del_timer(bs->block_timer);
|
|
|
|
qemu_free_timer(bs->block_timer);
|
|
|
|
bs->block_timer = NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->slice_start = 0;
|
|
|
|
bs->slice_end = 0;
|
|
|
|
bs->slice_time = 0;
|
|
|
|
memset(&bs->io_base, 0, sizeof(bs->io_base));
|
|
|
|
}
|
|
|
|
|
2011-11-03 09:57:25 +01:00
|
|
|
static void bdrv_block_timer(void *opaque)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs = opaque;
|
|
|
|
|
|
|
|
qemu_co_queue_next(&bs->throttled_reqs);
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_io_limits_enable(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
qemu_co_queue_init(&bs->throttled_reqs);
|
|
|
|
bs->block_timer = qemu_new_timer_ns(vm_clock, bdrv_block_timer, bs);
|
|
|
|
bs->slice_time = 5 * BLOCK_IO_SLICE_TIME;
|
|
|
|
bs->slice_start = qemu_get_clock_ns(vm_clock);
|
|
|
|
bs->slice_end = bs->slice_start + bs->slice_time;
|
|
|
|
memset(&bs->io_base, 0, sizeof(bs->io_base));
|
|
|
|
bs->io_limits_enabled = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool bdrv_io_limits_enabled(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockIOLimit *io_limits = &bs->io_limits;
|
|
|
|
return io_limits->bps[BLOCK_IO_LIMIT_READ]
|
|
|
|
|| io_limits->bps[BLOCK_IO_LIMIT_WRITE]
|
|
|
|
|| io_limits->bps[BLOCK_IO_LIMIT_TOTAL]
|
|
|
|
|| io_limits->iops[BLOCK_IO_LIMIT_READ]
|
|
|
|
|| io_limits->iops[BLOCK_IO_LIMIT_WRITE]
|
|
|
|
|| io_limits->iops[BLOCK_IO_LIMIT_TOTAL];
|
|
|
|
}
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
static void bdrv_io_limits_intercept(BlockDriverState *bs,
|
|
|
|
bool is_write, int nb_sectors)
|
|
|
|
{
|
|
|
|
int64_t wait_time = -1;
|
|
|
|
|
|
|
|
if (!qemu_co_queue_empty(&bs->throttled_reqs)) {
|
|
|
|
qemu_co_queue_wait(&bs->throttled_reqs);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* In fact, we hope to keep each request's timing, in FIFO mode. The next
|
|
|
|
* throttled requests will not be dequeued until the current request is
|
|
|
|
* allowed to be serviced. So if the current request still exceeds the
|
|
|
|
* limits, it will be inserted to the head. All requests followed it will
|
|
|
|
* be still in throttled_reqs queue.
|
|
|
|
*/
|
|
|
|
|
|
|
|
while (bdrv_exceed_io_limits(bs, nb_sectors, is_write, &wait_time)) {
|
|
|
|
qemu_mod_timer(bs->block_timer,
|
|
|
|
wait_time + qemu_get_clock_ns(vm_clock));
|
|
|
|
qemu_co_queue_wait_insert_head(&bs->throttled_reqs);
|
|
|
|
}
|
|
|
|
|
|
|
|
qemu_co_queue_next(&bs->throttled_reqs);
|
|
|
|
}
|
|
|
|
|
2010-12-09 12:53:00 +01:00
|
|
|
/* check if the path starts with "<protocol>:" */
|
|
|
|
static int path_has_protocol(const char *path)
|
|
|
|
{
|
2012-05-08 16:51:48 +02:00
|
|
|
const char *p;
|
|
|
|
|
2010-12-09 12:53:00 +01:00
|
|
|
#ifdef _WIN32
|
|
|
|
if (is_windows_drive(path) ||
|
|
|
|
is_windows_drive_prefix(path)) {
|
|
|
|
return 0;
|
|
|
|
}
|
2012-05-08 16:51:48 +02:00
|
|
|
p = path + strcspn(path, ":/\\");
|
|
|
|
#else
|
|
|
|
p = path + strcspn(path, ":/");
|
2010-12-09 12:53:00 +01:00
|
|
|
#endif
|
|
|
|
|
2012-05-08 16:51:48 +02:00
|
|
|
return *p == ':';
|
2010-12-09 12:53:00 +01:00
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
int path_is_absolute(const char *path)
|
2005-10-30 19:30:10 +01:00
|
|
|
{
|
2007-01-07 19:22:37 +01:00
|
|
|
#ifdef _WIN32
|
|
|
|
/* specific case for names like: "\\.\d:" */
|
2012-05-08 16:51:47 +02:00
|
|
|
if (is_windows_drive(path) || is_windows_drive_prefix(path)) {
|
2007-01-07 19:22:37 +01:00
|
|
|
return 1;
|
2012-05-08 16:51:47 +02:00
|
|
|
}
|
|
|
|
return (*path == '/' || *path == '\\');
|
2007-01-07 18:27:07 +01:00
|
|
|
#else
|
2012-05-08 16:51:47 +02:00
|
|
|
return (*path == '/');
|
2007-01-07 18:27:07 +01:00
|
|
|
#endif
|
2005-10-30 19:30:10 +01:00
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
/* if filename is absolute, just copy it to dest. Otherwise, build a
|
|
|
|
path to it by considering it is relative to base_path. URL are
|
|
|
|
supported. */
|
|
|
|
void path_combine(char *dest, int dest_size,
|
|
|
|
const char *base_path,
|
|
|
|
const char *filename)
|
2005-10-30 19:30:10 +01:00
|
|
|
{
|
2006-08-01 18:21:11 +02:00
|
|
|
const char *p, *p1;
|
|
|
|
int len;
|
|
|
|
|
|
|
|
if (dest_size <= 0)
|
|
|
|
return;
|
|
|
|
if (path_is_absolute(filename)) {
|
|
|
|
pstrcpy(dest, dest_size, filename);
|
|
|
|
} else {
|
|
|
|
p = strchr(base_path, ':');
|
|
|
|
if (p)
|
|
|
|
p++;
|
|
|
|
else
|
|
|
|
p = base_path;
|
2007-01-07 18:27:07 +01:00
|
|
|
p1 = strrchr(base_path, '/');
|
|
|
|
#ifdef _WIN32
|
|
|
|
{
|
|
|
|
const char *p2;
|
|
|
|
p2 = strrchr(base_path, '\\');
|
|
|
|
if (!p1 || p2 > p1)
|
|
|
|
p1 = p2;
|
|
|
|
}
|
|
|
|
#endif
|
2006-08-01 18:21:11 +02:00
|
|
|
if (p1)
|
|
|
|
p1++;
|
|
|
|
else
|
|
|
|
p1 = base_path;
|
|
|
|
if (p1 > p)
|
|
|
|
p = p1;
|
|
|
|
len = p - base_path;
|
|
|
|
if (len > dest_size - 1)
|
|
|
|
len = dest_size - 1;
|
|
|
|
memcpy(dest, base_path, len);
|
|
|
|
dest[len] = '\0';
|
|
|
|
pstrcat(dest, dest_size, filename);
|
2005-10-30 19:30:10 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-05-08 16:51:50 +02:00
|
|
|
void bdrv_get_full_backing_filename(BlockDriverState *bs, char *dest, size_t sz)
|
|
|
|
{
|
|
|
|
if (bs->backing_file[0] == '\0' || path_has_protocol(bs->backing_file)) {
|
|
|
|
pstrcpy(dest, sz, bs->backing_file);
|
|
|
|
} else {
|
|
|
|
path_combine(dest, sz, bs->filename, bs->backing_file);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2009-05-10 00:03:42 +02:00
|
|
|
void bdrv_register(BlockDriver *bdrv)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2011-10-13 22:09:28 +02:00
|
|
|
/* Block drivers without coroutine functions need emulation */
|
|
|
|
if (!bdrv->bdrv_co_readv) {
|
2011-07-15 13:50:26 +02:00
|
|
|
bdrv->bdrv_co_readv = bdrv_co_readv_em;
|
|
|
|
bdrv->bdrv_co_writev = bdrv_co_writev_em;
|
|
|
|
|
2011-10-13 22:09:31 +02:00
|
|
|
/* bdrv_co_readv_em()/brdv_co_writev_em() work in terms of aio, so if
|
|
|
|
* the block driver lacks aio we need to emulate that too.
|
|
|
|
*/
|
2011-07-15 13:50:26 +02:00
|
|
|
if (!bdrv->bdrv_aio_readv) {
|
|
|
|
/* add AIO emulation layer */
|
|
|
|
bdrv->bdrv_aio_readv = bdrv_aio_readv_em;
|
|
|
|
bdrv->bdrv_aio_writev = bdrv_aio_writev_em;
|
|
|
|
}
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
2009-09-04 19:01:49 +02:00
|
|
|
|
2010-04-13 11:29:33 +02:00
|
|
|
QLIST_INSERT_HEAD(&bdrv_drivers, bdrv, list);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2004-03-14 22:38:54 +01:00
|
|
|
|
|
|
|
/* create a new block device (by default it is empty) */
|
|
|
|
BlockDriverState *bdrv_new(const char *device_name)
|
|
|
|
{
|
2010-04-10 08:02:42 +02:00
|
|
|
BlockDriverState *bs;
|
2004-03-14 22:38:54 +01:00
|
|
|
|
2011-08-21 05:09:37 +02:00
|
|
|
bs = g_malloc0(sizeof(BlockDriverState));
|
2004-03-14 22:38:54 +01:00
|
|
|
pstrcpy(bs->device_name, sizeof(bs->device_name), device_name);
|
2004-08-01 23:59:26 +02:00
|
|
|
if (device_name[0] != '\0') {
|
2010-04-10 08:02:42 +02:00
|
|
|
QTAILQ_INSERT_TAIL(&bdrv_states, bs, list);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2011-09-26 22:43:50 +02:00
|
|
|
bdrv_iostatus_disable(bs);
|
2012-08-23 11:20:36 +02:00
|
|
|
notifier_list_init(&bs->close_notifiers);
|
|
|
|
|
2004-03-14 22:38:54 +01:00
|
|
|
return bs;
|
|
|
|
}
|
|
|
|
|
2012-08-23 11:20:36 +02:00
|
|
|
void bdrv_add_close_notifier(BlockDriverState *bs, Notifier *notify)
|
|
|
|
{
|
|
|
|
notifier_list_add(&bs->close_notifiers, notify);
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
BlockDriver *bdrv_find_format(const char *format_name)
|
|
|
|
{
|
|
|
|
BlockDriver *drv1;
|
2010-04-13 11:29:33 +02:00
|
|
|
QLIST_FOREACH(drv1, &bdrv_drivers, list) {
|
|
|
|
if (!strcmp(drv1->format_name, format_name)) {
|
2004-08-01 23:59:26 +02:00
|
|
|
return drv1;
|
2010-04-13 11:29:33 +02:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2009-10-27 18:41:44 +01:00
|
|
|
static int bdrv_is_whitelisted(BlockDriver *drv)
|
|
|
|
{
|
|
|
|
static const char *whitelist[] = {
|
|
|
|
CONFIG_BDRV_WHITELIST
|
|
|
|
};
|
|
|
|
const char **p;
|
|
|
|
|
|
|
|
if (!whitelist[0])
|
|
|
|
return 1; /* no whitelist, anything goes */
|
|
|
|
|
|
|
|
for (p = whitelist; *p; p++) {
|
|
|
|
if (!strcmp(drv->format_name, *p)) {
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
BlockDriver *bdrv_find_whitelisted_format(const char *format_name)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bdrv_find_format(format_name);
|
|
|
|
return drv && bdrv_is_whitelisted(drv) ? drv : NULL;
|
|
|
|
}
|
|
|
|
|
2012-05-07 10:50:42 +02:00
|
|
|
typedef struct CreateCo {
|
|
|
|
BlockDriver *drv;
|
|
|
|
char *filename;
|
|
|
|
QEMUOptionParameter *options;
|
|
|
|
int ret;
|
|
|
|
} CreateCo;
|
|
|
|
|
|
|
|
static void coroutine_fn bdrv_create_co_entry(void *opaque)
|
|
|
|
{
|
|
|
|
CreateCo *cco = opaque;
|
|
|
|
assert(cco->drv);
|
|
|
|
|
|
|
|
cco->ret = cco->drv->bdrv_create(cco->filename, cco->options);
|
|
|
|
}
|
|
|
|
|
2009-05-18 16:42:10 +02:00
|
|
|
int bdrv_create(BlockDriver *drv, const char* filename,
|
|
|
|
QEMUOptionParameter *options)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2012-05-07 10:50:42 +02:00
|
|
|
int ret;
|
|
|
|
|
|
|
|
Coroutine *co;
|
|
|
|
CreateCo cco = {
|
|
|
|
.drv = drv,
|
|
|
|
.filename = g_strdup(filename),
|
|
|
|
.options = options,
|
|
|
|
.ret = NOT_DONE,
|
|
|
|
};
|
|
|
|
|
|
|
|
if (!drv->bdrv_create) {
|
2012-10-17 21:45:25 +02:00
|
|
|
ret = -ENOTSUP;
|
|
|
|
goto out;
|
2012-05-07 10:50:42 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (qemu_in_coroutine()) {
|
|
|
|
/* Fast-path if already in coroutine context */
|
|
|
|
bdrv_create_co_entry(&cco);
|
|
|
|
} else {
|
|
|
|
co = qemu_coroutine_create(bdrv_create_co_entry);
|
|
|
|
qemu_coroutine_enter(co, &cco);
|
|
|
|
while (cco.ret == NOT_DONE) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = cco.ret;
|
2009-05-18 16:42:10 +02:00
|
|
|
|
2012-10-17 21:45:25 +02:00
|
|
|
out:
|
|
|
|
g_free(cco.filename);
|
2012-05-07 10:50:42 +02:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2010-04-07 22:30:24 +02:00
|
|
|
int bdrv_create_file(const char* filename, QEMUOptionParameter *options)
|
|
|
|
{
|
|
|
|
BlockDriver *drv;
|
|
|
|
|
2010-05-26 04:35:36 +02:00
|
|
|
drv = bdrv_find_protocol(filename);
|
2010-04-07 22:30:24 +02:00
|
|
|
if (drv == NULL) {
|
2010-11-30 16:14:14 +01:00
|
|
|
return -ENOENT;
|
2010-04-07 22:30:24 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return bdrv_create(drv, filename, options);
|
|
|
|
}
|
|
|
|
|
2012-05-28 09:27:54 +02:00
|
|
|
/*
|
|
|
|
* Create a uniquely-named empty temporary file.
|
|
|
|
* Return 0 upon success, otherwise a negative errno value.
|
|
|
|
*/
|
|
|
|
int get_tmp_filename(char *filename, int size)
|
2004-08-03 23:14:23 +02:00
|
|
|
{
|
2012-05-28 09:27:54 +02:00
|
|
|
#ifdef _WIN32
|
2007-01-07 18:27:07 +01:00
|
|
|
char temp_dir[MAX_PATH];
|
2012-05-28 09:27:54 +02:00
|
|
|
/* GetTempFileName requires that its output buffer (4th param)
|
|
|
|
have length MAX_PATH or greater. */
|
|
|
|
assert(size >= MAX_PATH);
|
|
|
|
return (GetTempPath(MAX_PATH, temp_dir)
|
|
|
|
&& GetTempFileName(temp_dir, "qem", 0, filename)
|
|
|
|
? 0 : -GetLastError());
|
2004-08-03 23:14:23 +02:00
|
|
|
#else
|
2004-04-01 01:37:16 +02:00
|
|
|
int fd;
|
2008-09-14 08:45:34 +02:00
|
|
|
const char *tmpdir;
|
2008-03-10 01:05:34 +01:00
|
|
|
tmpdir = getenv("TMPDIR");
|
|
|
|
if (!tmpdir)
|
|
|
|
tmpdir = "/tmp";
|
2012-05-28 09:27:54 +02:00
|
|
|
if (snprintf(filename, size, "%s/vl.XXXXXX", tmpdir) >= size) {
|
|
|
|
return -EOVERFLOW;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
fd = mkstemp(filename);
|
2012-09-05 15:26:22 +02:00
|
|
|
if (fd < 0) {
|
|
|
|
return -errno;
|
|
|
|
}
|
|
|
|
if (close(fd) != 0) {
|
|
|
|
unlink(filename);
|
2012-05-28 09:27:54 +02:00
|
|
|
return -errno;
|
|
|
|
}
|
|
|
|
return 0;
|
2004-08-03 23:14:23 +02:00
|
|
|
#endif
|
2012-05-28 09:27:54 +02:00
|
|
|
}
|
2003-06-30 12:03:06 +02:00
|
|
|
|
2010-04-07 22:30:24 +02:00
|
|
|
/*
|
|
|
|
* Detect host devices. By convention, /dev/cdrom[N] is always
|
|
|
|
* recognized as a host CDROM.
|
|
|
|
*/
|
|
|
|
static BlockDriver *find_hdev_driver(const char *filename)
|
|
|
|
{
|
|
|
|
int score_max = 0, score;
|
|
|
|
BlockDriver *drv = NULL, *d;
|
|
|
|
|
|
|
|
QLIST_FOREACH(d, &bdrv_drivers, list) {
|
|
|
|
if (d->bdrv_probe_device) {
|
|
|
|
score = d->bdrv_probe_device(filename);
|
|
|
|
if (score > score_max) {
|
|
|
|
score_max = score;
|
|
|
|
drv = d;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return drv;
|
|
|
|
}
|
|
|
|
|
2010-05-26 04:35:36 +02:00
|
|
|
BlockDriver *bdrv_find_protocol(const char *filename)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv1;
|
|
|
|
char protocol[128];
|
2009-07-02 15:12:26 +02:00
|
|
|
int len;
|
2006-08-01 18:21:11 +02:00
|
|
|
const char *p;
|
2006-08-19 13:45:59 +02:00
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
/* TODO Drivers without bdrv_file_open must be specified explicitly */
|
|
|
|
|
2010-06-23 12:25:17 +02:00
|
|
|
/*
|
|
|
|
* XXX(hch): we really should not let host device detection
|
|
|
|
* override an explicit protocol specification, but moving this
|
|
|
|
* later breaks access to device names with colons in them.
|
|
|
|
* Thanks to the brain-dead persistent naming schemes on udev-
|
|
|
|
* based Linux systems those actually are quite common.
|
|
|
|
*/
|
|
|
|
drv1 = find_hdev_driver(filename);
|
|
|
|
if (drv1) {
|
|
|
|
return drv1;
|
|
|
|
}
|
|
|
|
|
2010-12-09 12:53:00 +01:00
|
|
|
if (!path_has_protocol(filename)) {
|
2010-06-23 12:25:17 +02:00
|
|
|
return bdrv_find_format("file");
|
2010-04-07 22:30:24 +02:00
|
|
|
}
|
2010-12-09 12:53:00 +01:00
|
|
|
p = strchr(filename, ':');
|
|
|
|
assert(p != NULL);
|
2009-07-02 15:12:26 +02:00
|
|
|
len = p - filename;
|
|
|
|
if (len > sizeof(protocol) - 1)
|
|
|
|
len = sizeof(protocol) - 1;
|
|
|
|
memcpy(protocol, filename, len);
|
|
|
|
protocol[len] = '\0';
|
2010-04-13 11:29:33 +02:00
|
|
|
QLIST_FOREACH(drv1, &bdrv_drivers, list) {
|
2007-09-16 23:08:06 +02:00
|
|
|
if (drv1->protocol_name &&
|
2010-04-13 11:29:33 +02:00
|
|
|
!strcmp(drv1->protocol_name, protocol)) {
|
2006-08-01 18:21:11 +02:00
|
|
|
return drv1;
|
2010-04-13 11:29:33 +02:00
|
|
|
}
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2012-11-12 17:35:27 +01:00
|
|
|
static int find_image_format(BlockDriverState *bs, const char *filename,
|
|
|
|
BlockDriver **pdrv)
|
2009-06-15 13:55:19 +02:00
|
|
|
{
|
2012-11-12 17:35:27 +01:00
|
|
|
int score, score_max;
|
2009-06-15 13:55:19 +02:00
|
|
|
BlockDriver *drv1, *drv;
|
|
|
|
uint8_t buf[2048];
|
2012-11-12 17:35:27 +01:00
|
|
|
int ret = 0;
|
2010-05-17 18:45:57 +02:00
|
|
|
|
2010-06-01 18:37:31 +02:00
|
|
|
/* Return the raw BlockDriver * to scsi-generic devices or empty drives */
|
2013-01-10 15:39:27 +01:00
|
|
|
if (bs->sg || !bdrv_is_inserted(bs) || bdrv_getlength(bs) == 0) {
|
2010-07-21 21:51:51 +02:00
|
|
|
drv = bdrv_find_format("raw");
|
|
|
|
if (!drv) {
|
|
|
|
ret = -ENOENT;
|
|
|
|
}
|
|
|
|
*pdrv = drv;
|
|
|
|
return ret;
|
2010-05-27 17:56:28 +02:00
|
|
|
}
|
2010-05-17 18:45:57 +02:00
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
ret = bdrv_pread(bs, 0, buf, sizeof(buf));
|
|
|
|
if (ret < 0) {
|
2010-07-21 21:51:51 +02:00
|
|
|
*pdrv = NULL;
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
score_max = 0;
|
2010-04-07 22:30:24 +02:00
|
|
|
drv = NULL;
|
2010-04-13 11:29:33 +02:00
|
|
|
QLIST_FOREACH(drv1, &bdrv_drivers, list) {
|
2006-08-01 18:21:11 +02:00
|
|
|
if (drv1->bdrv_probe) {
|
|
|
|
score = drv1->bdrv_probe(buf, ret, filename);
|
|
|
|
if (score > score_max) {
|
|
|
|
score_max = score;
|
|
|
|
drv = drv1;
|
|
|
|
}
|
2003-07-01 01:17:31 +02:00
|
|
|
}
|
2003-06-30 12:03:06 +02:00
|
|
|
}
|
2010-07-21 21:51:51 +02:00
|
|
|
if (!drv) {
|
|
|
|
ret = -ENOENT;
|
|
|
|
}
|
|
|
|
*pdrv = drv;
|
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2010-04-19 17:56:41 +02:00
|
|
|
/**
|
|
|
|
* Set the current 'total_sectors' value
|
|
|
|
*/
|
|
|
|
static int refresh_total_sectors(BlockDriverState *bs, int64_t hint)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
2010-05-17 18:46:04 +02:00
|
|
|
/* Do not attempt drv->bdrv_getlength() on scsi-generic devices */
|
|
|
|
if (bs->sg)
|
|
|
|
return 0;
|
|
|
|
|
2010-04-19 17:56:41 +02:00
|
|
|
/* query actual device if possible, otherwise just trust the hint */
|
|
|
|
if (drv->bdrv_getlength) {
|
|
|
|
int64_t length = drv->bdrv_getlength(bs);
|
|
|
|
if (length < 0) {
|
|
|
|
return length;
|
|
|
|
}
|
|
|
|
hint = length >> BDRV_SECTOR_BITS;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->total_sectors = hint;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2011-08-04 13:26:51 +02:00
|
|
|
/**
|
|
|
|
* Set open flags for a given cache mode
|
|
|
|
*
|
|
|
|
* Return 0 on success, -1 if the cache mode was invalid.
|
|
|
|
*/
|
|
|
|
int bdrv_parse_cache_flags(const char *mode, int *flags)
|
|
|
|
{
|
|
|
|
*flags &= ~BDRV_O_CACHE_MASK;
|
|
|
|
|
|
|
|
if (!strcmp(mode, "off") || !strcmp(mode, "none")) {
|
|
|
|
*flags |= BDRV_O_NOCACHE | BDRV_O_CACHE_WB;
|
2011-08-04 13:26:52 +02:00
|
|
|
} else if (!strcmp(mode, "directsync")) {
|
|
|
|
*flags |= BDRV_O_NOCACHE;
|
2011-08-04 13:26:51 +02:00
|
|
|
} else if (!strcmp(mode, "writeback")) {
|
|
|
|
*flags |= BDRV_O_CACHE_WB;
|
|
|
|
} else if (!strcmp(mode, "unsafe")) {
|
|
|
|
*flags |= BDRV_O_CACHE_WB;
|
|
|
|
*flags |= BDRV_O_NO_FLUSH;
|
|
|
|
} else if (!strcmp(mode, "writethrough")) {
|
|
|
|
/* this is the default */
|
|
|
|
} else {
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2011-11-28 17:08:47 +01:00
|
|
|
/**
|
|
|
|
* The copy-on-read flag is actually a reference count so multiple users may
|
|
|
|
* use the feature without worrying about clobbering its previous state.
|
|
|
|
* Copy-on-read stays enabled until all users have called to disable it.
|
|
|
|
*/
|
|
|
|
void bdrv_enable_copy_on_read(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
bs->copy_on_read++;
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_disable_copy_on_read(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
assert(bs->copy_on_read > 0);
|
|
|
|
bs->copy_on_read--;
|
|
|
|
}
|
|
|
|
|
2012-11-12 17:05:39 +01:00
|
|
|
static int bdrv_open_flags(BlockDriverState *bs, int flags)
|
|
|
|
{
|
|
|
|
int open_flags = flags | BDRV_O_CACHE_WB;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Clear flags that are internal to the block layer before opening the
|
|
|
|
* image.
|
|
|
|
*/
|
|
|
|
open_flags &= ~(BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Snapshots should be writable.
|
|
|
|
*/
|
|
|
|
if (bs->is_temporary) {
|
|
|
|
open_flags |= BDRV_O_RDWR;
|
|
|
|
}
|
|
|
|
|
|
|
|
return open_flags;
|
|
|
|
}
|
|
|
|
|
2010-04-14 15:24:50 +02:00
|
|
|
/*
|
|
|
|
* Common part for opening disk images and files
|
|
|
|
*/
|
2012-11-12 17:35:27 +01:00
|
|
|
static int bdrv_open_common(BlockDriverState *bs, BlockDriverState *file,
|
|
|
|
const char *filename,
|
2010-04-14 15:24:50 +02:00
|
|
|
int flags, BlockDriver *drv)
|
|
|
|
{
|
|
|
|
int ret, open_flags;
|
|
|
|
|
|
|
|
assert(drv != NULL);
|
2012-05-08 16:51:49 +02:00
|
|
|
assert(bs->file == NULL);
|
2010-04-14 15:24:50 +02:00
|
|
|
|
2011-09-22 21:14:12 +02:00
|
|
|
trace_bdrv_open_common(bs, filename, flags, drv->format_name);
|
|
|
|
|
2010-04-14 15:24:50 +02:00
|
|
|
bs->open_flags = flags;
|
|
|
|
bs->buffer_alignment = 512;
|
|
|
|
|
2011-11-28 17:08:47 +01:00
|
|
|
assert(bs->copy_on_read == 0); /* bdrv_new() and bdrv_close() make it so */
|
|
|
|
if ((flags & BDRV_O_RDWR) && (flags & BDRV_O_COPY_ON_READ)) {
|
|
|
|
bdrv_enable_copy_on_read(bs);
|
|
|
|
}
|
|
|
|
|
2010-04-14 15:24:50 +02:00
|
|
|
pstrcpy(bs->filename, sizeof(bs->filename), filename);
|
|
|
|
|
|
|
|
if (use_bdrv_whitelist && !bdrv_is_whitelisted(drv)) {
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->drv = drv;
|
2011-08-21 05:09:37 +02:00
|
|
|
bs->opaque = g_malloc0(drv->instance_size);
|
2010-04-14 15:24:50 +02:00
|
|
|
|
2011-10-27 11:54:28 +02:00
|
|
|
bs->enable_write_cache = !!(flags & BDRV_O_CACHE_WB);
|
2012-11-12 17:05:39 +01:00
|
|
|
open_flags = bdrv_open_flags(bs, flags);
|
2010-04-14 15:24:50 +02:00
|
|
|
|
block: correctly set the keep_read_only flag
I believe the bs->keep_read_only flag is supposed to reflect
the initial open state of the device. If the device is initially
opened R/O, then commit operations, or reopen operations changing
to R/W, are prohibited.
Currently, the keep_read_only flag is only accurate for the active
layer, and its backing file. Subsequent images end up always having
the keep_read_only flag set.
For instance, what happens now:
[ base ] kro = 1, ro = 1
|
v
[ snap-1 ] kro = 1, ro = 1
|
v
[ snap-2 ] kro = 0, ro = 1
|
v
[ active ] kro = 0, ro = 0
What we want:
[ base ] kro = 0, ro = 1
|
v
[ snap-1 ] kro = 0, ro = 1
|
v
[ snap-2 ] kro = 0, ro = 1
|
v
[ active ] kro = 0, ro = 0
Signed-off-by: Jeff Cody <jcody@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2012-09-20 21:13:17 +02:00
|
|
|
bs->read_only = !(open_flags & BDRV_O_RDWR);
|
2011-10-27 11:54:27 +02:00
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
/* Open the image, either directly or using a protocol */
|
|
|
|
if (drv->bdrv_file_open) {
|
2012-11-12 17:35:27 +01:00
|
|
|
if (file != NULL) {
|
|
|
|
bdrv_swap(file, bs);
|
|
|
|
ret = 0;
|
|
|
|
} else {
|
|
|
|
ret = drv->bdrv_file_open(bs, filename, open_flags);
|
2010-04-14 14:17:38 +02:00
|
|
|
}
|
2012-11-12 17:35:27 +01:00
|
|
|
} else {
|
|
|
|
assert(file != NULL);
|
|
|
|
bs->file = file;
|
|
|
|
ret = drv->bdrv_open(bs, open_flags);
|
2010-04-14 14:17:38 +02:00
|
|
|
}
|
|
|
|
|
2010-04-14 15:24:50 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
goto free_and_fail;
|
|
|
|
}
|
|
|
|
|
2010-04-19 17:56:41 +02:00
|
|
|
ret = refresh_total_sectors(bs, bs->total_sectors);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto free_and_fail;
|
2010-04-14 15:24:50 +02:00
|
|
|
}
|
2010-04-19 17:56:41 +02:00
|
|
|
|
2010-04-14 15:24:50 +02:00
|
|
|
#ifndef _WIN32
|
|
|
|
if (bs->is_temporary) {
|
|
|
|
unlink(filename);
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
free_and_fail:
|
2012-11-12 17:35:27 +01:00
|
|
|
bs->file = NULL;
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(bs->opaque);
|
2010-04-14 15:24:50 +02:00
|
|
|
bs->opaque = NULL;
|
|
|
|
bs->drv = NULL;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2010-04-12 16:37:13 +02:00
|
|
|
/*
|
|
|
|
* Opens a file using a protocol (file, host_device, nbd, ...)
|
|
|
|
*/
|
2006-08-01 18:21:11 +02:00
|
|
|
int bdrv_file_open(BlockDriverState **pbs, const char *filename, int flags)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2006-08-01 18:21:11 +02:00
|
|
|
BlockDriverState *bs;
|
2010-04-05 16:53:57 +02:00
|
|
|
BlockDriver *drv;
|
2006-08-01 18:21:11 +02:00
|
|
|
int ret;
|
|
|
|
|
2010-05-26 04:35:36 +02:00
|
|
|
drv = bdrv_find_protocol(filename);
|
2010-04-05 16:53:57 +02:00
|
|
|
if (!drv) {
|
|
|
|
return -ENOENT;
|
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
bs = bdrv_new("");
|
2012-11-12 17:35:27 +01:00
|
|
|
ret = bdrv_open_common(bs, NULL, filename, flags, drv);
|
2006-08-01 18:21:11 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
bdrv_delete(bs);
|
|
|
|
return ret;
|
2005-10-30 19:30:10 +01:00
|
|
|
}
|
2009-03-03 18:37:16 +01:00
|
|
|
bs->growable = 1;
|
2006-08-01 18:21:11 +02:00
|
|
|
*pbs = bs;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-10-18 16:49:17 +02:00
|
|
|
int bdrv_open_backing_file(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
char backing_filename[PATH_MAX];
|
|
|
|
int back_flags, ret;
|
|
|
|
BlockDriver *back_drv = NULL;
|
|
|
|
|
|
|
|
if (bs->backing_hd != NULL) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->open_flags &= ~BDRV_O_NO_BACKING;
|
|
|
|
if (bs->backing_file[0] == '\0') {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->backing_hd = bdrv_new("");
|
|
|
|
bdrv_get_full_backing_filename(bs, backing_filename,
|
|
|
|
sizeof(backing_filename));
|
|
|
|
|
|
|
|
if (bs->backing_format[0] != '\0') {
|
|
|
|
back_drv = bdrv_find_format(bs->backing_format);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* backing files always opened read-only */
|
|
|
|
back_flags = bs->open_flags & ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT);
|
|
|
|
|
|
|
|
ret = bdrv_open(bs->backing_hd, backing_filename, back_flags, back_drv);
|
|
|
|
if (ret < 0) {
|
|
|
|
bdrv_delete(bs->backing_hd);
|
|
|
|
bs->backing_hd = NULL;
|
|
|
|
bs->open_flags |= BDRV_O_NO_BACKING;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2010-04-12 16:37:13 +02:00
|
|
|
/*
|
|
|
|
* Opens a disk image (raw, qcow2, vmdk, ...)
|
|
|
|
*/
|
2010-03-31 14:40:27 +02:00
|
|
|
int bdrv_open(BlockDriverState *bs, const char *filename, int flags,
|
|
|
|
BlockDriver *drv)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2010-04-12 16:37:13 +02:00
|
|
|
int ret;
|
2012-11-22 07:25:48 +01:00
|
|
|
/* TODO: extra byte is a hack to ensure MAX_PATH space on Windows. */
|
|
|
|
char tmp_filename[PATH_MAX + 1];
|
2012-11-12 17:35:27 +01:00
|
|
|
BlockDriverState *file = NULL;
|
2005-04-28 23:09:32 +02:00
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
if (flags & BDRV_O_SNAPSHOT) {
|
2004-08-01 23:59:26 +02:00
|
|
|
BlockDriverState *bs1;
|
|
|
|
int64_t total_size;
|
2008-09-12 19:54:13 +02:00
|
|
|
int is_protocol = 0;
|
2009-05-27 14:48:06 +02:00
|
|
|
BlockDriver *bdrv_qcow2;
|
|
|
|
QEMUOptionParameter *options;
|
2010-04-12 16:37:13 +02:00
|
|
|
char backing_filename[PATH_MAX];
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/* if snapshot, we create a temporary backing file and open it
|
|
|
|
instead of opening 'filename' directly */
|
2003-07-06 19:15:21 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/* if there is a backing file, use it */
|
|
|
|
bs1 = bdrv_new("");
|
2010-03-31 14:40:27 +02:00
|
|
|
ret = bdrv_open(bs1, filename, 0, drv);
|
2009-03-06 00:00:29 +01:00
|
|
|
if (ret < 0) {
|
2004-08-01 23:59:26 +02:00
|
|
|
bdrv_delete(bs1);
|
2009-03-06 00:00:29 +01:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2010-05-27 16:20:30 +02:00
|
|
|
total_size = bdrv_getlength(bs1) & BDRV_SECTOR_MASK;
|
2008-09-12 19:54:13 +02:00
|
|
|
|
|
|
|
if (bs1->drv && bs1->drv->protocol_name)
|
|
|
|
is_protocol = 1;
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
bdrv_delete(bs1);
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2012-05-28 09:27:54 +02:00
|
|
|
ret = get_tmp_filename(tmp_filename, sizeof(tmp_filename));
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
2008-09-12 19:54:13 +02:00
|
|
|
|
|
|
|
/* Real path is meaningless for protocols */
|
|
|
|
if (is_protocol)
|
|
|
|
snprintf(backing_filename, sizeof(backing_filename),
|
|
|
|
"%s", filename);
|
2009-12-25 19:19:22 +01:00
|
|
|
else if (!realpath(filename, backing_filename))
|
|
|
|
return -errno;
|
2008-09-12 19:54:13 +02:00
|
|
|
|
2009-05-27 14:48:06 +02:00
|
|
|
bdrv_qcow2 = bdrv_find_format("qcow2");
|
|
|
|
options = parse_option_parameters("", bdrv_qcow2->create_options, NULL);
|
|
|
|
|
2010-05-27 16:20:30 +02:00
|
|
|
set_option_parameter_int(options, BLOCK_OPT_SIZE, total_size);
|
2009-05-27 14:48:06 +02:00
|
|
|
set_option_parameter(options, BLOCK_OPT_BACKING_FILE, backing_filename);
|
|
|
|
if (drv) {
|
|
|
|
set_option_parameter(options, BLOCK_OPT_BACKING_FMT,
|
|
|
|
drv->format_name);
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = bdrv_create(bdrv_qcow2, tmp_filename, options);
|
2010-04-29 18:24:50 +02:00
|
|
|
free_option_parameters(options);
|
2009-03-06 00:00:29 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2009-05-27 14:48:06 +02:00
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
filename = tmp_filename;
|
2009-05-27 14:48:06 +02:00
|
|
|
drv = bdrv_qcow2;
|
2004-08-01 23:59:26 +02:00
|
|
|
bs->is_temporary = 1;
|
|
|
|
}
|
2005-04-28 23:09:32 +02:00
|
|
|
|
2012-11-12 17:35:27 +01:00
|
|
|
/* Open image file without format layer */
|
|
|
|
if (flags & BDRV_O_RDWR) {
|
|
|
|
flags |= BDRV_O_ALLOW_RDWR;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = bdrv_file_open(&file, filename, bdrv_open_flags(bs, flags));
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2010-04-12 16:37:13 +02:00
|
|
|
/* Find the right image format driver */
|
2010-04-05 16:53:57 +02:00
|
|
|
if (!drv) {
|
2012-11-12 17:35:27 +01:00
|
|
|
ret = find_image_format(file, filename, &drv);
|
2009-03-06 00:00:29 +01:00
|
|
|
}
|
2010-01-20 18:13:25 +01:00
|
|
|
|
2009-03-06 00:00:29 +01:00
|
|
|
if (!drv) {
|
|
|
|
goto unlink_and_fail;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2010-04-12 16:37:13 +02:00
|
|
|
|
|
|
|
/* Open the image */
|
2012-11-12 17:35:27 +01:00
|
|
|
ret = bdrv_open_common(bs, file, filename, flags, drv);
|
2010-04-12 16:37:13 +02:00
|
|
|
if (ret < 0) {
|
2010-01-20 18:13:25 +01:00
|
|
|
goto unlink_and_fail;
|
|
|
|
}
|
|
|
|
|
2012-11-12 17:35:27 +01:00
|
|
|
if (bs->file != file) {
|
|
|
|
bdrv_delete(file);
|
|
|
|
file = NULL;
|
|
|
|
}
|
|
|
|
|
2010-04-12 16:37:13 +02:00
|
|
|
/* If there is a backing file, use it */
|
2012-10-18 16:49:17 +02:00
|
|
|
if ((flags & BDRV_O_NO_BACKING) == 0) {
|
|
|
|
ret = bdrv_open_backing_file(bs);
|
2010-04-12 16:37:13 +02:00
|
|
|
if (ret < 0) {
|
|
|
|
bdrv_close(bs);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!bdrv_key_required(bs)) {
|
2011-09-06 18:58:59 +02:00
|
|
|
bdrv_dev_change_media_cb(bs, true);
|
2010-04-12 16:37:13 +02:00
|
|
|
}
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
/* throttling disk I/O limits */
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
bdrv_io_limits_enable(bs);
|
|
|
|
}
|
|
|
|
|
2010-04-12 16:37:13 +02:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
unlink_and_fail:
|
2012-11-12 17:35:27 +01:00
|
|
|
if (file != NULL) {
|
|
|
|
bdrv_delete(file);
|
|
|
|
}
|
2010-04-12 16:37:13 +02:00
|
|
|
if (bs->is_temporary) {
|
|
|
|
unlink(filename);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-09-20 21:13:19 +02:00
|
|
|
typedef struct BlockReopenQueueEntry {
|
|
|
|
bool prepared;
|
|
|
|
BDRVReopenState state;
|
|
|
|
QSIMPLEQ_ENTRY(BlockReopenQueueEntry) entry;
|
|
|
|
} BlockReopenQueueEntry;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Adds a BlockDriverState to a simple queue for an atomic, transactional
|
|
|
|
* reopen of multiple devices.
|
|
|
|
*
|
|
|
|
* bs_queue can either be an existing BlockReopenQueue that has had QSIMPLE_INIT
|
|
|
|
* already performed, or alternatively may be NULL a new BlockReopenQueue will
|
|
|
|
* be created and initialized. This newly created BlockReopenQueue should be
|
|
|
|
* passed back in for subsequent calls that are intended to be of the same
|
|
|
|
* atomic 'set'.
|
|
|
|
*
|
|
|
|
* bs is the BlockDriverState to add to the reopen queue.
|
|
|
|
*
|
|
|
|
* flags contains the open flags for the associated bs
|
|
|
|
*
|
|
|
|
* returns a pointer to bs_queue, which is either the newly allocated
|
|
|
|
* bs_queue, or the existing bs_queue being used.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
BlockReopenQueue *bdrv_reopen_queue(BlockReopenQueue *bs_queue,
|
|
|
|
BlockDriverState *bs, int flags)
|
|
|
|
{
|
|
|
|
assert(bs != NULL);
|
|
|
|
|
|
|
|
BlockReopenQueueEntry *bs_entry;
|
|
|
|
if (bs_queue == NULL) {
|
|
|
|
bs_queue = g_new0(BlockReopenQueue, 1);
|
|
|
|
QSIMPLEQ_INIT(bs_queue);
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bs->file) {
|
|
|
|
bdrv_reopen_queue(bs_queue, bs->file, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
bs_entry = g_new0(BlockReopenQueueEntry, 1);
|
|
|
|
QSIMPLEQ_INSERT_TAIL(bs_queue, bs_entry, entry);
|
|
|
|
|
|
|
|
bs_entry->state.bs = bs;
|
|
|
|
bs_entry->state.flags = flags;
|
|
|
|
|
|
|
|
return bs_queue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Reopen multiple BlockDriverStates atomically & transactionally.
|
|
|
|
*
|
|
|
|
* The queue passed in (bs_queue) must have been built up previous
|
|
|
|
* via bdrv_reopen_queue().
|
|
|
|
*
|
|
|
|
* Reopens all BDS specified in the queue, with the appropriate
|
|
|
|
* flags. All devices are prepared for reopen, and failure of any
|
|
|
|
* device will cause all device changes to be abandonded, and intermediate
|
|
|
|
* data cleaned up.
|
|
|
|
*
|
|
|
|
* If all devices prepare successfully, then the changes are committed
|
|
|
|
* to all devices.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int bdrv_reopen_multiple(BlockReopenQueue *bs_queue, Error **errp)
|
|
|
|
{
|
|
|
|
int ret = -1;
|
|
|
|
BlockReopenQueueEntry *bs_entry, *next;
|
|
|
|
Error *local_err = NULL;
|
|
|
|
|
|
|
|
assert(bs_queue != NULL);
|
|
|
|
|
|
|
|
bdrv_drain_all();
|
|
|
|
|
|
|
|
QSIMPLEQ_FOREACH(bs_entry, bs_queue, entry) {
|
|
|
|
if (bdrv_reopen_prepare(&bs_entry->state, bs_queue, &local_err)) {
|
|
|
|
error_propagate(errp, local_err);
|
|
|
|
goto cleanup;
|
|
|
|
}
|
|
|
|
bs_entry->prepared = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If we reach this point, we have success and just need to apply the
|
|
|
|
* changes
|
|
|
|
*/
|
|
|
|
QSIMPLEQ_FOREACH(bs_entry, bs_queue, entry) {
|
|
|
|
bdrv_reopen_commit(&bs_entry->state);
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
cleanup:
|
|
|
|
QSIMPLEQ_FOREACH_SAFE(bs_entry, bs_queue, entry, next) {
|
|
|
|
if (ret && bs_entry->prepared) {
|
|
|
|
bdrv_reopen_abort(&bs_entry->state);
|
|
|
|
}
|
|
|
|
g_free(bs_entry);
|
|
|
|
}
|
|
|
|
g_free(bs_queue);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/* Reopen a single BlockDriverState with the specified flags. */
|
|
|
|
int bdrv_reopen(BlockDriverState *bs, int bdrv_flags, Error **errp)
|
|
|
|
{
|
|
|
|
int ret = -1;
|
|
|
|
Error *local_err = NULL;
|
|
|
|
BlockReopenQueue *queue = bdrv_reopen_queue(NULL, bs, bdrv_flags);
|
|
|
|
|
|
|
|
ret = bdrv_reopen_multiple(queue, &local_err);
|
|
|
|
if (local_err != NULL) {
|
|
|
|
error_propagate(errp, local_err);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Prepares a BlockDriverState for reopen. All changes are staged in the
|
|
|
|
* 'opaque' field of the BDRVReopenState, which is used and allocated by
|
|
|
|
* the block driver layer .bdrv_reopen_prepare()
|
|
|
|
*
|
|
|
|
* bs is the BlockDriverState to reopen
|
|
|
|
* flags are the new open flags
|
|
|
|
* queue is the reopen queue
|
|
|
|
*
|
|
|
|
* Returns 0 on success, non-zero on error. On error errp will be set
|
|
|
|
* as well.
|
|
|
|
*
|
|
|
|
* On failure, bdrv_reopen_abort() will be called to clean up any data.
|
|
|
|
* It is the responsibility of the caller to then call the abort() or
|
|
|
|
* commit() for any other BDS that have been left in a prepare() state
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int bdrv_reopen_prepare(BDRVReopenState *reopen_state, BlockReopenQueue *queue,
|
|
|
|
Error **errp)
|
|
|
|
{
|
|
|
|
int ret = -1;
|
|
|
|
Error *local_err = NULL;
|
|
|
|
BlockDriver *drv;
|
|
|
|
|
|
|
|
assert(reopen_state != NULL);
|
|
|
|
assert(reopen_state->bs->drv != NULL);
|
|
|
|
drv = reopen_state->bs->drv;
|
|
|
|
|
|
|
|
/* if we are to stay read-only, do not allow permission change
|
|
|
|
* to r/w */
|
|
|
|
if (!(reopen_state->bs->open_flags & BDRV_O_ALLOW_RDWR) &&
|
|
|
|
reopen_state->flags & BDRV_O_RDWR) {
|
|
|
|
error_set(errp, QERR_DEVICE_IS_READ_ONLY,
|
|
|
|
reopen_state->bs->device_name);
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
ret = bdrv_flush(reopen_state->bs);
|
|
|
|
if (ret) {
|
|
|
|
error_set(errp, ERROR_CLASS_GENERIC_ERROR, "Error (%s) flushing drive",
|
|
|
|
strerror(-ret));
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (drv->bdrv_reopen_prepare) {
|
|
|
|
ret = drv->bdrv_reopen_prepare(reopen_state, queue, &local_err);
|
|
|
|
if (ret) {
|
|
|
|
if (local_err != NULL) {
|
|
|
|
error_propagate(errp, local_err);
|
|
|
|
} else {
|
|
|
|
error_set(errp, QERR_OPEN_FILE_FAILED,
|
|
|
|
reopen_state->bs->filename);
|
|
|
|
}
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/* It is currently mandatory to have a bdrv_reopen_prepare()
|
|
|
|
* handler for each supported drv. */
|
|
|
|
error_set(errp, QERR_BLOCK_FORMAT_FEATURE_NOT_SUPPORTED,
|
|
|
|
drv->format_name, reopen_state->bs->device_name,
|
|
|
|
"reopening of file");
|
|
|
|
ret = -1;
|
|
|
|
goto error;
|
|
|
|
}
|
|
|
|
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
error:
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Takes the staged changes for the reopen from bdrv_reopen_prepare(), and
|
|
|
|
* makes them final by swapping the staging BlockDriverState contents into
|
|
|
|
* the active BlockDriverState contents.
|
|
|
|
*/
|
|
|
|
void bdrv_reopen_commit(BDRVReopenState *reopen_state)
|
|
|
|
{
|
|
|
|
BlockDriver *drv;
|
|
|
|
|
|
|
|
assert(reopen_state != NULL);
|
|
|
|
drv = reopen_state->bs->drv;
|
|
|
|
assert(drv != NULL);
|
|
|
|
|
|
|
|
/* If there are any driver level actions to take */
|
|
|
|
if (drv->bdrv_reopen_commit) {
|
|
|
|
drv->bdrv_reopen_commit(reopen_state);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* set BDS specific flags now */
|
|
|
|
reopen_state->bs->open_flags = reopen_state->flags;
|
|
|
|
reopen_state->bs->enable_write_cache = !!(reopen_state->flags &
|
|
|
|
BDRV_O_CACHE_WB);
|
|
|
|
reopen_state->bs->read_only = !(reopen_state->flags & BDRV_O_RDWR);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Abort the reopen, and delete and free the staged changes in
|
|
|
|
* reopen_state
|
|
|
|
*/
|
|
|
|
void bdrv_reopen_abort(BDRVReopenState *reopen_state)
|
|
|
|
{
|
|
|
|
BlockDriver *drv;
|
|
|
|
|
|
|
|
assert(reopen_state != NULL);
|
|
|
|
drv = reopen_state->bs->drv;
|
|
|
|
assert(drv != NULL);
|
|
|
|
|
|
|
|
if (drv->bdrv_reopen_abort) {
|
|
|
|
drv->bdrv_reopen_abort(reopen_state);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2003-06-30 12:03:06 +02:00
|
|
|
void bdrv_close(BlockDriverState *bs)
|
|
|
|
{
|
2012-04-20 11:10:56 +02:00
|
|
|
bdrv_flush(bs);
|
2012-10-19 11:36:48 +02:00
|
|
|
if (bs->job) {
|
|
|
|
block_job_cancel_sync(bs->job);
|
|
|
|
}
|
|
|
|
bdrv_drain_all();
|
2012-08-23 11:20:36 +02:00
|
|
|
notifier_list_notify(&bs->close_notifiers, bs);
|
2012-04-11 11:06:37 +02:00
|
|
|
|
2012-10-19 11:36:48 +02:00
|
|
|
if (bs->drv) {
|
2010-06-25 10:33:39 +02:00
|
|
|
if (bs == bs_snapshots) {
|
|
|
|
bs_snapshots = NULL;
|
|
|
|
}
|
2010-04-17 11:49:06 +02:00
|
|
|
if (bs->backing_hd) {
|
2004-08-01 23:59:26 +02:00
|
|
|
bdrv_delete(bs->backing_hd);
|
2010-04-17 11:49:06 +02:00
|
|
|
bs->backing_hd = NULL;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
bs->drv->bdrv_close(bs);
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(bs->opaque);
|
2004-08-01 23:59:26 +02:00
|
|
|
#ifdef _WIN32
|
|
|
|
if (bs->is_temporary) {
|
|
|
|
unlink(bs->filename);
|
|
|
|
}
|
2004-04-01 01:37:16 +02:00
|
|
|
#endif
|
2004-08-01 23:59:26 +02:00
|
|
|
bs->opaque = NULL;
|
|
|
|
bs->drv = NULL;
|
2011-11-28 17:08:47 +01:00
|
|
|
bs->copy_on_read = 0;
|
2012-05-08 16:51:43 +02:00
|
|
|
bs->backing_file[0] = '\0';
|
|
|
|
bs->backing_format[0] = '\0';
|
2012-05-08 16:51:49 +02:00
|
|
|
bs->total_sectors = 0;
|
|
|
|
bs->encrypted = 0;
|
|
|
|
bs->valid_key = 0;
|
|
|
|
bs->sg = 0;
|
|
|
|
bs->growable = 0;
|
2004-03-14 22:38:54 +01:00
|
|
|
|
2010-04-14 14:17:38 +02:00
|
|
|
if (bs->file != NULL) {
|
2012-05-08 16:51:44 +02:00
|
|
|
bdrv_delete(bs->file);
|
|
|
|
bs->file = NULL;
|
2010-04-14 14:17:38 +02:00
|
|
|
}
|
2004-03-14 22:38:54 +01:00
|
|
|
}
|
2011-11-08 06:00:14 +01:00
|
|
|
|
2012-08-09 12:44:48 +02:00
|
|
|
bdrv_dev_change_media_cb(bs, false);
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
/*throttling disk I/O limits*/
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
bdrv_io_limits_disable(bs);
|
|
|
|
}
|
2004-03-14 22:38:54 +01:00
|
|
|
}
|
|
|
|
|
2010-05-28 04:44:57 +02:00
|
|
|
void bdrv_close_all(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
bdrv_close(bs);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-11-30 13:23:43 +01:00
|
|
|
/*
|
|
|
|
* Wait for pending requests to complete across all BlockDriverStates
|
|
|
|
*
|
|
|
|
* This function does not flush data to disk, use bdrv_flush_all() for that
|
|
|
|
* after calling this function.
|
2012-04-12 14:00:57 +02:00
|
|
|
*
|
|
|
|
* Note that completion of an asynchronous I/O operation can trigger any
|
|
|
|
* number of other I/O operations on other devices---for example a coroutine
|
|
|
|
* can be arbitrarily complex and a constant flow of I/O can come until the
|
|
|
|
* coroutine is complete. Because of this, it is not possible to have a
|
|
|
|
* function to drain a single device's I/O queue.
|
2011-11-30 13:23:43 +01:00
|
|
|
*/
|
|
|
|
void bdrv_drain_all(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
2012-04-12 14:00:57 +02:00
|
|
|
bool busy;
|
|
|
|
|
|
|
|
do {
|
|
|
|
busy = qemu_aio_wait();
|
2011-11-30 13:23:43 +01:00
|
|
|
|
2012-04-12 14:00:57 +02:00
|
|
|
/* FIXME: We do not have timer support here, so this is effectively
|
|
|
|
* a busy wait.
|
|
|
|
*/
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
if (!qemu_co_queue_empty(&bs->throttled_reqs)) {
|
|
|
|
qemu_co_queue_restart_all(&bs->throttled_reqs);
|
|
|
|
busy = true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
} while (busy);
|
2011-11-30 13:23:43 +01:00
|
|
|
|
|
|
|
/* If requests are still pending there is a bug somewhere */
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
assert(QLIST_EMPTY(&bs->tracked_requests));
|
|
|
|
assert(qemu_co_queue_empty(&bs->throttled_reqs));
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
Do not delete BlockDriverState when deleting the drive
When removing a drive from the host-side via drive_del we currently have
the following path:
drive_del
qemu_aio_flush()
bdrv_close() // zaps bs->drv, which makes any subsequent I/O get
// dropped. Works as designed
drive_uninit()
bdrv_delete() // frees the bs. Since the device is still connected to
// bs, any subsequent I/O is a use-after-free.
The value of bs->drv becomes unpredictable on free. As long as it
remains null, I/O still gets dropped, however it could become non-null
at any point after the free resulting SEGVs or other QEMU state
corruption.
To resolve this issue as simply as possible, we can chose to not
actually delete the BlockDriverState pointer. Since bdrv_close()
handles setting the drv pointer to NULL, we just need to remove the
BlockDriverState from the QLIST that is used to enumerate the block
devices. This is currently handled within bdrv_delete, so move this
into its own function, bdrv_make_anon().
The result is that we can now invoke drive_del, this closes the file
descriptors and sets BlockDriverState->drv to NULL which prevents futher
IO to the device, and since we do not free BlockDriverState, we don't
have to worry about the copy retained in the block devices.
We also don't attempt to remove the qdev property since we are no longer
deleting the BlockDriverState on drives with associated drives. This
also allows for removing Drives with no devices associated either.
Reported-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Ryan Harper <ryanh@us.ibm.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2011-03-30 03:51:47 +02:00
|
|
|
/* make a BlockDriverState anonymous by removing from bdrv_state list.
|
|
|
|
Also, NULL terminate the device_name to prevent double remove */
|
|
|
|
void bdrv_make_anon(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->device_name[0] != '\0') {
|
|
|
|
QTAILQ_REMOVE(&bdrv_states, bs, list);
|
|
|
|
}
|
|
|
|
bs->device_name[0] = '\0';
|
|
|
|
}
|
|
|
|
|
2012-05-08 16:51:41 +02:00
|
|
|
static void bdrv_rebind(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->drv && bs->drv->bdrv_rebind) {
|
|
|
|
bs->drv->bdrv_rebind(bs);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
static void bdrv_move_feature_fields(BlockDriverState *bs_dest,
|
|
|
|
BlockDriverState *bs_src)
|
2012-02-28 21:54:06 +01:00
|
|
|
{
|
2012-06-14 16:55:02 +02:00
|
|
|
/* move some fields that need to stay attached to the device */
|
|
|
|
bs_dest->open_flags = bs_src->open_flags;
|
2012-02-28 21:54:06 +01:00
|
|
|
|
|
|
|
/* dev info */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->dev_ops = bs_src->dev_ops;
|
|
|
|
bs_dest->dev_opaque = bs_src->dev_opaque;
|
|
|
|
bs_dest->dev = bs_src->dev;
|
|
|
|
bs_dest->buffer_alignment = bs_src->buffer_alignment;
|
|
|
|
bs_dest->copy_on_read = bs_src->copy_on_read;
|
2012-02-28 21:54:06 +01:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->enable_write_cache = bs_src->enable_write_cache;
|
2012-06-06 00:04:51 +02:00
|
|
|
|
2012-02-28 21:54:06 +01:00
|
|
|
/* i/o timing parameters */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->slice_time = bs_src->slice_time;
|
|
|
|
bs_dest->slice_start = bs_src->slice_start;
|
|
|
|
bs_dest->slice_end = bs_src->slice_end;
|
|
|
|
bs_dest->io_limits = bs_src->io_limits;
|
|
|
|
bs_dest->io_base = bs_src->io_base;
|
|
|
|
bs_dest->throttled_reqs = bs_src->throttled_reqs;
|
|
|
|
bs_dest->block_timer = bs_src->block_timer;
|
|
|
|
bs_dest->io_limits_enabled = bs_src->io_limits_enabled;
|
2012-02-28 21:54:06 +01:00
|
|
|
|
|
|
|
/* r/w error */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->on_read_error = bs_src->on_read_error;
|
|
|
|
bs_dest->on_write_error = bs_src->on_write_error;
|
2012-02-28 21:54:06 +01:00
|
|
|
|
|
|
|
/* i/o status */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->iostatus_enabled = bs_src->iostatus_enabled;
|
|
|
|
bs_dest->iostatus = bs_src->iostatus;
|
2012-02-28 21:54:06 +01:00
|
|
|
|
2012-06-14 16:55:01 +02:00
|
|
|
/* dirty bitmap */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->dirty_count = bs_src->dirty_count;
|
|
|
|
bs_dest->dirty_bitmap = bs_src->dirty_bitmap;
|
2012-06-14 16:55:01 +02:00
|
|
|
|
|
|
|
/* job */
|
2012-06-14 16:55:02 +02:00
|
|
|
bs_dest->in_use = bs_src->in_use;
|
|
|
|
bs_dest->job = bs_src->job;
|
2012-06-14 16:55:01 +02:00
|
|
|
|
2012-02-28 21:54:06 +01:00
|
|
|
/* keep the same entry in bdrv_states */
|
2012-06-14 16:55:02 +02:00
|
|
|
pstrcpy(bs_dest->device_name, sizeof(bs_dest->device_name),
|
|
|
|
bs_src->device_name);
|
|
|
|
bs_dest->list = bs_src->list;
|
|
|
|
}
|
2012-02-28 21:54:06 +01:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
/*
|
|
|
|
* Swap bs contents for two image chains while they are live,
|
|
|
|
* while keeping required fields on the BlockDriverState that is
|
|
|
|
* actually attached to a device.
|
|
|
|
*
|
|
|
|
* This will modify the BlockDriverState fields, and swap contents
|
|
|
|
* between bs_new and bs_old. Both bs_new and bs_old are modified.
|
|
|
|
*
|
|
|
|
* bs_new is required to be anonymous.
|
|
|
|
*
|
|
|
|
* This function does not create any image files.
|
|
|
|
*/
|
|
|
|
void bdrv_swap(BlockDriverState *bs_new, BlockDriverState *bs_old)
|
|
|
|
{
|
|
|
|
BlockDriverState tmp;
|
2012-03-27 22:30:19 +02:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
/* bs_new must be anonymous and shouldn't have anything fancy enabled */
|
|
|
|
assert(bs_new->device_name[0] == '\0');
|
|
|
|
assert(bs_new->dirty_bitmap == NULL);
|
|
|
|
assert(bs_new->job == NULL);
|
|
|
|
assert(bs_new->dev == NULL);
|
|
|
|
assert(bs_new->in_use == 0);
|
|
|
|
assert(bs_new->io_limits_enabled == false);
|
|
|
|
assert(bs_new->block_timer == NULL);
|
2012-02-28 21:54:06 +01:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
tmp = *bs_new;
|
|
|
|
*bs_new = *bs_old;
|
|
|
|
*bs_old = tmp;
|
2012-06-14 16:55:01 +02:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
/* there are some fields that should not be swapped, move them back */
|
|
|
|
bdrv_move_feature_fields(&tmp, bs_old);
|
|
|
|
bdrv_move_feature_fields(bs_old, bs_new);
|
|
|
|
bdrv_move_feature_fields(bs_new, &tmp);
|
2012-02-28 21:54:06 +01:00
|
|
|
|
2012-06-14 16:55:02 +02:00
|
|
|
/* bs_new shouldn't be in bdrv_states even after the swap! */
|
|
|
|
assert(bs_new->device_name[0] == '\0');
|
|
|
|
|
|
|
|
/* Check a few fields that should remain attached to the device */
|
|
|
|
assert(bs_new->dev == NULL);
|
|
|
|
assert(bs_new->job == NULL);
|
|
|
|
assert(bs_new->in_use == 0);
|
|
|
|
assert(bs_new->io_limits_enabled == false);
|
|
|
|
assert(bs_new->block_timer == NULL);
|
2012-05-08 16:51:41 +02:00
|
|
|
|
|
|
|
bdrv_rebind(bs_new);
|
2012-06-14 16:55:02 +02:00
|
|
|
bdrv_rebind(bs_old);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Add new bs contents at the top of an image chain while the chain is
|
|
|
|
* live, while keeping required fields on the top layer.
|
|
|
|
*
|
|
|
|
* This will modify the BlockDriverState fields, and swap contents
|
|
|
|
* between bs_new and bs_top. Both bs_new and bs_top are modified.
|
|
|
|
*
|
|
|
|
* bs_new is required to be anonymous.
|
|
|
|
*
|
|
|
|
* This function does not create any image files.
|
|
|
|
*/
|
|
|
|
void bdrv_append(BlockDriverState *bs_new, BlockDriverState *bs_top)
|
|
|
|
{
|
|
|
|
bdrv_swap(bs_new, bs_top);
|
|
|
|
|
|
|
|
/* The contents of 'tmp' will become bs_top, as we are
|
|
|
|
* swapping bs_new and bs_top contents. */
|
|
|
|
bs_top->backing_hd = bs_new;
|
|
|
|
bs_top->open_flags &= ~BDRV_O_NO_BACKING;
|
|
|
|
pstrcpy(bs_top->backing_file, sizeof(bs_top->backing_file),
|
|
|
|
bs_new->filename);
|
|
|
|
pstrcpy(bs_top->backing_format, sizeof(bs_top->backing_format),
|
|
|
|
bs_new->drv ? bs_new->drv->format_name : "");
|
2012-02-28 21:54:06 +01:00
|
|
|
}
|
|
|
|
|
2004-03-14 22:38:54 +01:00
|
|
|
void bdrv_delete(BlockDriverState *bs)
|
|
|
|
{
|
2011-08-03 15:07:40 +02:00
|
|
|
assert(!bs->dev);
|
2012-03-30 13:17:11 +02:00
|
|
|
assert(!bs->job);
|
|
|
|
assert(!bs->in_use);
|
2010-06-29 16:58:30 +02:00
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
/* remove from list, if necessary */
|
Do not delete BlockDriverState when deleting the drive
When removing a drive from the host-side via drive_del we currently have
the following path:
drive_del
qemu_aio_flush()
bdrv_close() // zaps bs->drv, which makes any subsequent I/O get
// dropped. Works as designed
drive_uninit()
bdrv_delete() // frees the bs. Since the device is still connected to
// bs, any subsequent I/O is a use-after-free.
The value of bs->drv becomes unpredictable on free. As long as it
remains null, I/O still gets dropped, however it could become non-null
at any point after the free resulting SEGVs or other QEMU state
corruption.
To resolve this issue as simply as possible, we can chose to not
actually delete the BlockDriverState pointer. Since bdrv_close()
handles setting the drv pointer to NULL, we just need to remove the
BlockDriverState from the QLIST that is used to enumerate the block
devices. This is currently handled within bdrv_delete, so move this
into its own function, bdrv_make_anon().
The result is that we can now invoke drive_del, this closes the file
descriptors and sets BlockDriverState->drv to NULL which prevents futher
IO to the device, and since we do not free BlockDriverState, we don't
have to worry about the copy retained in the block devices.
We also don't attempt to remove the qdev property since we are no longer
deleting the BlockDriverState on drives with associated drives. This
also allows for removing Drives with no devices associated either.
Reported-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Ryan Harper <ryanh@us.ibm.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2011-03-30 03:51:47 +02:00
|
|
|
bdrv_make_anon(bs);
|
2008-04-08 21:51:21 +02:00
|
|
|
|
2004-03-14 22:38:54 +01:00
|
|
|
bdrv_close(bs);
|
2010-04-14 14:17:38 +02:00
|
|
|
|
2010-06-25 10:33:39 +02:00
|
|
|
assert(bs != bs_snapshots);
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(bs);
|
2003-06-30 12:03:06 +02:00
|
|
|
}
|
|
|
|
|
2011-08-03 15:07:40 +02:00
|
|
|
int bdrv_attach_dev(BlockDriverState *bs, void *dev)
|
|
|
|
/* TODO change to DeviceState *dev when all users are qdevified */
|
2010-06-29 16:58:30 +02:00
|
|
|
{
|
2011-08-03 15:07:40 +02:00
|
|
|
if (bs->dev) {
|
2010-06-29 16:58:30 +02:00
|
|
|
return -EBUSY;
|
|
|
|
}
|
2011-08-03 15:07:40 +02:00
|
|
|
bs->dev = dev;
|
2011-09-26 22:43:50 +02:00
|
|
|
bdrv_iostatus_reset(bs);
|
2010-06-29 16:58:30 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2011-08-03 15:07:40 +02:00
|
|
|
/* TODO qdevified devices don't use this, remove when devices are qdevified */
|
|
|
|
void bdrv_attach_dev_nofail(BlockDriverState *bs, void *dev)
|
2010-06-29 16:58:30 +02:00
|
|
|
{
|
2011-08-03 15:07:40 +02:00
|
|
|
if (bdrv_attach_dev(bs, dev) < 0) {
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_detach_dev(BlockDriverState *bs, void *dev)
|
|
|
|
/* TODO change to DeviceState *dev when all users are qdevified */
|
|
|
|
{
|
|
|
|
assert(bs->dev == dev);
|
|
|
|
bs->dev = NULL;
|
2011-08-03 15:07:41 +02:00
|
|
|
bs->dev_ops = NULL;
|
|
|
|
bs->dev_opaque = NULL;
|
2011-09-06 18:58:57 +02:00
|
|
|
bs->buffer_alignment = 512;
|
2010-06-29 16:58:30 +02:00
|
|
|
}
|
|
|
|
|
2011-08-03 15:07:40 +02:00
|
|
|
/* TODO change to return DeviceState * when all users are qdevified */
|
|
|
|
void *bdrv_get_attached_dev(BlockDriverState *bs)
|
2010-06-29 16:58:30 +02:00
|
|
|
{
|
2011-08-03 15:07:40 +02:00
|
|
|
return bs->dev;
|
2010-06-29 16:58:30 +02:00
|
|
|
}
|
|
|
|
|
2011-08-03 15:07:41 +02:00
|
|
|
void bdrv_set_dev_ops(BlockDriverState *bs, const BlockDevOps *ops,
|
|
|
|
void *opaque)
|
|
|
|
{
|
|
|
|
bs->dev_ops = ops;
|
|
|
|
bs->dev_opaque = opaque;
|
block: Clean up remaining users of "removable"
BlockDriverState member removable is a confused mess. It is true when
an ide-cd, scsi-cd or floppy qdev is attached, or when the
BlockDriverState was created with -drive if={floppy,sd} or -drive
if={ide,scsi,xen,none},media=cdrom ("created removable"), except when
an ide-hd, scsi-hd, scsi-generic or virtio-blk qdev is attached.
Three users remain:
1. eject_device(), via bdrv_is_removable() uses it to determine
whether a block device can eject media.
2. bdrv_info() is monitor command "info block". QMP documentation
says "true if the device is removable, false otherwise". From the
monitor user's point of view, the only sensible interpretation of
"is removable" is "can eject media with monitor commands eject and
change".
A block device can eject media unless a device is attached that
doesn't support it. Switch the two users over to new
bdrv_dev_has_removable_media() that returns exactly that.
3. bdrv_getlength() uses to suppress its length cache when media can
change (see commit 46a4e4e6). Media change is either monitor
command change (updates the length cache), monitor command eject
(doesn't update the length cache, easily fixable), or physical
media change (invalidates length cache, not so easily fixable).
I'm refraining from improving anything here, because this series is
long enough already. Instead, I simply switch it over to
bdrv_dev_has_removable_media() as well.
This changes the behavior of the length cache and of monitor commands
eject and change in two cases:
a. drive not created removable, no device attached
The commit makes the drive removable, and defeats the length cache.
Example: -drive if=none
b. drive created removable, but the attached drive is non-removable,
and doesn't call bdrv_set_removable(..., 0) (most devices don't)
The commit makes the drive non-removable, and enables the length
cache.
Example: -drive if=xen,media=cdrom -M xenpv
The other non-removable devices that don't call
bdrv_set_removable() can't currently use a drive created removable,
either because they aren't qdevified, or because they lack a drive
property. Won't stay that way.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2011-09-06 18:58:51 +02:00
|
|
|
if (bdrv_dev_has_removable_media(bs) && bs == bs_snapshots) {
|
|
|
|
bs_snapshots = NULL;
|
|
|
|
}
|
2011-08-03 15:07:41 +02:00
|
|
|
}
|
|
|
|
|
block: introduce block job error
The following behaviors are possible:
'report': The behavior is the same as in 1.1. An I/O error,
respectively during a read or a write, will complete the job immediately
with an error code.
'ignore': An I/O error, respectively during a read or a write, will be
ignored. For streaming, the job will complete with an error and the
backing file will be left in place. For mirroring, the sector will be
marked again as dirty and re-examined later.
'stop': The job will be paused and the job iostatus will be set to
failed or nospace, while the VM will keep running. This can only be
specified if the block device has rerror=stop and werror=stop or enospc.
'enospc': Behaves as 'stop' for ENOSPC errors, 'report' for others.
In all cases, even for 'report', the I/O error is reported as a QMP
event BLOCK_JOB_ERROR, with the same arguments as BLOCK_IO_ERROR.
It is possible that while stopping the VM a BLOCK_IO_ERROR event will be
reported and will clobber the event from BLOCK_JOB_ERROR, or vice versa.
This is not really avoidable since stopping the VM completes all pending
I/O requests. In fact, it is already possible now that a series of
BLOCK_IO_ERROR events are reported with rerror=stop, because vm_stop
calls bdrv_drain_all and this can generate further errors.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2012-09-28 17:22:58 +02:00
|
|
|
void bdrv_emit_qmp_error_event(const BlockDriverState *bdrv,
|
|
|
|
enum MonitorEvent ev,
|
|
|
|
BlockErrorAction action, bool is_read)
|
2012-01-25 19:59:43 +01:00
|
|
|
{
|
|
|
|
QObject *data;
|
|
|
|
const char *action_str;
|
|
|
|
|
|
|
|
switch (action) {
|
|
|
|
case BDRV_ACTION_REPORT:
|
|
|
|
action_str = "report";
|
|
|
|
break;
|
|
|
|
case BDRV_ACTION_IGNORE:
|
|
|
|
action_str = "ignore";
|
|
|
|
break;
|
|
|
|
case BDRV_ACTION_STOP:
|
|
|
|
action_str = "stop";
|
|
|
|
break;
|
|
|
|
default:
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
|
|
|
|
data = qobject_from_jsonf("{ 'device': %s, 'action': %s, 'operation': %s }",
|
|
|
|
bdrv->device_name,
|
|
|
|
action_str,
|
|
|
|
is_read ? "read" : "write");
|
block: introduce block job error
The following behaviors are possible:
'report': The behavior is the same as in 1.1. An I/O error,
respectively during a read or a write, will complete the job immediately
with an error code.
'ignore': An I/O error, respectively during a read or a write, will be
ignored. For streaming, the job will complete with an error and the
backing file will be left in place. For mirroring, the sector will be
marked again as dirty and re-examined later.
'stop': The job will be paused and the job iostatus will be set to
failed or nospace, while the VM will keep running. This can only be
specified if the block device has rerror=stop and werror=stop or enospc.
'enospc': Behaves as 'stop' for ENOSPC errors, 'report' for others.
In all cases, even for 'report', the I/O error is reported as a QMP
event BLOCK_JOB_ERROR, with the same arguments as BLOCK_IO_ERROR.
It is possible that while stopping the VM a BLOCK_IO_ERROR event will be
reported and will clobber the event from BLOCK_JOB_ERROR, or vice versa.
This is not really avoidable since stopping the VM completes all pending
I/O requests. In fact, it is already possible now that a series of
BLOCK_IO_ERROR events are reported with rerror=stop, because vm_stop
calls bdrv_drain_all and this can generate further errors.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2012-09-28 17:22:58 +02:00
|
|
|
monitor_protocol_event(ev, data);
|
2012-01-25 19:59:43 +01:00
|
|
|
|
|
|
|
qobject_decref(data);
|
|
|
|
}
|
|
|
|
|
2012-02-14 16:41:13 +01:00
|
|
|
static void bdrv_emit_qmp_eject_event(BlockDriverState *bs, bool ejected)
|
|
|
|
{
|
|
|
|
QObject *data;
|
|
|
|
|
|
|
|
data = qobject_from_jsonf("{ 'device': %s, 'tray-open': %i }",
|
|
|
|
bdrv_get_device_name(bs), ejected);
|
|
|
|
monitor_protocol_event(QEVENT_DEVICE_TRAY_MOVED, data);
|
|
|
|
|
|
|
|
qobject_decref(data);
|
|
|
|
}
|
|
|
|
|
2011-09-06 18:58:59 +02:00
|
|
|
static void bdrv_dev_change_media_cb(BlockDriverState *bs, bool load)
|
2011-08-03 15:07:41 +02:00
|
|
|
{
|
2011-08-03 15:07:42 +02:00
|
|
|
if (bs->dev_ops && bs->dev_ops->change_media_cb) {
|
2012-02-14 16:41:13 +01:00
|
|
|
bool tray_was_closed = !bdrv_dev_is_tray_open(bs);
|
2011-09-06 18:58:59 +02:00
|
|
|
bs->dev_ops->change_media_cb(bs->dev_opaque, load);
|
2012-02-14 16:41:13 +01:00
|
|
|
if (tray_was_closed) {
|
|
|
|
/* tray open */
|
|
|
|
bdrv_emit_qmp_eject_event(bs, true);
|
|
|
|
}
|
|
|
|
if (load) {
|
|
|
|
/* tray close */
|
|
|
|
bdrv_emit_qmp_eject_event(bs, false);
|
|
|
|
}
|
2011-08-03 15:07:42 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
block: Clean up remaining users of "removable"
BlockDriverState member removable is a confused mess. It is true when
an ide-cd, scsi-cd or floppy qdev is attached, or when the
BlockDriverState was created with -drive if={floppy,sd} or -drive
if={ide,scsi,xen,none},media=cdrom ("created removable"), except when
an ide-hd, scsi-hd, scsi-generic or virtio-blk qdev is attached.
Three users remain:
1. eject_device(), via bdrv_is_removable() uses it to determine
whether a block device can eject media.
2. bdrv_info() is monitor command "info block". QMP documentation
says "true if the device is removable, false otherwise". From the
monitor user's point of view, the only sensible interpretation of
"is removable" is "can eject media with monitor commands eject and
change".
A block device can eject media unless a device is attached that
doesn't support it. Switch the two users over to new
bdrv_dev_has_removable_media() that returns exactly that.
3. bdrv_getlength() uses to suppress its length cache when media can
change (see commit 46a4e4e6). Media change is either monitor
command change (updates the length cache), monitor command eject
(doesn't update the length cache, easily fixable), or physical
media change (invalidates length cache, not so easily fixable).
I'm refraining from improving anything here, because this series is
long enough already. Instead, I simply switch it over to
bdrv_dev_has_removable_media() as well.
This changes the behavior of the length cache and of monitor commands
eject and change in two cases:
a. drive not created removable, no device attached
The commit makes the drive removable, and defeats the length cache.
Example: -drive if=none
b. drive created removable, but the attached drive is non-removable,
and doesn't call bdrv_set_removable(..., 0) (most devices don't)
The commit makes the drive non-removable, and enables the length
cache.
Example: -drive if=xen,media=cdrom -M xenpv
The other non-removable devices that don't call
bdrv_set_removable() can't currently use a drive created removable,
either because they aren't qdevified, or because they lack a drive
property. Won't stay that way.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2011-09-06 18:58:51 +02:00
|
|
|
bool bdrv_dev_has_removable_media(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return !bs->dev || (bs->dev_ops && bs->dev_ops->change_media_cb);
|
|
|
|
}
|
|
|
|
|
2011-11-07 17:50:13 +01:00
|
|
|
void bdrv_dev_eject_request(BlockDriverState *bs, bool force)
|
|
|
|
{
|
|
|
|
if (bs->dev_ops && bs->dev_ops->eject_request_cb) {
|
|
|
|
bs->dev_ops->eject_request_cb(bs->dev_opaque, force);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-09-06 18:58:53 +02:00
|
|
|
bool bdrv_dev_is_tray_open(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->dev_ops && bs->dev_ops->is_tray_open) {
|
|
|
|
return bs->dev_ops->is_tray_open(bs->dev_opaque);
|
|
|
|
}
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2011-08-03 15:07:42 +02:00
|
|
|
static void bdrv_dev_resize_cb(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->dev_ops && bs->dev_ops->resize_cb) {
|
|
|
|
bs->dev_ops->resize_cb(bs->dev_opaque);
|
2011-08-03 15:07:41 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-09-06 18:58:46 +02:00
|
|
|
bool bdrv_dev_is_medium_locked(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->dev_ops && bs->dev_ops->is_medium_locked) {
|
|
|
|
return bs->dev_ops->is_medium_locked(bs->dev_opaque);
|
|
|
|
}
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
2009-04-22 01:11:50 +02:00
|
|
|
/*
|
|
|
|
* Run consistency checks on an image
|
|
|
|
*
|
2010-06-29 11:43:13 +02:00
|
|
|
* Returns 0 if the check could be completed (it doesn't mean that the image is
|
2011-04-28 17:20:38 +02:00
|
|
|
* free of errors) or -errno when an internal error occurred. The results of the
|
2010-06-29 11:43:13 +02:00
|
|
|
* check are stored in res.
|
2009-04-22 01:11:50 +02:00
|
|
|
*/
|
2012-05-11 16:07:02 +02:00
|
|
|
int bdrv_check(BlockDriverState *bs, BdrvCheckResult *res, BdrvCheckMode fix)
|
2009-04-22 01:11:50 +02:00
|
|
|
{
|
|
|
|
if (bs->drv->bdrv_check == NULL) {
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
2010-06-29 11:43:13 +02:00
|
|
|
memset(res, 0, sizeof(*res));
|
2012-05-11 16:07:02 +02:00
|
|
|
return bs->drv->bdrv_check(bs, res, fix);
|
2009-04-22 01:11:50 +02:00
|
|
|
}
|
|
|
|
|
2010-07-16 17:17:01 +02:00
|
|
|
#define COMMIT_BUF_SECTORS 2048
|
|
|
|
|
2003-07-06 19:15:21 +02:00
|
|
|
/* commit COW file into the raw image */
|
|
|
|
int bdrv_commit(BlockDriverState *bs)
|
|
|
|
{
|
2006-08-19 13:45:59 +02:00
|
|
|
BlockDriver *drv = bs->drv;
|
2010-07-16 17:17:01 +02:00
|
|
|
int64_t sector, total_sectors;
|
|
|
|
int n, ro, open_flags;
|
2012-09-20 21:13:34 +02:00
|
|
|
int ret = 0;
|
2010-07-16 17:17:01 +02:00
|
|
|
uint8_t *buf;
|
2012-10-04 13:09:46 +02:00
|
|
|
char filename[PATH_MAX];
|
2003-07-06 19:15:21 +02:00
|
|
|
|
2006-08-19 13:45:59 +02:00
|
|
|
if (!drv)
|
|
|
|
return -ENOMEDIUM;
|
2010-02-14 12:39:18 +01:00
|
|
|
|
|
|
|
if (!bs->backing_hd) {
|
|
|
|
return -ENOTSUP;
|
2003-07-06 19:15:21 +02:00
|
|
|
}
|
|
|
|
|
2012-01-18 15:40:41 +01:00
|
|
|
if (bdrv_in_use(bs) || bdrv_in_use(bs->backing_hd)) {
|
|
|
|
return -EBUSY;
|
|
|
|
}
|
|
|
|
|
2010-02-14 12:39:18 +01:00
|
|
|
ro = bs->backing_hd->read_only;
|
2012-10-04 13:09:46 +02:00
|
|
|
/* Use pstrcpy (not strncpy): filename must be NUL-terminated. */
|
|
|
|
pstrcpy(filename, sizeof(filename), bs->backing_hd->filename);
|
2010-02-14 12:39:18 +01:00
|
|
|
open_flags = bs->backing_hd->open_flags;
|
|
|
|
|
|
|
|
if (ro) {
|
2012-09-20 21:13:34 +02:00
|
|
|
if (bdrv_reopen(bs->backing_hd, open_flags | BDRV_O_RDWR, NULL)) {
|
|
|
|
return -EACCES;
|
2010-02-14 12:39:18 +01:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2003-07-06 19:15:21 +02:00
|
|
|
|
2009-11-30 18:21:19 +01:00
|
|
|
total_sectors = bdrv_getlength(bs) >> BDRV_SECTOR_BITS;
|
2011-08-21 05:09:37 +02:00
|
|
|
buf = g_malloc(COMMIT_BUF_SECTORS * BDRV_SECTOR_SIZE);
|
2010-07-16 17:17:01 +02:00
|
|
|
|
|
|
|
for (sector = 0; sector < total_sectors; sector += n) {
|
2011-11-14 13:44:18 +01:00
|
|
|
if (bdrv_is_allocated(bs, sector, COMMIT_BUF_SECTORS, &n)) {
|
2010-07-16 17:17:01 +02:00
|
|
|
|
|
|
|
if (bdrv_read(bs, sector, buf, n) != 0) {
|
|
|
|
ret = -EIO;
|
|
|
|
goto ro_cleanup;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bdrv_write(bs->backing_hd, sector, buf, n) != 0) {
|
|
|
|
ret = -EIO;
|
|
|
|
goto ro_cleanup;
|
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2003-07-06 19:15:21 +02:00
|
|
|
}
|
2005-12-18 19:28:15 +01:00
|
|
|
|
2010-01-17 12:32:30 +01:00
|
|
|
if (drv->bdrv_make_empty) {
|
|
|
|
ret = drv->bdrv_make_empty(bs);
|
|
|
|
bdrv_flush(bs);
|
|
|
|
}
|
2005-12-18 19:28:15 +01:00
|
|
|
|
2010-01-12 13:49:23 +01:00
|
|
|
/*
|
|
|
|
* Make sure all data we wrote to the backing device is actually
|
|
|
|
* stable on disk.
|
|
|
|
*/
|
|
|
|
if (bs->backing_hd)
|
|
|
|
bdrv_flush(bs->backing_hd);
|
2010-02-14 12:39:18 +01:00
|
|
|
|
|
|
|
ro_cleanup:
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(buf);
|
2010-02-14 12:39:18 +01:00
|
|
|
|
|
|
|
if (ro) {
|
2012-09-20 21:13:34 +02:00
|
|
|
/* ignoring error return here */
|
|
|
|
bdrv_reopen(bs->backing_hd, open_flags & ~BDRV_O_RDWR, NULL);
|
2010-02-14 12:39:18 +01:00
|
|
|
}
|
|
|
|
|
2010-01-17 12:32:30 +01:00
|
|
|
return ret;
|
2003-07-06 19:15:21 +02:00
|
|
|
}
|
|
|
|
|
2012-03-05 19:10:11 +01:00
|
|
|
int bdrv_commit_all(void)
|
2010-06-02 18:55:18 +02:00
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
2012-03-05 19:10:11 +01:00
|
|
|
int ret = bdrv_commit(bs);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
2010-06-02 18:55:18 +02:00
|
|
|
}
|
2012-03-05 19:10:11 +01:00
|
|
|
return 0;
|
2010-06-02 18:55:18 +02:00
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
struct BdrvTrackedRequest {
|
|
|
|
BlockDriverState *bs;
|
|
|
|
int64_t sector_num;
|
|
|
|
int nb_sectors;
|
|
|
|
bool is_write;
|
|
|
|
QLIST_ENTRY(BdrvTrackedRequest) list;
|
2011-11-30 13:23:42 +01:00
|
|
|
Coroutine *co; /* owner, used for deadlock detection */
|
2011-11-17 14:40:29 +01:00
|
|
|
CoQueue wait_queue; /* coroutines blocked on this request */
|
2011-11-17 14:40:27 +01:00
|
|
|
};
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Remove an active request from the tracked requests list
|
|
|
|
*
|
|
|
|
* This function should be called when a tracked request is completing.
|
|
|
|
*/
|
|
|
|
static void tracked_request_end(BdrvTrackedRequest *req)
|
|
|
|
{
|
|
|
|
QLIST_REMOVE(req, list);
|
2011-11-17 14:40:29 +01:00
|
|
|
qemu_co_queue_restart_all(&req->wait_queue);
|
2011-11-17 14:40:27 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Add an active request to the tracked requests list
|
|
|
|
*/
|
|
|
|
static void tracked_request_begin(BdrvTrackedRequest *req,
|
|
|
|
BlockDriverState *bs,
|
|
|
|
int64_t sector_num,
|
|
|
|
int nb_sectors, bool is_write)
|
|
|
|
{
|
|
|
|
*req = (BdrvTrackedRequest){
|
|
|
|
.bs = bs,
|
|
|
|
.sector_num = sector_num,
|
|
|
|
.nb_sectors = nb_sectors,
|
|
|
|
.is_write = is_write,
|
2011-11-30 13:23:42 +01:00
|
|
|
.co = qemu_coroutine_self(),
|
2011-11-17 14:40:27 +01:00
|
|
|
};
|
|
|
|
|
2011-11-17 14:40:29 +01:00
|
|
|
qemu_co_queue_init(&req->wait_queue);
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
QLIST_INSERT_HEAD(&bs->tracked_requests, req, list);
|
|
|
|
}
|
|
|
|
|
2011-11-23 12:47:56 +01:00
|
|
|
/**
|
|
|
|
* Round a region to cluster boundaries
|
|
|
|
*/
|
|
|
|
static void round_to_clusters(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
int64_t *cluster_sector_num,
|
|
|
|
int *cluster_nb_sectors)
|
|
|
|
{
|
|
|
|
BlockDriverInfo bdi;
|
|
|
|
|
|
|
|
if (bdrv_get_info(bs, &bdi) < 0 || bdi.cluster_size == 0) {
|
|
|
|
*cluster_sector_num = sector_num;
|
|
|
|
*cluster_nb_sectors = nb_sectors;
|
|
|
|
} else {
|
|
|
|
int64_t c = bdi.cluster_size / BDRV_SECTOR_SIZE;
|
|
|
|
*cluster_sector_num = QEMU_ALIGN_DOWN(sector_num, c);
|
|
|
|
*cluster_nb_sectors = QEMU_ALIGN_UP(sector_num - *cluster_sector_num +
|
|
|
|
nb_sectors, c);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:29 +01:00
|
|
|
static bool tracked_request_overlaps(BdrvTrackedRequest *req,
|
|
|
|
int64_t sector_num, int nb_sectors) {
|
2011-11-23 12:47:56 +01:00
|
|
|
/* aaaa bbbb */
|
|
|
|
if (sector_num >= req->sector_num + req->nb_sectors) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
/* bbbb aaaa */
|
|
|
|
if (req->sector_num >= sector_num + nb_sectors) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
return true;
|
2011-11-17 14:40:29 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
static void coroutine_fn wait_for_overlapping_requests(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors)
|
|
|
|
{
|
|
|
|
BdrvTrackedRequest *req;
|
2011-11-23 12:47:56 +01:00
|
|
|
int64_t cluster_sector_num;
|
|
|
|
int cluster_nb_sectors;
|
2011-11-17 14:40:29 +01:00
|
|
|
bool retry;
|
|
|
|
|
2011-11-23 12:47:56 +01:00
|
|
|
/* If we touch the same cluster it counts as an overlap. This guarantees
|
|
|
|
* that allocating writes will be serialized and not race with each other
|
|
|
|
* for the same cluster. For example, in copy-on-read it ensures that the
|
|
|
|
* CoR read and write operations are atomic and guest writes cannot
|
|
|
|
* interleave between them.
|
|
|
|
*/
|
|
|
|
round_to_clusters(bs, sector_num, nb_sectors,
|
|
|
|
&cluster_sector_num, &cluster_nb_sectors);
|
|
|
|
|
2011-11-17 14:40:29 +01:00
|
|
|
do {
|
|
|
|
retry = false;
|
|
|
|
QLIST_FOREACH(req, &bs->tracked_requests, list) {
|
2011-11-23 12:47:56 +01:00
|
|
|
if (tracked_request_overlaps(req, cluster_sector_num,
|
|
|
|
cluster_nb_sectors)) {
|
2011-11-30 13:23:42 +01:00
|
|
|
/* Hitting this means there was a reentrant request, for
|
|
|
|
* example, a block driver issuing nested requests. This must
|
|
|
|
* never happen since it means deadlock.
|
|
|
|
*/
|
|
|
|
assert(qemu_coroutine_self() != req->co);
|
|
|
|
|
2011-11-17 14:40:29 +01:00
|
|
|
qemu_co_queue_wait(&req->wait_queue);
|
|
|
|
retry = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
} while (retry);
|
|
|
|
}
|
|
|
|
|
2010-01-12 12:55:17 +01:00
|
|
|
/*
|
|
|
|
* Return values:
|
|
|
|
* 0 - success
|
|
|
|
* -EINVAL - backing format specified, but no file
|
|
|
|
* -ENOSPC - can't update the backing file because no space is left in the
|
|
|
|
* image file header
|
|
|
|
* -ENOTSUP - format driver doesn't support changing the backing file
|
|
|
|
*/
|
|
|
|
int bdrv_change_backing_file(BlockDriverState *bs,
|
|
|
|
const char *backing_file, const char *backing_fmt)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2012-04-12 14:01:02 +02:00
|
|
|
int ret;
|
2010-01-12 12:55:17 +01:00
|
|
|
|
2012-04-12 14:01:01 +02:00
|
|
|
/* Backing file format doesn't make sense without a backing file */
|
|
|
|
if (backing_fmt && !backing_file) {
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
2010-01-12 12:55:17 +01:00
|
|
|
if (drv->bdrv_change_backing_file != NULL) {
|
2012-04-12 14:01:02 +02:00
|
|
|
ret = drv->bdrv_change_backing_file(bs, backing_file, backing_fmt);
|
2010-01-12 12:55:17 +01:00
|
|
|
} else {
|
2012-04-12 14:01:02 +02:00
|
|
|
ret = -ENOTSUP;
|
2010-01-12 12:55:17 +01:00
|
|
|
}
|
2012-04-12 14:01:02 +02:00
|
|
|
|
|
|
|
if (ret == 0) {
|
|
|
|
pstrcpy(bs->backing_file, sizeof(bs->backing_file), backing_file ?: "");
|
|
|
|
pstrcpy(bs->backing_format, sizeof(bs->backing_format), backing_fmt ?: "");
|
|
|
|
}
|
|
|
|
return ret;
|
2010-01-12 12:55:17 +01:00
|
|
|
}
|
|
|
|
|
2012-09-27 19:29:12 +02:00
|
|
|
/*
|
|
|
|
* Finds the image layer in the chain that has 'bs' as its backing file.
|
|
|
|
*
|
|
|
|
* active is the current topmost image.
|
|
|
|
*
|
|
|
|
* Returns NULL if bs is not found in active's image chain,
|
|
|
|
* or if active == bs.
|
|
|
|
*/
|
|
|
|
BlockDriverState *bdrv_find_overlay(BlockDriverState *active,
|
|
|
|
BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriverState *overlay = NULL;
|
|
|
|
BlockDriverState *intermediate;
|
|
|
|
|
|
|
|
assert(active != NULL);
|
|
|
|
assert(bs != NULL);
|
|
|
|
|
|
|
|
/* if bs is the same as active, then by definition it has no overlay
|
|
|
|
*/
|
|
|
|
if (active == bs) {
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
intermediate = active;
|
|
|
|
while (intermediate->backing_hd) {
|
|
|
|
if (intermediate->backing_hd == bs) {
|
|
|
|
overlay = intermediate;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
intermediate = intermediate->backing_hd;
|
|
|
|
}
|
|
|
|
|
|
|
|
return overlay;
|
|
|
|
}
|
|
|
|
|
|
|
|
typedef struct BlkIntermediateStates {
|
|
|
|
BlockDriverState *bs;
|
|
|
|
QSIMPLEQ_ENTRY(BlkIntermediateStates) entry;
|
|
|
|
} BlkIntermediateStates;
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Drops images above 'base' up to and including 'top', and sets the image
|
|
|
|
* above 'top' to have base as its backing file.
|
|
|
|
*
|
|
|
|
* Requires that the overlay to 'top' is opened r/w, so that the backing file
|
|
|
|
* information in 'bs' can be properly updated.
|
|
|
|
*
|
|
|
|
* E.g., this will convert the following chain:
|
|
|
|
* bottom <- base <- intermediate <- top <- active
|
|
|
|
*
|
|
|
|
* to
|
|
|
|
*
|
|
|
|
* bottom <- base <- active
|
|
|
|
*
|
|
|
|
* It is allowed for bottom==base, in which case it converts:
|
|
|
|
*
|
|
|
|
* base <- intermediate <- top <- active
|
|
|
|
*
|
|
|
|
* to
|
|
|
|
*
|
|
|
|
* base <- active
|
|
|
|
*
|
|
|
|
* Error conditions:
|
|
|
|
* if active == top, that is considered an error
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int bdrv_drop_intermediate(BlockDriverState *active, BlockDriverState *top,
|
|
|
|
BlockDriverState *base)
|
|
|
|
{
|
|
|
|
BlockDriverState *intermediate;
|
|
|
|
BlockDriverState *base_bs = NULL;
|
|
|
|
BlockDriverState *new_top_bs = NULL;
|
|
|
|
BlkIntermediateStates *intermediate_state, *next;
|
|
|
|
int ret = -EIO;
|
|
|
|
|
|
|
|
QSIMPLEQ_HEAD(states_to_delete, BlkIntermediateStates) states_to_delete;
|
|
|
|
QSIMPLEQ_INIT(&states_to_delete);
|
|
|
|
|
|
|
|
if (!top->drv || !base->drv) {
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
|
|
|
new_top_bs = bdrv_find_overlay(active, top);
|
|
|
|
|
|
|
|
if (new_top_bs == NULL) {
|
|
|
|
/* we could not find the image above 'top', this is an error */
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* special case of new_top_bs->backing_hd already pointing to base - nothing
|
|
|
|
* to do, no intermediate images */
|
|
|
|
if (new_top_bs->backing_hd == base) {
|
|
|
|
ret = 0;
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
|
|
|
intermediate = top;
|
|
|
|
|
|
|
|
/* now we will go down through the list, and add each BDS we find
|
|
|
|
* into our deletion queue, until we hit the 'base'
|
|
|
|
*/
|
|
|
|
while (intermediate) {
|
|
|
|
intermediate_state = g_malloc0(sizeof(BlkIntermediateStates));
|
|
|
|
intermediate_state->bs = intermediate;
|
|
|
|
QSIMPLEQ_INSERT_TAIL(&states_to_delete, intermediate_state, entry);
|
|
|
|
|
|
|
|
if (intermediate->backing_hd == base) {
|
|
|
|
base_bs = intermediate->backing_hd;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
intermediate = intermediate->backing_hd;
|
|
|
|
}
|
|
|
|
if (base_bs == NULL) {
|
|
|
|
/* something went wrong, we did not end at the base. safely
|
|
|
|
* unravel everything, and exit with error */
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* success - we can delete the intermediate states, and link top->base */
|
|
|
|
ret = bdrv_change_backing_file(new_top_bs, base_bs->filename,
|
|
|
|
base_bs->drv ? base_bs->drv->format_name : "");
|
|
|
|
if (ret) {
|
|
|
|
goto exit;
|
|
|
|
}
|
|
|
|
new_top_bs->backing_hd = base_bs;
|
|
|
|
|
|
|
|
|
|
|
|
QSIMPLEQ_FOREACH_SAFE(intermediate_state, &states_to_delete, entry, next) {
|
|
|
|
/* so that bdrv_close() does not recursively close the chain */
|
|
|
|
intermediate_state->bs->backing_hd = NULL;
|
|
|
|
bdrv_delete(intermediate_state->bs);
|
|
|
|
}
|
|
|
|
ret = 0;
|
|
|
|
|
|
|
|
exit:
|
|
|
|
QSIMPLEQ_FOREACH_SAFE(intermediate_state, &states_to_delete, entry, next) {
|
|
|
|
g_free(intermediate_state);
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2009-03-03 18:37:16 +01:00
|
|
|
static int bdrv_check_byte_request(BlockDriverState *bs, int64_t offset,
|
|
|
|
size_t size)
|
|
|
|
{
|
|
|
|
int64_t len;
|
|
|
|
|
|
|
|
if (!bdrv_is_inserted(bs))
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
|
|
|
|
if (bs->growable)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
len = bdrv_getlength(bs);
|
|
|
|
|
2009-05-08 14:47:24 +02:00
|
|
|
if (offset < 0)
|
|
|
|
return -EIO;
|
|
|
|
|
|
|
|
if ((offset > len) || (len - offset < size))
|
2009-03-03 18:37:16 +01:00
|
|
|
return -EIO;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int bdrv_check_request(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
int nb_sectors)
|
|
|
|
{
|
2010-05-27 16:20:31 +02:00
|
|
|
return bdrv_check_byte_request(bs, sector_num * BDRV_SECTOR_SIZE,
|
|
|
|
nb_sectors * BDRV_SECTOR_SIZE);
|
2009-03-03 18:37:16 +01:00
|
|
|
}
|
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
typedef struct RwCo {
|
|
|
|
BlockDriverState *bs;
|
|
|
|
int64_t sector_num;
|
|
|
|
int nb_sectors;
|
|
|
|
QEMUIOVector *qiov;
|
|
|
|
bool is_write;
|
|
|
|
int ret;
|
|
|
|
} RwCo;
|
|
|
|
|
|
|
|
static void coroutine_fn bdrv_rw_co_entry(void *opaque)
|
2003-06-30 12:03:06 +02:00
|
|
|
{
|
2011-10-13 14:08:22 +02:00
|
|
|
RwCo *rwco = opaque;
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
if (!rwco->is_write) {
|
|
|
|
rwco->ret = bdrv_co_do_readv(rwco->bs, rwco->sector_num,
|
2012-01-18 15:40:42 +01:00
|
|
|
rwco->nb_sectors, rwco->qiov, 0);
|
2011-10-13 14:08:22 +02:00
|
|
|
} else {
|
|
|
|
rwco->ret = bdrv_co_do_writev(rwco->bs, rwco->sector_num,
|
2012-02-07 14:27:25 +01:00
|
|
|
rwco->nb_sectors, rwco->qiov, 0);
|
2011-10-13 14:08:22 +02:00
|
|
|
}
|
|
|
|
}
|
2011-07-15 16:05:00 +02:00
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
/*
|
|
|
|
* Process a synchronous request using coroutines
|
|
|
|
*/
|
|
|
|
static int bdrv_rw_co(BlockDriverState *bs, int64_t sector_num, uint8_t *buf,
|
|
|
|
int nb_sectors, bool is_write)
|
|
|
|
{
|
|
|
|
QEMUIOVector qiov;
|
|
|
|
struct iovec iov = {
|
|
|
|
.iov_base = (void *)buf,
|
|
|
|
.iov_len = nb_sectors * BDRV_SECTOR_SIZE,
|
|
|
|
};
|
|
|
|
Coroutine *co;
|
|
|
|
RwCo rwco = {
|
|
|
|
.bs = bs,
|
|
|
|
.sector_num = sector_num,
|
|
|
|
.nb_sectors = nb_sectors,
|
|
|
|
.qiov = &qiov,
|
|
|
|
.is_write = is_write,
|
|
|
|
.ret = NOT_DONE,
|
|
|
|
};
|
2011-07-15 16:05:00 +02:00
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
qemu_iovec_init_external(&qiov, &iov, 1);
|
2011-07-15 16:05:00 +02:00
|
|
|
|
2012-04-02 12:59:34 +02:00
|
|
|
/**
|
|
|
|
* In sync call context, when the vcpu is blocked, this throttling timer
|
|
|
|
* will not fire; so the I/O throttling function has to be disabled here
|
|
|
|
* if it has been enabled.
|
|
|
|
*/
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
fprintf(stderr, "Disabling I/O throttling on '%s' due "
|
|
|
|
"to synchronous I/O.\n", bdrv_get_device_name(bs));
|
|
|
|
bdrv_io_limits_disable(bs);
|
|
|
|
}
|
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
if (qemu_in_coroutine()) {
|
|
|
|
/* Fast-path if already in coroutine context */
|
|
|
|
bdrv_rw_co_entry(&rwco);
|
|
|
|
} else {
|
|
|
|
co = qemu_coroutine_create(bdrv_rw_co_entry);
|
|
|
|
qemu_coroutine_enter(co, &rwco);
|
|
|
|
while (rwco.ret == NOT_DONE) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return rwco.ret;
|
|
|
|
}
|
2004-03-14 22:38:54 +01:00
|
|
|
|
2011-10-13 14:08:22 +02:00
|
|
|
/* return < 0 if error. See bdrv_write() for the return codes */
|
|
|
|
int bdrv_read(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
uint8_t *buf, int nb_sectors)
|
|
|
|
{
|
|
|
|
return bdrv_rw_co(bs, sector_num, buf, nb_sectors, false);
|
2003-06-30 12:03:06 +02:00
|
|
|
}
|
|
|
|
|
2012-06-29 17:34:29 +02:00
|
|
|
/* Just like bdrv_read(), but with I/O throttling temporarily disabled */
|
|
|
|
int bdrv_read_unthrottled(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
uint8_t *buf, int nb_sectors)
|
|
|
|
{
|
|
|
|
bool enabled;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
enabled = bs->io_limits_enabled;
|
|
|
|
bs->io_limits_enabled = false;
|
|
|
|
ret = bdrv_read(bs, 0, buf, 1);
|
|
|
|
bs->io_limits_enabled = enabled;
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-04-12 14:01:04 +02:00
|
|
|
#define BITS_PER_LONG (sizeof(unsigned long) * 8)
|
|
|
|
|
2009-11-02 14:40:41 +01:00
|
|
|
static void set_dirty_bitmap(BlockDriverState *bs, int64_t sector_num,
|
2009-11-30 18:21:19 +01:00
|
|
|
int nb_sectors, int dirty)
|
2009-11-02 14:40:41 +01:00
|
|
|
{
|
|
|
|
int64_t start, end;
|
2009-11-30 18:21:20 +01:00
|
|
|
unsigned long val, idx, bit;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2009-11-30 18:21:19 +01:00
|
|
|
start = sector_num / BDRV_SECTORS_PER_DIRTY_CHUNK;
|
2009-11-30 18:21:20 +01:00
|
|
|
end = (sector_num + nb_sectors - 1) / BDRV_SECTORS_PER_DIRTY_CHUNK;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
|
|
|
for (; start <= end; start++) {
|
2012-04-12 14:01:04 +02:00
|
|
|
idx = start / BITS_PER_LONG;
|
|
|
|
bit = start % BITS_PER_LONG;
|
2009-11-30 18:21:20 +01:00
|
|
|
val = bs->dirty_bitmap[idx];
|
|
|
|
if (dirty) {
|
2010-11-08 20:02:54 +01:00
|
|
|
if (!(val & (1UL << bit))) {
|
2010-01-26 09:31:48 +01:00
|
|
|
bs->dirty_count++;
|
2010-11-08 20:02:54 +01:00
|
|
|
val |= 1UL << bit;
|
2010-01-26 09:31:48 +01:00
|
|
|
}
|
2009-11-30 18:21:20 +01:00
|
|
|
} else {
|
2010-11-08 20:02:54 +01:00
|
|
|
if (val & (1UL << bit)) {
|
2010-01-26 09:31:48 +01:00
|
|
|
bs->dirty_count--;
|
2010-11-08 20:02:54 +01:00
|
|
|
val &= ~(1UL << bit);
|
2010-01-26 09:31:48 +01:00
|
|
|
}
|
2009-11-30 18:21:20 +01:00
|
|
|
}
|
|
|
|
bs->dirty_bitmap[idx] = val;
|
2009-11-02 14:40:41 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
/* Return < 0 if error. Important errors are:
|
2006-08-19 13:45:59 +02:00
|
|
|
-EIO generic I/O error (may happen for all errors)
|
|
|
|
-ENOMEDIUM No media inserted.
|
|
|
|
-EINVAL Invalid sector number or nb_sectors
|
|
|
|
-EACCES Trying to write a read-only device
|
|
|
|
*/
|
2007-09-16 23:08:06 +02:00
|
|
|
int bdrv_write(BlockDriverState *bs, int64_t sector_num,
|
2003-06-30 12:03:06 +02:00
|
|
|
const uint8_t *buf, int nb_sectors)
|
|
|
|
{
|
2011-10-13 14:08:22 +02:00
|
|
|
return bdrv_rw_co(bs, sector_num, (uint8_t *)buf, nb_sectors, true);
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2009-03-12 20:57:16 +01:00
|
|
|
int bdrv_pread(BlockDriverState *bs, int64_t offset,
|
|
|
|
void *buf, int count1)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
2009-11-30 18:21:19 +01:00
|
|
|
uint8_t tmp_buf[BDRV_SECTOR_SIZE];
|
2006-08-01 18:21:11 +02:00
|
|
|
int len, nb_sectors, count;
|
|
|
|
int64_t sector_num;
|
2010-01-20 15:03:02 +01:00
|
|
|
int ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
|
|
|
|
count = count1;
|
|
|
|
/* first read to align to sector start */
|
2009-11-30 18:21:19 +01:00
|
|
|
len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
|
2006-08-01 18:21:11 +02:00
|
|
|
if (len > count)
|
|
|
|
len = count;
|
2009-11-30 18:21:19 +01:00
|
|
|
sector_num = offset >> BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (len > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2009-11-30 18:21:19 +01:00
|
|
|
memcpy(buf, tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), len);
|
2006-08-01 18:21:11 +02:00
|
|
|
count -= len;
|
|
|
|
if (count == 0)
|
|
|
|
return count1;
|
|
|
|
sector_num++;
|
|
|
|
buf += len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* read the sectors "in place" */
|
2009-11-30 18:21:19 +01:00
|
|
|
nb_sectors = count >> BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (nb_sectors > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_read(bs, sector_num, buf, nb_sectors)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
sector_num += nb_sectors;
|
2009-11-30 18:21:19 +01:00
|
|
|
len = nb_sectors << BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
buf += len;
|
|
|
|
count -= len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* add data from the last sector */
|
|
|
|
if (count > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
memcpy(buf, tmp_buf, count);
|
|
|
|
}
|
|
|
|
return count1;
|
|
|
|
}
|
|
|
|
|
2009-03-12 20:57:16 +01:00
|
|
|
int bdrv_pwrite(BlockDriverState *bs, int64_t offset,
|
|
|
|
const void *buf, int count1)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
2009-11-30 18:21:19 +01:00
|
|
|
uint8_t tmp_buf[BDRV_SECTOR_SIZE];
|
2006-08-01 18:21:11 +02:00
|
|
|
int len, nb_sectors, count;
|
|
|
|
int64_t sector_num;
|
2010-01-20 15:03:02 +01:00
|
|
|
int ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
|
|
|
|
count = count1;
|
|
|
|
/* first write to align to sector start */
|
2009-11-30 18:21:19 +01:00
|
|
|
len = (BDRV_SECTOR_SIZE - offset) & (BDRV_SECTOR_SIZE - 1);
|
2006-08-01 18:21:11 +02:00
|
|
|
if (len > count)
|
|
|
|
len = count;
|
2009-11-30 18:21:19 +01:00
|
|
|
sector_num = offset >> BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (len > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2009-11-30 18:21:19 +01:00
|
|
|
memcpy(tmp_buf + (offset & (BDRV_SECTOR_SIZE - 1)), buf, len);
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
count -= len;
|
|
|
|
if (count == 0)
|
|
|
|
return count1;
|
|
|
|
sector_num++;
|
|
|
|
buf += len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* write the sectors "in place" */
|
2009-11-30 18:21:19 +01:00
|
|
|
nb_sectors = count >> BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (nb_sectors > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_write(bs, sector_num, buf, nb_sectors)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
sector_num += nb_sectors;
|
2009-11-30 18:21:19 +01:00
|
|
|
len = nb_sectors << BDRV_SECTOR_BITS;
|
2006-08-01 18:21:11 +02:00
|
|
|
buf += len;
|
|
|
|
count -= len;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* add data from the last sector */
|
|
|
|
if (count > 0) {
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_read(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
memcpy(tmp_buf, buf, count);
|
2010-01-20 15:03:02 +01:00
|
|
|
if ((ret = bdrv_write(bs, sector_num, tmp_buf, 1)) < 0)
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
return count1;
|
|
|
|
}
|
|
|
|
|
2010-06-16 16:38:15 +02:00
|
|
|
/*
|
|
|
|
* Writes to the file and ensures that no writes are reordered across this
|
|
|
|
* request (acts as a barrier)
|
|
|
|
*
|
|
|
|
* Returns 0 on success, -errno in error cases.
|
|
|
|
*/
|
|
|
|
int bdrv_pwrite_sync(BlockDriverState *bs, int64_t offset,
|
|
|
|
const void *buf, int count)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
ret = bdrv_pwrite(bs, offset, buf, count);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2012-06-06 00:04:49 +02:00
|
|
|
/* No flush needed for cache modes that already do it */
|
|
|
|
if (bs->enable_write_cache) {
|
2010-06-16 16:38:15 +02:00
|
|
|
bdrv_flush(bs);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
static int coroutine_fn bdrv_co_do_copy_on_readv(BlockDriverState *bs,
|
2011-11-17 14:40:31 +01:00
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov)
|
|
|
|
{
|
|
|
|
/* Perform I/O through a temporary buffer so that users who scribble over
|
|
|
|
* their read buffer while the operation is in progress do not end up
|
|
|
|
* modifying the image file. This is critical for zero-copy guest I/O
|
|
|
|
* where anything might happen inside guest memory.
|
|
|
|
*/
|
|
|
|
void *bounce_buffer;
|
|
|
|
|
2012-02-07 14:27:26 +01:00
|
|
|
BlockDriver *drv = bs->drv;
|
2011-11-17 14:40:31 +01:00
|
|
|
struct iovec iov;
|
|
|
|
QEMUIOVector bounce_qiov;
|
|
|
|
int64_t cluster_sector_num;
|
|
|
|
int cluster_nb_sectors;
|
|
|
|
size_t skip_bytes;
|
|
|
|
int ret;
|
|
|
|
|
|
|
|
/* Cover entire cluster so no additional backing file I/O is required when
|
|
|
|
* allocating cluster in the image file.
|
|
|
|
*/
|
|
|
|
round_to_clusters(bs, sector_num, nb_sectors,
|
|
|
|
&cluster_sector_num, &cluster_nb_sectors);
|
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
trace_bdrv_co_do_copy_on_readv(bs, sector_num, nb_sectors,
|
|
|
|
cluster_sector_num, cluster_nb_sectors);
|
2011-11-17 14:40:31 +01:00
|
|
|
|
|
|
|
iov.iov_len = cluster_nb_sectors * BDRV_SECTOR_SIZE;
|
|
|
|
iov.iov_base = bounce_buffer = qemu_blockalign(bs, iov.iov_len);
|
|
|
|
qemu_iovec_init_external(&bounce_qiov, &iov, 1);
|
|
|
|
|
2012-02-07 14:27:26 +01:00
|
|
|
ret = drv->bdrv_co_readv(bs, cluster_sector_num, cluster_nb_sectors,
|
|
|
|
&bounce_qiov);
|
2011-11-17 14:40:31 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
2012-02-07 14:27:26 +01:00
|
|
|
if (drv->bdrv_co_write_zeroes &&
|
|
|
|
buffer_is_zero(bounce_buffer, iov.iov_len)) {
|
2012-03-20 15:12:58 +01:00
|
|
|
ret = bdrv_co_do_write_zeroes(bs, cluster_sector_num,
|
|
|
|
cluster_nb_sectors);
|
2012-02-07 14:27:26 +01:00
|
|
|
} else {
|
2012-06-06 00:04:49 +02:00
|
|
|
/* This does not change the data on the disk, it is not necessary
|
|
|
|
* to flush even in cache=writethrough mode.
|
|
|
|
*/
|
2012-02-07 14:27:26 +01:00
|
|
|
ret = drv->bdrv_co_writev(bs, cluster_sector_num, cluster_nb_sectors,
|
2011-11-17 14:40:31 +01:00
|
|
|
&bounce_qiov);
|
2012-02-07 14:27:26 +01:00
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:31 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
/* It might be okay to ignore write errors for guest requests. If this
|
|
|
|
* is a deliberate copy-on-read then we don't want to ignore the error.
|
|
|
|
* Simply report it in all cases.
|
|
|
|
*/
|
|
|
|
goto err;
|
|
|
|
}
|
|
|
|
|
|
|
|
skip_bytes = (sector_num - cluster_sector_num) * BDRV_SECTOR_SIZE;
|
allow qemu_iovec_from_buffer() to specify offset from which to start copying
Similar to
qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
int c, size_t bytes);
the new prototype is:
qemu_iovec_from_buf(QEMUIOVector *qiov, size_t offset,
const void *buf, size_t bytes);
The processing starts at offset bytes within qiov.
This way, we may copy a bounce buffer directly to
a middle of qiov.
This is exactly the same function as iov_from_buf() from
iov.c, so use the existing implementation and rename it
to qemu_iovec_from_buf() to be shorter and to match the
utility function.
As with utility implementation, we now assert that the
offset is inside actual iovec. Nothing changed for
current callers, because `offset' parameter is new.
While at it, stop using "bounce-qiov" in block/qcow2.c
and copy decrypted data directly from cluster_data
instead of recreating a temp qiov for doing that.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-06-07 18:17:55 +02:00
|
|
|
qemu_iovec_from_buf(qiov, 0, bounce_buffer + skip_bytes,
|
|
|
|
nb_sectors * BDRV_SECTOR_SIZE);
|
2011-11-17 14:40:31 +01:00
|
|
|
|
|
|
|
err:
|
|
|
|
qemu_vfree(bounce_buffer);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2011-10-05 18:17:03 +02:00
|
|
|
/*
|
|
|
|
* Handle a read request in coroutine context
|
|
|
|
*/
|
|
|
|
static int coroutine_fn bdrv_co_do_readv(BlockDriverState *bs,
|
2012-01-18 15:40:42 +01:00
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov,
|
|
|
|
BdrvRequestFlags flags)
|
2011-07-14 17:27:13 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2011-11-17 14:40:27 +01:00
|
|
|
BdrvTrackedRequest req;
|
|
|
|
int ret;
|
2011-07-14 17:27:13 +02:00
|
|
|
|
|
|
|
if (!drv) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
}
|
|
|
|
if (bdrv_check_request(bs, sector_num, nb_sectors)) {
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
/* throttling disk read I/O */
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
bdrv_io_limits_intercept(bs, false, nb_sectors);
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:29 +01:00
|
|
|
if (bs->copy_on_read) {
|
2012-01-18 15:40:42 +01:00
|
|
|
flags |= BDRV_REQ_COPY_ON_READ;
|
|
|
|
}
|
|
|
|
if (flags & BDRV_REQ_COPY_ON_READ) {
|
|
|
|
bs->copy_on_read_in_flight++;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bs->copy_on_read_in_flight) {
|
2011-11-17 14:40:29 +01:00
|
|
|
wait_for_overlapping_requests(bs, sector_num, nb_sectors);
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
tracked_request_begin(&req, bs, sector_num, nb_sectors, false);
|
2011-11-17 14:40:31 +01:00
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
if (flags & BDRV_REQ_COPY_ON_READ) {
|
2011-11-17 14:40:31 +01:00
|
|
|
int pnum;
|
|
|
|
|
|
|
|
ret = bdrv_co_is_allocated(bs, sector_num, nb_sectors, &pnum);
|
|
|
|
if (ret < 0) {
|
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!ret || pnum != nb_sectors) {
|
2012-01-18 15:40:42 +01:00
|
|
|
ret = bdrv_co_do_copy_on_readv(bs, sector_num, nb_sectors, qiov);
|
2011-11-17 14:40:31 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
ret = drv->bdrv_co_readv(bs, sector_num, nb_sectors, qiov);
|
2011-11-17 14:40:31 +01:00
|
|
|
|
|
|
|
out:
|
2011-11-17 14:40:27 +01:00
|
|
|
tracked_request_end(&req);
|
2012-01-18 15:40:42 +01:00
|
|
|
|
|
|
|
if (flags & BDRV_REQ_COPY_ON_READ) {
|
|
|
|
bs->copy_on_read_in_flight--;
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
return ret;
|
2011-07-14 17:27:13 +02:00
|
|
|
}
|
|
|
|
|
2011-10-05 18:17:03 +02:00
|
|
|
int coroutine_fn bdrv_co_readv(BlockDriverState *bs, int64_t sector_num,
|
2011-07-14 17:27:13 +02:00
|
|
|
int nb_sectors, QEMUIOVector *qiov)
|
|
|
|
{
|
2011-10-05 18:17:03 +02:00
|
|
|
trace_bdrv_co_readv(bs, sector_num, nb_sectors);
|
2011-07-14 17:27:13 +02:00
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
return bdrv_co_do_readv(bs, sector_num, nb_sectors, qiov, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
int coroutine_fn bdrv_co_copy_on_readv(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov)
|
|
|
|
{
|
|
|
|
trace_bdrv_co_copy_on_readv(bs, sector_num, nb_sectors);
|
|
|
|
|
|
|
|
return bdrv_co_do_readv(bs, sector_num, nb_sectors, qiov,
|
|
|
|
BDRV_REQ_COPY_ON_READ);
|
2011-10-05 18:17:03 +02:00
|
|
|
}
|
|
|
|
|
2012-02-07 14:27:25 +01:00
|
|
|
static int coroutine_fn bdrv_co_do_write_zeroes(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
QEMUIOVector qiov;
|
|
|
|
struct iovec iov;
|
|
|
|
int ret;
|
|
|
|
|
2012-03-20 15:12:58 +01:00
|
|
|
/* TODO Emulate only part of misaligned requests instead of letting block
|
|
|
|
* drivers return -ENOTSUP and emulate everything */
|
|
|
|
|
2012-02-07 14:27:25 +01:00
|
|
|
/* First try the efficient write zeroes operation */
|
|
|
|
if (drv->bdrv_co_write_zeroes) {
|
2012-03-20 15:12:58 +01:00
|
|
|
ret = drv->bdrv_co_write_zeroes(bs, sector_num, nb_sectors);
|
|
|
|
if (ret != -ENOTSUP) {
|
|
|
|
return ret;
|
|
|
|
}
|
2012-02-07 14:27:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Fall back to bounce buffer if write zeroes is unsupported */
|
|
|
|
iov.iov_len = nb_sectors * BDRV_SECTOR_SIZE;
|
|
|
|
iov.iov_base = qemu_blockalign(bs, iov.iov_len);
|
|
|
|
memset(iov.iov_base, 0, iov.iov_len);
|
|
|
|
qemu_iovec_init_external(&qiov, &iov, 1);
|
|
|
|
|
|
|
|
ret = drv->bdrv_co_writev(bs, sector_num, nb_sectors, &qiov);
|
|
|
|
|
|
|
|
qemu_vfree(iov.iov_base);
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
2011-10-05 18:17:03 +02:00
|
|
|
/*
|
|
|
|
* Handle a write request in coroutine context
|
|
|
|
*/
|
|
|
|
static int coroutine_fn bdrv_co_do_writev(BlockDriverState *bs,
|
2012-02-07 14:27:25 +01:00
|
|
|
int64_t sector_num, int nb_sectors, QEMUIOVector *qiov,
|
|
|
|
BdrvRequestFlags flags)
|
2011-10-05 18:17:03 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2011-11-17 14:40:27 +01:00
|
|
|
BdrvTrackedRequest req;
|
2011-10-13 14:08:24 +02:00
|
|
|
int ret;
|
2011-07-14 17:27:13 +02:00
|
|
|
|
|
|
|
if (!bs->drv) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
}
|
|
|
|
if (bs->read_only) {
|
|
|
|
return -EACCES;
|
|
|
|
}
|
|
|
|
if (bdrv_check_request(bs, sector_num, nb_sectors)) {
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
/* throttling disk write I/O */
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
bdrv_io_limits_intercept(bs, true, nb_sectors);
|
|
|
|
}
|
|
|
|
|
2012-01-18 15:40:42 +01:00
|
|
|
if (bs->copy_on_read_in_flight) {
|
2011-11-17 14:40:29 +01:00
|
|
|
wait_for_overlapping_requests(bs, sector_num, nb_sectors);
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
tracked_request_begin(&req, bs, sector_num, nb_sectors, true);
|
|
|
|
|
2012-02-07 14:27:25 +01:00
|
|
|
if (flags & BDRV_REQ_ZERO_WRITE) {
|
|
|
|
ret = bdrv_co_do_write_zeroes(bs, sector_num, nb_sectors);
|
|
|
|
} else {
|
|
|
|
ret = drv->bdrv_co_writev(bs, sector_num, nb_sectors, qiov);
|
|
|
|
}
|
2011-10-13 14:08:24 +02:00
|
|
|
|
2012-06-06 00:04:49 +02:00
|
|
|
if (ret == 0 && !bs->enable_write_cache) {
|
|
|
|
ret = bdrv_co_flush(bs);
|
|
|
|
}
|
|
|
|
|
2011-07-14 17:27:13 +02:00
|
|
|
if (bs->dirty_bitmap) {
|
2012-10-18 16:49:18 +02:00
|
|
|
bdrv_set_dirty(bs, sector_num, nb_sectors);
|
2011-07-14 17:27:13 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if (bs->wr_highest_sector < sector_num + nb_sectors - 1) {
|
|
|
|
bs->wr_highest_sector = sector_num + nb_sectors - 1;
|
|
|
|
}
|
|
|
|
|
2011-11-17 14:40:27 +01:00
|
|
|
tracked_request_end(&req);
|
|
|
|
|
2011-10-13 14:08:24 +02:00
|
|
|
return ret;
|
2011-07-14 17:27:13 +02:00
|
|
|
}
|
|
|
|
|
2011-10-05 18:17:03 +02:00
|
|
|
int coroutine_fn bdrv_co_writev(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
int nb_sectors, QEMUIOVector *qiov)
|
|
|
|
{
|
|
|
|
trace_bdrv_co_writev(bs, sector_num, nb_sectors);
|
|
|
|
|
2012-02-07 14:27:25 +01:00
|
|
|
return bdrv_co_do_writev(bs, sector_num, nb_sectors, qiov, 0);
|
|
|
|
}
|
|
|
|
|
|
|
|
int coroutine_fn bdrv_co_write_zeroes(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors)
|
|
|
|
{
|
|
|
|
trace_bdrv_co_write_zeroes(bs, sector_num, nb_sectors);
|
|
|
|
|
|
|
|
return bdrv_co_do_writev(bs, sector_num, nb_sectors, NULL,
|
|
|
|
BDRV_REQ_ZERO_WRITE);
|
2011-10-05 18:17:03 +02:00
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
/**
|
|
|
|
* Truncate file to 'offset' bytes (needed only for file protocols)
|
|
|
|
*/
|
|
|
|
int bdrv_truncate(BlockDriverState *bs, int64_t offset)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2010-04-19 17:56:41 +02:00
|
|
|
int ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2006-08-01 18:21:11 +02:00
|
|
|
if (!drv->bdrv_truncate)
|
|
|
|
return -ENOTSUP;
|
2009-10-26 15:25:16 +01:00
|
|
|
if (bs->read_only)
|
|
|
|
return -EACCES;
|
2011-01-26 15:12:35 +01:00
|
|
|
if (bdrv_in_use(bs))
|
|
|
|
return -EBUSY;
|
2010-04-19 17:56:41 +02:00
|
|
|
ret = drv->bdrv_truncate(bs, offset);
|
|
|
|
if (ret == 0) {
|
|
|
|
ret = refresh_total_sectors(bs, offset >> BDRV_SECTOR_BITS);
|
2011-08-03 15:07:42 +02:00
|
|
|
bdrv_dev_resize_cb(bs);
|
2010-04-19 17:56:41 +02:00
|
|
|
}
|
|
|
|
return ret;
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2011-07-12 13:56:39 +02:00
|
|
|
/**
|
|
|
|
* Length of a allocated file in bytes. Sparse files are counted by actual
|
|
|
|
* allocated space. Return < 0 if error or unknown.
|
|
|
|
*/
|
|
|
|
int64_t bdrv_get_allocated_file_size(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
}
|
|
|
|
if (drv->bdrv_get_allocated_file_size) {
|
|
|
|
return drv->bdrv_get_allocated_file_size(bs);
|
|
|
|
}
|
|
|
|
if (bs->file) {
|
|
|
|
return bdrv_get_allocated_file_size(bs->file);
|
|
|
|
}
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
/**
|
|
|
|
* Length of a file in bytes. Return < 0 if error or unknown.
|
|
|
|
*/
|
|
|
|
int64_t bdrv_getlength(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2010-04-19 17:56:41 +02:00
|
|
|
|
block: Clean up remaining users of "removable"
BlockDriverState member removable is a confused mess. It is true when
an ide-cd, scsi-cd or floppy qdev is attached, or when the
BlockDriverState was created with -drive if={floppy,sd} or -drive
if={ide,scsi,xen,none},media=cdrom ("created removable"), except when
an ide-hd, scsi-hd, scsi-generic or virtio-blk qdev is attached.
Three users remain:
1. eject_device(), via bdrv_is_removable() uses it to determine
whether a block device can eject media.
2. bdrv_info() is monitor command "info block". QMP documentation
says "true if the device is removable, false otherwise". From the
monitor user's point of view, the only sensible interpretation of
"is removable" is "can eject media with monitor commands eject and
change".
A block device can eject media unless a device is attached that
doesn't support it. Switch the two users over to new
bdrv_dev_has_removable_media() that returns exactly that.
3. bdrv_getlength() uses to suppress its length cache when media can
change (see commit 46a4e4e6). Media change is either monitor
command change (updates the length cache), monitor command eject
(doesn't update the length cache, easily fixable), or physical
media change (invalidates length cache, not so easily fixable).
I'm refraining from improving anything here, because this series is
long enough already. Instead, I simply switch it over to
bdrv_dev_has_removable_media() as well.
This changes the behavior of the length cache and of monitor commands
eject and change in two cases:
a. drive not created removable, no device attached
The commit makes the drive removable, and defeats the length cache.
Example: -drive if=none
b. drive created removable, but the attached drive is non-removable,
and doesn't call bdrv_set_removable(..., 0) (most devices don't)
The commit makes the drive non-removable, and enables the length
cache.
Example: -drive if=xen,media=cdrom -M xenpv
The other non-removable devices that don't call
bdrv_set_removable() can't currently use a drive created removable,
either because they aren't qdevified, or because they lack a drive
property. Won't stay that way.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2011-09-06 18:58:51 +02:00
|
|
|
if (bs->growable || bdrv_dev_has_removable_media(bs)) {
|
2011-03-29 21:04:41 +02:00
|
|
|
if (drv->bdrv_getlength) {
|
|
|
|
return drv->bdrv_getlength(bs);
|
|
|
|
}
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
2011-03-29 21:04:41 +02:00
|
|
|
return bs->total_sectors * BDRV_SECTOR_SIZE;
|
2003-06-30 12:03:06 +02:00
|
|
|
}
|
|
|
|
|
2006-08-19 13:45:59 +02:00
|
|
|
/* return 0 as number of sectors if no device present or error */
|
2007-12-17 02:35:20 +01:00
|
|
|
void bdrv_get_geometry(BlockDriverState *bs, uint64_t *nb_sectors_ptr)
|
2003-06-30 12:03:06 +02:00
|
|
|
{
|
2006-08-19 13:45:59 +02:00
|
|
|
int64_t length;
|
|
|
|
length = bdrv_getlength(bs);
|
|
|
|
if (length < 0)
|
|
|
|
length = 0;
|
|
|
|
else
|
2009-11-30 18:21:19 +01:00
|
|
|
length = length >> BDRV_SECTOR_BITS;
|
2006-08-19 13:45:59 +02:00
|
|
|
*nb_sectors_ptr = length;
|
2003-06-30 12:03:06 +02:00
|
|
|
}
|
2004-02-16 22:56:36 +01:00
|
|
|
|
2011-11-03 09:57:25 +01:00
|
|
|
/* throttling disk io limits */
|
|
|
|
void bdrv_set_io_limits(BlockDriverState *bs,
|
|
|
|
BlockIOLimit *io_limits)
|
|
|
|
{
|
|
|
|
bs->io_limits = *io_limits;
|
|
|
|
bs->io_limits_enabled = bdrv_io_limits_enabled(bs);
|
|
|
|
}
|
|
|
|
|
2012-09-28 17:22:54 +02:00
|
|
|
void bdrv_set_on_error(BlockDriverState *bs, BlockdevOnError on_read_error,
|
|
|
|
BlockdevOnError on_write_error)
|
2010-06-02 18:55:17 +02:00
|
|
|
{
|
|
|
|
bs->on_read_error = on_read_error;
|
|
|
|
bs->on_write_error = on_write_error;
|
|
|
|
}
|
|
|
|
|
2012-09-28 17:22:56 +02:00
|
|
|
BlockdevOnError bdrv_get_on_error(BlockDriverState *bs, bool is_read)
|
2010-06-02 18:55:17 +02:00
|
|
|
{
|
|
|
|
return is_read ? bs->on_read_error : bs->on_write_error;
|
|
|
|
}
|
|
|
|
|
2012-09-28 17:22:57 +02:00
|
|
|
BlockErrorAction bdrv_get_error_action(BlockDriverState *bs, bool is_read, int error)
|
|
|
|
{
|
|
|
|
BlockdevOnError on_err = is_read ? bs->on_read_error : bs->on_write_error;
|
|
|
|
|
|
|
|
switch (on_err) {
|
|
|
|
case BLOCKDEV_ON_ERROR_ENOSPC:
|
|
|
|
return (error == ENOSPC) ? BDRV_ACTION_STOP : BDRV_ACTION_REPORT;
|
|
|
|
case BLOCKDEV_ON_ERROR_STOP:
|
|
|
|
return BDRV_ACTION_STOP;
|
|
|
|
case BLOCKDEV_ON_ERROR_REPORT:
|
|
|
|
return BDRV_ACTION_REPORT;
|
|
|
|
case BLOCKDEV_ON_ERROR_IGNORE:
|
|
|
|
return BDRV_ACTION_IGNORE;
|
|
|
|
default:
|
|
|
|
abort();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* This is done by device models because, while the block layer knows
|
|
|
|
* about the error, it does not know whether an operation comes from
|
|
|
|
* the device or the block layer (from a job, for example).
|
|
|
|
*/
|
|
|
|
void bdrv_error_action(BlockDriverState *bs, BlockErrorAction action,
|
|
|
|
bool is_read, int error)
|
|
|
|
{
|
|
|
|
assert(error >= 0);
|
block: introduce block job error
The following behaviors are possible:
'report': The behavior is the same as in 1.1. An I/O error,
respectively during a read or a write, will complete the job immediately
with an error code.
'ignore': An I/O error, respectively during a read or a write, will be
ignored. For streaming, the job will complete with an error and the
backing file will be left in place. For mirroring, the sector will be
marked again as dirty and re-examined later.
'stop': The job will be paused and the job iostatus will be set to
failed or nospace, while the VM will keep running. This can only be
specified if the block device has rerror=stop and werror=stop or enospc.
'enospc': Behaves as 'stop' for ENOSPC errors, 'report' for others.
In all cases, even for 'report', the I/O error is reported as a QMP
event BLOCK_JOB_ERROR, with the same arguments as BLOCK_IO_ERROR.
It is possible that while stopping the VM a BLOCK_IO_ERROR event will be
reported and will clobber the event from BLOCK_JOB_ERROR, or vice versa.
This is not really avoidable since stopping the VM completes all pending
I/O requests. In fact, it is already possible now that a series of
BLOCK_IO_ERROR events are reported with rerror=stop, because vm_stop
calls bdrv_drain_all and this can generate further errors.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2012-09-28 17:22:58 +02:00
|
|
|
bdrv_emit_qmp_error_event(bs, QEVENT_BLOCK_IO_ERROR, action, is_read);
|
2012-09-28 17:22:57 +02:00
|
|
|
if (action == BDRV_ACTION_STOP) {
|
|
|
|
vm_stop(RUN_STATE_IO_ERROR);
|
|
|
|
bdrv_iostatus_set_err(bs, error);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2004-03-14 22:38:54 +01:00
|
|
|
int bdrv_is_read_only(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->read_only;
|
|
|
|
}
|
|
|
|
|
2007-12-24 17:10:43 +01:00
|
|
|
int bdrv_is_sg(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->sg;
|
|
|
|
}
|
|
|
|
|
2009-09-04 19:01:15 +02:00
|
|
|
int bdrv_enable_write_cache(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->enable_write_cache;
|
|
|
|
}
|
|
|
|
|
2012-06-06 00:04:52 +02:00
|
|
|
void bdrv_set_enable_write_cache(BlockDriverState *bs, bool wce)
|
|
|
|
{
|
|
|
|
bs->enable_write_cache = wce;
|
2012-09-20 21:13:18 +02:00
|
|
|
|
|
|
|
/* so a reopen() will preserve wce */
|
|
|
|
if (wce) {
|
|
|
|
bs->open_flags |= BDRV_O_CACHE_WB;
|
|
|
|
} else {
|
|
|
|
bs->open_flags &= ~BDRV_O_CACHE_WB;
|
|
|
|
}
|
2012-06-06 00:04:52 +02:00
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
int bdrv_is_encrypted(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->backing_hd && bs->backing_hd->encrypted)
|
|
|
|
return 1;
|
|
|
|
return bs->encrypted;
|
|
|
|
}
|
|
|
|
|
2009-03-06 00:01:01 +01:00
|
|
|
int bdrv_key_required(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriverState *backing_hd = bs->backing_hd;
|
|
|
|
|
|
|
|
if (backing_hd && backing_hd->encrypted && !backing_hd->valid_key)
|
|
|
|
return 1;
|
|
|
|
return (bs->encrypted && !bs->valid_key);
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
int bdrv_set_key(BlockDriverState *bs, const char *key)
|
|
|
|
{
|
|
|
|
int ret;
|
|
|
|
if (bs->backing_hd && bs->backing_hd->encrypted) {
|
|
|
|
ret = bdrv_set_key(bs->backing_hd, key);
|
|
|
|
if (ret < 0)
|
|
|
|
return ret;
|
|
|
|
if (!bs->encrypted)
|
|
|
|
return 0;
|
|
|
|
}
|
2010-03-05 23:26:13 +01:00
|
|
|
if (!bs->encrypted) {
|
|
|
|
return -EINVAL;
|
|
|
|
} else if (!bs->drv || !bs->drv->bdrv_set_key) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
}
|
2009-03-06 00:01:01 +01:00
|
|
|
ret = bs->drv->bdrv_set_key(bs, key);
|
2009-03-06 00:01:15 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
bs->valid_key = 0;
|
|
|
|
} else if (!bs->valid_key) {
|
|
|
|
bs->valid_key = 1;
|
|
|
|
/* call the change callback now, we skipped it on open */
|
2011-09-06 18:58:59 +02:00
|
|
|
bdrv_dev_change_media_cb(bs, true);
|
2009-03-06 00:01:15 +01:00
|
|
|
}
|
2009-03-06 00:01:01 +01:00
|
|
|
return ret;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2012-06-13 10:11:48 +02:00
|
|
|
const char *bdrv_get_format_name(BlockDriverState *bs)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2012-06-13 10:11:48 +02:00
|
|
|
return bs->drv ? bs->drv->format_name : NULL;
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
void bdrv_iterate_format(void (*it)(void *opaque, const char *name),
|
2004-08-01 23:59:26 +02:00
|
|
|
void *opaque)
|
|
|
|
{
|
|
|
|
BlockDriver *drv;
|
|
|
|
|
2010-04-13 11:29:33 +02:00
|
|
|
QLIST_FOREACH(drv, &bdrv_drivers, list) {
|
2004-08-01 23:59:26 +02:00
|
|
|
it(opaque, drv->format_name);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2004-03-14 22:38:54 +01:00
|
|
|
BlockDriverState *bdrv_find(const char *name)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
if (!strcmp(name, bs->device_name)) {
|
2004-03-14 22:38:54 +01:00
|
|
|
return bs;
|
2010-04-10 08:02:42 +02:00
|
|
|
}
|
2004-03-14 22:38:54 +01:00
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2010-06-02 18:55:20 +02:00
|
|
|
BlockDriverState *bdrv_next(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (!bs) {
|
|
|
|
return QTAILQ_FIRST(&bdrv_states);
|
|
|
|
}
|
|
|
|
return QTAILQ_NEXT(bs, list);
|
|
|
|
}
|
|
|
|
|
2009-03-06 00:00:43 +01:00
|
|
|
void bdrv_iterate(void (*it)(void *opaque, BlockDriverState *bs), void *opaque)
|
2004-07-14 19:21:37 +02:00
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
2009-03-06 00:00:43 +01:00
|
|
|
it(opaque, bs);
|
2004-07-14 19:21:37 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
const char *bdrv_get_device_name(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->device_name;
|
|
|
|
}
|
|
|
|
|
2012-06-05 16:49:24 +02:00
|
|
|
int bdrv_get_flags(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->open_flags;
|
|
|
|
}
|
|
|
|
|
2008-10-06 15:55:43 +02:00
|
|
|
void bdrv_flush_all(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
2012-03-12 18:26:01 +01:00
|
|
|
bdrv_flush(bs);
|
2010-04-10 08:02:42 +02:00
|
|
|
}
|
2008-10-06 15:55:43 +02:00
|
|
|
}
|
|
|
|
|
2010-04-14 17:30:35 +02:00
|
|
|
int bdrv_has_zero_init(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
assert(bs->drv);
|
|
|
|
|
2010-07-28 11:26:29 +02:00
|
|
|
if (bs->drv->bdrv_has_zero_init) {
|
|
|
|
return bs->drv->bdrv_has_zero_init(bs);
|
2010-04-14 17:30:35 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2011-11-14 13:44:19 +01:00
|
|
|
typedef struct BdrvCoIsAllocatedData {
|
|
|
|
BlockDriverState *bs;
|
|
|
|
int64_t sector_num;
|
|
|
|
int nb_sectors;
|
|
|
|
int *pnum;
|
|
|
|
int ret;
|
|
|
|
bool done;
|
|
|
|
} BdrvCoIsAllocatedData;
|
|
|
|
|
2008-06-05 23:53:49 +02:00
|
|
|
/*
|
|
|
|
* Returns true iff the specified sector is present in the disk image. Drivers
|
|
|
|
* not implementing the functionality are assumed to not support backing files,
|
|
|
|
* hence all their sectors are reported as allocated.
|
|
|
|
*
|
2011-11-29 14:49:51 +01:00
|
|
|
* If 'sector_num' is beyond the end of the disk image the return value is 0
|
|
|
|
* and 'pnum' is set to 0.
|
|
|
|
*
|
2008-06-05 23:53:49 +02:00
|
|
|
* 'pnum' is set to the number of sectors (including and immediately following
|
|
|
|
* the specified sector) that are known to be in the same
|
|
|
|
* allocated/unallocated state.
|
|
|
|
*
|
2011-11-29 14:49:51 +01:00
|
|
|
* 'nb_sectors' is the max value 'pnum' should be set to. If nb_sectors goes
|
|
|
|
* beyond the end of the disk image it will be clamped.
|
2008-06-05 23:53:49 +02:00
|
|
|
*/
|
2011-11-14 13:44:26 +01:00
|
|
|
int coroutine_fn bdrv_co_is_allocated(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
int nb_sectors, int *pnum)
|
2008-06-05 23:53:49 +02:00
|
|
|
{
|
2011-11-29 14:49:51 +01:00
|
|
|
int64_t n;
|
|
|
|
|
|
|
|
if (sector_num >= bs->total_sectors) {
|
|
|
|
*pnum = 0;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
n = bs->total_sectors - sector_num;
|
|
|
|
if (n < nb_sectors) {
|
|
|
|
nb_sectors = n;
|
|
|
|
}
|
|
|
|
|
2011-11-14 13:44:25 +01:00
|
|
|
if (!bs->drv->bdrv_co_is_allocated) {
|
2011-11-29 14:49:51 +01:00
|
|
|
*pnum = nb_sectors;
|
2008-06-05 23:53:49 +02:00
|
|
|
return 1;
|
|
|
|
}
|
2011-11-14 13:44:25 +01:00
|
|
|
|
2011-11-14 13:44:26 +01:00
|
|
|
return bs->drv->bdrv_co_is_allocated(bs, sector_num, nb_sectors, pnum);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Coroutine wrapper for bdrv_is_allocated() */
|
|
|
|
static void coroutine_fn bdrv_is_allocated_co_entry(void *opaque)
|
|
|
|
{
|
|
|
|
BdrvCoIsAllocatedData *data = opaque;
|
|
|
|
BlockDriverState *bs = data->bs;
|
|
|
|
|
|
|
|
data->ret = bdrv_co_is_allocated(bs, data->sector_num, data->nb_sectors,
|
|
|
|
data->pnum);
|
|
|
|
data->done = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Synchronous wrapper around bdrv_co_is_allocated().
|
|
|
|
*
|
|
|
|
* See bdrv_co_is_allocated() for details.
|
|
|
|
*/
|
|
|
|
int bdrv_is_allocated(BlockDriverState *bs, int64_t sector_num, int nb_sectors,
|
|
|
|
int *pnum)
|
|
|
|
{
|
2011-11-14 13:44:25 +01:00
|
|
|
Coroutine *co;
|
|
|
|
BdrvCoIsAllocatedData data = {
|
|
|
|
.bs = bs,
|
|
|
|
.sector_num = sector_num,
|
|
|
|
.nb_sectors = nb_sectors,
|
|
|
|
.pnum = pnum,
|
|
|
|
.done = false,
|
|
|
|
};
|
|
|
|
|
|
|
|
co = qemu_coroutine_create(bdrv_is_allocated_co_entry);
|
|
|
|
qemu_coroutine_enter(co, &data);
|
|
|
|
while (!data.done) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
|
|
|
return data.ret;
|
2008-06-05 23:53:49 +02:00
|
|
|
}
|
|
|
|
|
2012-05-08 16:52:01 +02:00
|
|
|
/*
|
|
|
|
* Given an image chain: ... -> [BASE] -> [INTER1] -> [INTER2] -> [TOP]
|
|
|
|
*
|
|
|
|
* Return true if the given sector is allocated in any image between
|
|
|
|
* BASE and TOP (inclusive). BASE can be NULL to check if the given
|
|
|
|
* sector is allocated in any image of the chain. Return false otherwise.
|
|
|
|
*
|
|
|
|
* 'pnum' is set to the number of sectors (including and immediately following
|
|
|
|
* the specified sector) that are known to be in the same
|
|
|
|
* allocated/unallocated state.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
int coroutine_fn bdrv_co_is_allocated_above(BlockDriverState *top,
|
|
|
|
BlockDriverState *base,
|
|
|
|
int64_t sector_num,
|
|
|
|
int nb_sectors, int *pnum)
|
|
|
|
{
|
|
|
|
BlockDriverState *intermediate;
|
|
|
|
int ret, n = nb_sectors;
|
|
|
|
|
|
|
|
intermediate = top;
|
|
|
|
while (intermediate && intermediate != base) {
|
|
|
|
int pnum_inter;
|
|
|
|
ret = bdrv_co_is_allocated(intermediate, sector_num, nb_sectors,
|
|
|
|
&pnum_inter);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
} else if (ret) {
|
|
|
|
*pnum = pnum_inter;
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* [sector_num, nb_sectors] is unallocated on top but intermediate
|
|
|
|
* might have
|
|
|
|
*
|
|
|
|
* [sector_num+x, nr_sectors] allocated.
|
|
|
|
*/
|
|
|
|
if (n > pnum_inter) {
|
|
|
|
n = pnum_inter;
|
|
|
|
}
|
|
|
|
|
|
|
|
intermediate = intermediate->backing_hd;
|
|
|
|
}
|
|
|
|
|
|
|
|
*pnum = n;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
BlockInfo *bdrv_query_info(BlockDriverState *bs)
|
2004-03-14 22:38:54 +01:00
|
|
|
{
|
2012-10-18 16:49:15 +02:00
|
|
|
BlockInfo *info = g_malloc0(sizeof(*info));
|
|
|
|
info->device = g_strdup(bs->device_name);
|
|
|
|
info->type = g_strdup("unknown");
|
|
|
|
info->locked = bdrv_dev_is_medium_locked(bs);
|
|
|
|
info->removable = bdrv_dev_has_removable_media(bs);
|
2004-03-14 22:38:54 +01:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
if (bdrv_dev_has_removable_media(bs)) {
|
|
|
|
info->has_tray_open = true;
|
|
|
|
info->tray_open = bdrv_dev_is_tray_open(bs);
|
|
|
|
}
|
2009-12-10 20:16:06 +01:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
if (bdrv_iostatus_is_enabled(bs)) {
|
|
|
|
info->has_io_status = true;
|
|
|
|
info->io_status = bs->iostatus;
|
|
|
|
}
|
2009-12-10 20:16:06 +01:00
|
|
|
|
2012-08-01 15:23:44 +02:00
|
|
|
if (bs->dirty_bitmap) {
|
|
|
|
info->has_dirty = true;
|
|
|
|
info->dirty = g_malloc0(sizeof(*info->dirty));
|
|
|
|
info->dirty->count = bdrv_get_dirty_count(bs) *
|
|
|
|
BDRV_SECTORS_PER_DIRTY_CHUNK * BDRV_SECTOR_SIZE;
|
|
|
|
}
|
2009-12-10 20:16:06 +01:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
if (bs->drv) {
|
|
|
|
info->has_inserted = true;
|
|
|
|
info->inserted = g_malloc0(sizeof(*info->inserted));
|
|
|
|
info->inserted->file = g_strdup(bs->filename);
|
|
|
|
info->inserted->ro = bs->read_only;
|
|
|
|
info->inserted->drv = g_strdup(bs->drv->format_name);
|
|
|
|
info->inserted->encrypted = bs->encrypted;
|
|
|
|
info->inserted->encryption_key_missing = bdrv_key_required(bs);
|
|
|
|
|
|
|
|
if (bs->backing_file[0]) {
|
|
|
|
info->inserted->has_backing_file = true;
|
|
|
|
info->inserted->backing_file = g_strdup(bs->backing_file);
|
2011-09-06 18:58:53 +02:00
|
|
|
}
|
2011-09-26 22:43:54 +02:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
info->inserted->backing_file_depth = bdrv_get_backing_file_depth(bs);
|
|
|
|
|
|
|
|
if (bs->io_limits_enabled) {
|
|
|
|
info->inserted->bps =
|
|
|
|
bs->io_limits.bps[BLOCK_IO_LIMIT_TOTAL];
|
|
|
|
info->inserted->bps_rd =
|
|
|
|
bs->io_limits.bps[BLOCK_IO_LIMIT_READ];
|
|
|
|
info->inserted->bps_wr =
|
|
|
|
bs->io_limits.bps[BLOCK_IO_LIMIT_WRITE];
|
|
|
|
info->inserted->iops =
|
|
|
|
bs->io_limits.iops[BLOCK_IO_LIMIT_TOTAL];
|
|
|
|
info->inserted->iops_rd =
|
|
|
|
bs->io_limits.iops[BLOCK_IO_LIMIT_READ];
|
|
|
|
info->inserted->iops_wr =
|
|
|
|
bs->io_limits.iops[BLOCK_IO_LIMIT_WRITE];
|
2011-09-26 22:43:54 +02:00
|
|
|
}
|
2012-10-18 16:49:15 +02:00
|
|
|
}
|
|
|
|
return info;
|
|
|
|
}
|
2011-09-26 22:43:54 +02:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
BlockInfoList *qmp_query_block(Error **errp)
|
|
|
|
{
|
|
|
|
BlockInfoList *head = NULL, **p_next = &head;
|
|
|
|
BlockDriverState *bs;
|
2011-11-08 06:00:31 +01:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
BlockInfoList *info = g_malloc0(sizeof(*info));
|
|
|
|
info->value = bdrv_query_info(bs);
|
2009-12-10 20:16:06 +01:00
|
|
|
|
2012-10-18 16:49:15 +02:00
|
|
|
*p_next = info;
|
|
|
|
p_next = &info->next;
|
2004-03-14 22:38:54 +01:00
|
|
|
}
|
2009-12-10 20:16:06 +01:00
|
|
|
|
2011-09-21 22:16:47 +02:00
|
|
|
return head;
|
2004-03-14 22:38:54 +01:00
|
|
|
}
|
2007-12-02 06:18:19 +01:00
|
|
|
|
2012-10-18 16:49:16 +02:00
|
|
|
BlockStats *bdrv_query_stats(const BlockDriverState *bs)
|
2011-09-22 20:56:36 +02:00
|
|
|
{
|
|
|
|
BlockStats *s;
|
|
|
|
|
|
|
|
s = g_malloc0(sizeof(*s));
|
|
|
|
|
|
|
|
if (bs->device_name[0]) {
|
|
|
|
s->has_device = true;
|
|
|
|
s->device = g_strdup(bs->device_name);
|
2010-04-28 14:34:01 +02:00
|
|
|
}
|
|
|
|
|
2011-09-22 20:56:36 +02:00
|
|
|
s->stats = g_malloc0(sizeof(*s->stats));
|
|
|
|
s->stats->rd_bytes = bs->nr_bytes[BDRV_ACCT_READ];
|
|
|
|
s->stats->wr_bytes = bs->nr_bytes[BDRV_ACCT_WRITE];
|
|
|
|
s->stats->rd_operations = bs->nr_ops[BDRV_ACCT_READ];
|
|
|
|
s->stats->wr_operations = bs->nr_ops[BDRV_ACCT_WRITE];
|
|
|
|
s->stats->wr_highest_offset = bs->wr_highest_sector * BDRV_SECTOR_SIZE;
|
|
|
|
s->stats->flush_operations = bs->nr_ops[BDRV_ACCT_FLUSH];
|
|
|
|
s->stats->wr_total_time_ns = bs->total_time_ns[BDRV_ACCT_WRITE];
|
|
|
|
s->stats->rd_total_time_ns = bs->total_time_ns[BDRV_ACCT_READ];
|
|
|
|
s->stats->flush_total_time_ns = bs->total_time_ns[BDRV_ACCT_FLUSH];
|
|
|
|
|
2010-04-28 14:34:01 +02:00
|
|
|
if (bs->file) {
|
2011-09-22 20:56:36 +02:00
|
|
|
s->has_parent = true;
|
2012-10-18 16:49:16 +02:00
|
|
|
s->parent = bdrv_query_stats(bs->file);
|
2010-04-28 14:34:01 +02:00
|
|
|
}
|
|
|
|
|
2011-09-22 20:56:36 +02:00
|
|
|
return s;
|
2010-04-28 14:34:01 +02:00
|
|
|
}
|
|
|
|
|
2011-09-22 20:56:36 +02:00
|
|
|
BlockStatsList *qmp_query_blockstats(Error **errp)
|
2009-12-10 20:16:07 +01:00
|
|
|
{
|
2012-10-18 16:49:16 +02:00
|
|
|
BlockStatsList *head = NULL, **p_next = &head;
|
2007-12-02 06:18:19 +01:00
|
|
|
BlockDriverState *bs;
|
|
|
|
|
2010-04-10 08:02:42 +02:00
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
2011-09-22 20:56:36 +02:00
|
|
|
BlockStatsList *info = g_malloc0(sizeof(*info));
|
2012-10-18 16:49:16 +02:00
|
|
|
info->value = bdrv_query_stats(bs);
|
2011-09-22 20:56:36 +02:00
|
|
|
|
2012-10-18 16:49:16 +02:00
|
|
|
*p_next = info;
|
|
|
|
p_next = &info->next;
|
2007-12-02 06:18:19 +01:00
|
|
|
}
|
2009-12-10 20:16:07 +01:00
|
|
|
|
2011-09-22 20:56:36 +02:00
|
|
|
return head;
|
2007-12-02 06:18:19 +01:00
|
|
|
}
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2009-03-06 00:00:48 +01:00
|
|
|
const char *bdrv_get_encrypted_filename(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->backing_hd && bs->backing_hd->encrypted)
|
|
|
|
return bs->backing_file;
|
|
|
|
else if (bs->encrypted)
|
|
|
|
return bs->filename;
|
|
|
|
else
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
void bdrv_get_backing_filename(BlockDriverState *bs,
|
2006-08-01 18:21:11 +02:00
|
|
|
char *filename, int filename_size)
|
|
|
|
{
|
2011-10-26 11:02:11 +02:00
|
|
|
pstrcpy(filename, filename_size, bs->backing_file);
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
int bdrv_write_compressed(BlockDriverState *bs, int64_t sector_num,
|
2006-08-05 23:31:00 +02:00
|
|
|
const uint8_t *buf, int nb_sectors)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2006-08-05 23:31:00 +02:00
|
|
|
if (!drv->bdrv_write_compressed)
|
|
|
|
return -ENOTSUP;
|
2009-05-08 14:47:24 +02:00
|
|
|
if (bdrv_check_request(bs, sector_num, nb_sectors))
|
|
|
|
return -EIO;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2012-10-18 16:49:18 +02:00
|
|
|
assert(!bs->dirty_bitmap);
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2006-08-05 23:31:00 +02:00
|
|
|
return drv->bdrv_write_compressed(bs, sector_num, buf, nb_sectors);
|
|
|
|
}
|
2007-09-17 10:09:54 +02:00
|
|
|
|
2006-08-05 23:31:00 +02:00
|
|
|
int bdrv_get_info(BlockDriverState *bs, BlockDriverInfo *bdi)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2006-08-05 23:31:00 +02:00
|
|
|
if (!drv->bdrv_get_info)
|
|
|
|
return -ENOTSUP;
|
|
|
|
memset(bdi, 0, sizeof(*bdi));
|
|
|
|
return drv->bdrv_get_info(bs, bdi);
|
|
|
|
}
|
|
|
|
|
2009-07-10 23:11:57 +02:00
|
|
|
int bdrv_save_vmstate(BlockDriverState *bs, const uint8_t *buf,
|
|
|
|
int64_t pos, int size)
|
2009-04-05 21:10:55 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_save_vmstate)
|
|
|
|
return drv->bdrv_save_vmstate(bs, buf, pos, size);
|
|
|
|
if (bs->file)
|
|
|
|
return bdrv_save_vmstate(bs->file, buf, pos, size);
|
|
|
|
return -ENOTSUP;
|
2009-04-05 21:10:55 +02:00
|
|
|
}
|
|
|
|
|
2009-07-10 23:11:57 +02:00
|
|
|
int bdrv_load_vmstate(BlockDriverState *bs, uint8_t *buf,
|
|
|
|
int64_t pos, int size)
|
2009-04-05 21:10:55 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_load_vmstate)
|
|
|
|
return drv->bdrv_load_vmstate(bs, buf, pos, size);
|
|
|
|
if (bs->file)
|
|
|
|
return bdrv_load_vmstate(bs->file, buf, pos, size);
|
|
|
|
return -ENOTSUP;
|
2009-04-05 21:10:55 +02:00
|
|
|
}
|
|
|
|
|
2010-03-15 17:27:00 +01:00
|
|
|
void bdrv_debug_event(BlockDriverState *bs, BlkDebugEvent event)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
|
|
|
if (!drv || !drv->bdrv_debug_event) {
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
2012-07-08 08:56:53 +02:00
|
|
|
drv->bdrv_debug_event(bs, event);
|
2012-12-06 14:32:58 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_debug_breakpoint(BlockDriverState *bs, const char *event,
|
|
|
|
const char *tag)
|
|
|
|
{
|
|
|
|
while (bs && bs->drv && !bs->drv->bdrv_debug_breakpoint) {
|
|
|
|
bs = bs->file;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bs && bs->drv && bs->drv->bdrv_debug_breakpoint) {
|
|
|
|
return bs->drv->bdrv_debug_breakpoint(bs, event, tag);
|
|
|
|
}
|
|
|
|
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_debug_resume(BlockDriverState *bs, const char *tag)
|
|
|
|
{
|
|
|
|
while (bs && bs->drv && !bs->drv->bdrv_debug_resume) {
|
|
|
|
bs = bs->file;
|
|
|
|
}
|
2010-03-15 17:27:00 +01:00
|
|
|
|
2012-12-06 14:32:58 +01:00
|
|
|
if (bs && bs->drv && bs->drv->bdrv_debug_resume) {
|
|
|
|
return bs->drv->bdrv_debug_resume(bs, tag);
|
|
|
|
}
|
|
|
|
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
|
|
|
bool bdrv_debug_is_suspended(BlockDriverState *bs, const char *tag)
|
|
|
|
{
|
|
|
|
while (bs && bs->drv && !bs->drv->bdrv_debug_is_suspended) {
|
|
|
|
bs = bs->file;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (bs && bs->drv && bs->drv->bdrv_debug_is_suspended) {
|
|
|
|
return bs->drv->bdrv_debug_is_suspended(bs, tag);
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
2010-03-15 17:27:00 +01:00
|
|
|
}
|
|
|
|
|
2006-08-05 23:31:00 +02:00
|
|
|
/**************************************************************/
|
|
|
|
/* handling of snapshots */
|
|
|
|
|
savevm: Really verify if a drive supports snapshots
Both bdrv_can_snapshot() and bdrv_has_snapshot() does not work as advertized.
First issue: Their names implies different porpouses, but they do the same thing
and have exactly the same code. Maybe copied and pasted and forgotten?
bdrv_has_snapshot() is called in various places for actually checking if there
is snapshots or not.
Second issue: the way bdrv_can_snapshot() verifies if a block driver supports or
not snapshots does not catch all cases. E.g.: a raw image.
So when do_savevm() is called, first thing it does is to set a global
BlockDriverState to save the VM memory state calling get_bs_snapshots().
static BlockDriverState *get_bs_snapshots(void)
{
BlockDriverState *bs;
DriveInfo *dinfo;
if (bs_snapshots)
return bs_snapshots;
QTAILQ_FOREACH(dinfo, &drives, next) {
bs = dinfo->bdrv;
if (bdrv_can_snapshot(bs))
goto ok;
}
return NULL;
ok:
bs_snapshots = bs;
return bs;
}
bdrv_can_snapshot() may return a BlockDriverState that does not support
snapshots and do_savevm() goes on.
Later on in do_savevm(), we find:
QTAILQ_FOREACH(dinfo, &drives, next) {
bs1 = dinfo->bdrv;
if (bdrv_has_snapshot(bs1)) {
/* Write VM state size only to the image that contains the state */
sn->vm_state_size = (bs == bs1 ? vm_state_size : 0);
ret = bdrv_snapshot_create(bs1, sn);
if (ret < 0) {
monitor_printf(mon, "Error while creating snapshot on '%s'\n",
bdrv_get_device_name(bs1));
}
}
}
bdrv_has_snapshot(bs1) is not checking if the device does support or has
snapshots as explained above. Only in bdrv_snapshot_create() the device is
actually checked for snapshot support.
So, in cases where the first device supports snapshots, and the second does not,
the snapshot on the first will happen anyways. I believe this is not a good
behavior. It should be an all or nothing process.
This patch addresses these issues by making bdrv_can_snapshot() actually do
what it must do and enforces better tests to avoid errors in the middle of
do_savevm(). bdrv_has_snapshot() is removed and replaced by bdrv_can_snapshot()
where appropriate.
bdrv_can_snapshot() was moved from savevm.c to block.c. It makes more sense to me.
The loadvm_state() function was updated too to enforce that when loading a VM at
least all writable devices must support snapshots too.
Signed-off-by: Miguel Di Ciurcio Filho <miguel.filho@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2010-06-08 15:40:55 +02:00
|
|
|
int bdrv_can_snapshot(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2011-08-03 15:08:11 +02:00
|
|
|
if (!drv || !bdrv_is_inserted(bs) || bdrv_is_read_only(bs)) {
|
savevm: Really verify if a drive supports snapshots
Both bdrv_can_snapshot() and bdrv_has_snapshot() does not work as advertized.
First issue: Their names implies different porpouses, but they do the same thing
and have exactly the same code. Maybe copied and pasted and forgotten?
bdrv_has_snapshot() is called in various places for actually checking if there
is snapshots or not.
Second issue: the way bdrv_can_snapshot() verifies if a block driver supports or
not snapshots does not catch all cases. E.g.: a raw image.
So when do_savevm() is called, first thing it does is to set a global
BlockDriverState to save the VM memory state calling get_bs_snapshots().
static BlockDriverState *get_bs_snapshots(void)
{
BlockDriverState *bs;
DriveInfo *dinfo;
if (bs_snapshots)
return bs_snapshots;
QTAILQ_FOREACH(dinfo, &drives, next) {
bs = dinfo->bdrv;
if (bdrv_can_snapshot(bs))
goto ok;
}
return NULL;
ok:
bs_snapshots = bs;
return bs;
}
bdrv_can_snapshot() may return a BlockDriverState that does not support
snapshots and do_savevm() goes on.
Later on in do_savevm(), we find:
QTAILQ_FOREACH(dinfo, &drives, next) {
bs1 = dinfo->bdrv;
if (bdrv_has_snapshot(bs1)) {
/* Write VM state size only to the image that contains the state */
sn->vm_state_size = (bs == bs1 ? vm_state_size : 0);
ret = bdrv_snapshot_create(bs1, sn);
if (ret < 0) {
monitor_printf(mon, "Error while creating snapshot on '%s'\n",
bdrv_get_device_name(bs1));
}
}
}
bdrv_has_snapshot(bs1) is not checking if the device does support or has
snapshots as explained above. Only in bdrv_snapshot_create() the device is
actually checked for snapshot support.
So, in cases where the first device supports snapshots, and the second does not,
the snapshot on the first will happen anyways. I believe this is not a good
behavior. It should be an all or nothing process.
This patch addresses these issues by making bdrv_can_snapshot() actually do
what it must do and enforces better tests to avoid errors in the middle of
do_savevm(). bdrv_has_snapshot() is removed and replaced by bdrv_can_snapshot()
where appropriate.
bdrv_can_snapshot() was moved from savevm.c to block.c. It makes more sense to me.
The loadvm_state() function was updated too to enforce that when loading a VM at
least all writable devices must support snapshots too.
Signed-off-by: Miguel Di Ciurcio Filho <miguel.filho@gmail.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2010-06-08 15:40:55 +02:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!drv->bdrv_snapshot_create) {
|
|
|
|
if (bs->file != NULL) {
|
|
|
|
return bdrv_can_snapshot(bs->file);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 1;
|
|
|
|
}
|
|
|
|
|
2010-07-25 22:49:34 +02:00
|
|
|
int bdrv_is_snapshot(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return !!(bs->open_flags & BDRV_O_SNAPSHOT);
|
|
|
|
}
|
|
|
|
|
2010-06-25 10:33:39 +02:00
|
|
|
BlockDriverState *bdrv_snapshots(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
2010-07-01 09:30:38 +02:00
|
|
|
if (bs_snapshots) {
|
2010-06-25 10:33:39 +02:00
|
|
|
return bs_snapshots;
|
2010-07-01 09:30:38 +02:00
|
|
|
}
|
2010-06-25 10:33:39 +02:00
|
|
|
|
|
|
|
bs = NULL;
|
|
|
|
while ((bs = bdrv_next(bs))) {
|
|
|
|
if (bdrv_can_snapshot(bs)) {
|
2010-07-01 09:30:38 +02:00
|
|
|
bs_snapshots = bs;
|
|
|
|
return bs;
|
2010-06-25 10:33:39 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
int bdrv_snapshot_create(BlockDriverState *bs,
|
2006-08-05 23:31:00 +02:00
|
|
|
QEMUSnapshotInfo *sn_info)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_snapshot_create)
|
|
|
|
return drv->bdrv_snapshot_create(bs, sn_info);
|
|
|
|
if (bs->file)
|
|
|
|
return bdrv_snapshot_create(bs->file, sn_info);
|
|
|
|
return -ENOTSUP;
|
2006-08-05 23:31:00 +02:00
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
int bdrv_snapshot_goto(BlockDriverState *bs,
|
2006-08-05 23:31:00 +02:00
|
|
|
const char *snapshot_id)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2010-05-28 04:44:58 +02:00
|
|
|
int ret, open_ret;
|
|
|
|
|
2006-08-05 23:31:00 +02:00
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_snapshot_goto)
|
|
|
|
return drv->bdrv_snapshot_goto(bs, snapshot_id);
|
|
|
|
|
|
|
|
if (bs->file) {
|
|
|
|
drv->bdrv_close(bs);
|
|
|
|
ret = bdrv_snapshot_goto(bs->file, snapshot_id);
|
|
|
|
open_ret = drv->bdrv_open(bs, bs->open_flags);
|
|
|
|
if (open_ret < 0) {
|
|
|
|
bdrv_delete(bs->file);
|
|
|
|
bs->drv = NULL;
|
|
|
|
return open_ret;
|
|
|
|
}
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
return -ENOTSUP;
|
2006-08-05 23:31:00 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_snapshot_delete(BlockDriverState *bs, const char *snapshot_id)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_snapshot_delete)
|
|
|
|
return drv->bdrv_snapshot_delete(bs, snapshot_id);
|
|
|
|
if (bs->file)
|
|
|
|
return bdrv_snapshot_delete(bs->file, snapshot_id);
|
|
|
|
return -ENOTSUP;
|
2006-08-05 23:31:00 +02:00
|
|
|
}
|
|
|
|
|
2007-09-16 23:08:06 +02:00
|
|
|
int bdrv_snapshot_list(BlockDriverState *bs,
|
2006-08-05 23:31:00 +02:00
|
|
|
QEMUSnapshotInfo **psn_info)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv)
|
2006-08-19 13:45:59 +02:00
|
|
|
return -ENOMEDIUM;
|
2010-05-28 04:44:58 +02:00
|
|
|
if (drv->bdrv_snapshot_list)
|
|
|
|
return drv->bdrv_snapshot_list(bs, psn_info);
|
|
|
|
if (bs->file)
|
|
|
|
return bdrv_snapshot_list(bs->file, psn_info);
|
|
|
|
return -ENOTSUP;
|
2006-08-05 23:31:00 +02:00
|
|
|
}
|
|
|
|
|
2010-09-22 04:58:41 +02:00
|
|
|
int bdrv_snapshot_load_tmp(BlockDriverState *bs,
|
|
|
|
const char *snapshot_name)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
if (!drv) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
}
|
|
|
|
if (!bs->read_only) {
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
if (drv->bdrv_snapshot_load_tmp) {
|
|
|
|
return drv->bdrv_snapshot_load_tmp(bs, snapshot_name);
|
|
|
|
}
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
|
|
|
|
2012-10-16 21:49:09 +02:00
|
|
|
/* backing_file can either be relative, or absolute, or a protocol. If it is
|
|
|
|
* relative, it must be relative to the chain. So, passing in bs->filename
|
|
|
|
* from a BDS as backing_file should not be done, as that may be relative to
|
|
|
|
* the CWD rather than the chain. */
|
2012-01-18 15:40:51 +01:00
|
|
|
BlockDriverState *bdrv_find_backing_image(BlockDriverState *bs,
|
|
|
|
const char *backing_file)
|
|
|
|
{
|
2012-10-16 21:49:09 +02:00
|
|
|
char *filename_full = NULL;
|
|
|
|
char *backing_file_full = NULL;
|
|
|
|
char *filename_tmp = NULL;
|
|
|
|
int is_protocol = 0;
|
|
|
|
BlockDriverState *curr_bs = NULL;
|
|
|
|
BlockDriverState *retval = NULL;
|
|
|
|
|
|
|
|
if (!bs || !bs->drv || !backing_file) {
|
2012-01-18 15:40:51 +01:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2012-10-16 21:49:09 +02:00
|
|
|
filename_full = g_malloc(PATH_MAX);
|
|
|
|
backing_file_full = g_malloc(PATH_MAX);
|
|
|
|
filename_tmp = g_malloc(PATH_MAX);
|
|
|
|
|
|
|
|
is_protocol = path_has_protocol(backing_file);
|
|
|
|
|
|
|
|
for (curr_bs = bs; curr_bs->backing_hd; curr_bs = curr_bs->backing_hd) {
|
|
|
|
|
|
|
|
/* If either of the filename paths is actually a protocol, then
|
|
|
|
* compare unmodified paths; otherwise make paths relative */
|
|
|
|
if (is_protocol || path_has_protocol(curr_bs->backing_file)) {
|
|
|
|
if (strcmp(backing_file, curr_bs->backing_file) == 0) {
|
|
|
|
retval = curr_bs->backing_hd;
|
|
|
|
break;
|
|
|
|
}
|
2012-01-18 15:40:51 +01:00
|
|
|
} else {
|
2012-10-16 21:49:09 +02:00
|
|
|
/* If not an absolute filename path, make it relative to the current
|
|
|
|
* image's filename path */
|
|
|
|
path_combine(filename_tmp, PATH_MAX, curr_bs->filename,
|
|
|
|
backing_file);
|
|
|
|
|
|
|
|
/* We are going to compare absolute pathnames */
|
|
|
|
if (!realpath(filename_tmp, filename_full)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* We need to make sure the backing filename we are comparing against
|
|
|
|
* is relative to the current image filename (or absolute) */
|
|
|
|
path_combine(filename_tmp, PATH_MAX, curr_bs->filename,
|
|
|
|
curr_bs->backing_file);
|
|
|
|
|
|
|
|
if (!realpath(filename_tmp, backing_file_full)) {
|
|
|
|
continue;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (strcmp(backing_file_full, filename_full) == 0) {
|
|
|
|
retval = curr_bs->backing_hd;
|
|
|
|
break;
|
|
|
|
}
|
2012-01-18 15:40:51 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-10-16 21:49:09 +02:00
|
|
|
g_free(filename_full);
|
|
|
|
g_free(backing_file_full);
|
|
|
|
g_free(filename_tmp);
|
|
|
|
return retval;
|
2012-01-18 15:40:51 +01:00
|
|
|
}
|
|
|
|
|
2012-08-02 10:22:47 +02:00
|
|
|
int bdrv_get_backing_file_depth(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (!bs->drv) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!bs->backing_hd) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return 1 + bdrv_get_backing_file_depth(bs->backing_hd);
|
|
|
|
}
|
|
|
|
|
2012-09-27 19:29:15 +02:00
|
|
|
BlockDriverState *bdrv_find_base(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriverState *curr_bs = NULL;
|
|
|
|
|
|
|
|
if (!bs) {
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
curr_bs = bs;
|
|
|
|
|
|
|
|
while (curr_bs->backing_hd) {
|
|
|
|
curr_bs = curr_bs->backing_hd;
|
|
|
|
}
|
|
|
|
return curr_bs;
|
|
|
|
}
|
|
|
|
|
2006-08-05 23:31:00 +02:00
|
|
|
#define NB_SUFFIXES 4
|
|
|
|
|
|
|
|
char *get_human_readable_size(char *buf, int buf_size, int64_t size)
|
|
|
|
{
|
|
|
|
static const char suffixes[NB_SUFFIXES] = "KMGT";
|
|
|
|
int64_t base;
|
|
|
|
int i;
|
|
|
|
|
|
|
|
if (size <= 999) {
|
|
|
|
snprintf(buf, buf_size, "%" PRId64, size);
|
|
|
|
} else {
|
|
|
|
base = 1024;
|
|
|
|
for(i = 0; i < NB_SUFFIXES; i++) {
|
|
|
|
if (size < (10 * base)) {
|
2007-09-16 23:08:06 +02:00
|
|
|
snprintf(buf, buf_size, "%0.1f%c",
|
2006-08-05 23:31:00 +02:00
|
|
|
(double)size / base,
|
|
|
|
suffixes[i]);
|
|
|
|
break;
|
|
|
|
} else if (size < (1000 * base) || i == (NB_SUFFIXES - 1)) {
|
2007-09-16 23:08:06 +02:00
|
|
|
snprintf(buf, buf_size, "%" PRId64 "%c",
|
2006-08-05 23:31:00 +02:00
|
|
|
((size + (base >> 1)) / base),
|
|
|
|
suffixes[i]);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
base = base * 1024;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return buf;
|
|
|
|
}
|
|
|
|
|
|
|
|
char *bdrv_snapshot_dump(char *buf, int buf_size, QEMUSnapshotInfo *sn)
|
|
|
|
{
|
|
|
|
char buf1[128], date_buf[128], clock_buf[128];
|
|
|
|
struct tm tm;
|
|
|
|
time_t ti;
|
|
|
|
int64_t secs;
|
|
|
|
|
|
|
|
if (!sn) {
|
2007-09-16 23:08:06 +02:00
|
|
|
snprintf(buf, buf_size,
|
|
|
|
"%-10s%-20s%7s%20s%15s",
|
2006-08-05 23:31:00 +02:00
|
|
|
"ID", "TAG", "VM SIZE", "DATE", "VM CLOCK");
|
|
|
|
} else {
|
|
|
|
ti = sn->date_sec;
|
|
|
|
localtime_r(&ti, &tm);
|
|
|
|
strftime(date_buf, sizeof(date_buf),
|
|
|
|
"%Y-%m-%d %H:%M:%S", &tm);
|
|
|
|
secs = sn->vm_clock_nsec / 1000000000;
|
|
|
|
snprintf(clock_buf, sizeof(clock_buf),
|
|
|
|
"%02d:%02d:%02d.%03d",
|
|
|
|
(int)(secs / 3600),
|
|
|
|
(int)((secs / 60) % 60),
|
2007-09-16 23:08:06 +02:00
|
|
|
(int)(secs % 60),
|
2006-08-05 23:31:00 +02:00
|
|
|
(int)((sn->vm_clock_nsec / 1000000) % 1000));
|
|
|
|
snprintf(buf, buf_size,
|
2007-09-16 23:08:06 +02:00
|
|
|
"%-10s%-20s%7s%20s%15s",
|
2006-08-05 23:31:00 +02:00
|
|
|
sn->id_str, sn->name,
|
|
|
|
get_human_readable_size(buf1, sizeof(buf1), sn->vm_state_size),
|
|
|
|
date_buf,
|
|
|
|
clock_buf);
|
|
|
|
}
|
|
|
|
return buf;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
/**************************************************************/
|
2006-08-01 18:21:11 +02:00
|
|
|
/* async I/Os */
|
2004-08-01 23:59:26 +02:00
|
|
|
|
2009-01-22 17:59:24 +01:00
|
|
|
BlockDriverAIOCB *bdrv_aio_readv(BlockDriverState *bs, int64_t sector_num,
|
2009-04-07 20:43:24 +02:00
|
|
|
QEMUIOVector *qiov, int nb_sectors,
|
2009-01-22 17:59:24 +01:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
2010-10-05 15:28:53 +02:00
|
|
|
trace_bdrv_aio_readv(bs, sector_num, nb_sectors, opaque);
|
|
|
|
|
2011-10-13 14:08:23 +02:00
|
|
|
return bdrv_co_aio_rw_vector(bs, sector_num, qiov, nb_sectors,
|
2011-10-13 22:09:28 +02:00
|
|
|
cb, opaque, false);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
|
|
|
|
2009-04-07 20:43:24 +02:00
|
|
|
BlockDriverAIOCB *bdrv_aio_writev(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
QEMUIOVector *qiov, int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
2004-08-01 23:59:26 +02:00
|
|
|
{
|
2010-10-05 15:28:53 +02:00
|
|
|
trace_bdrv_aio_writev(bs, sector_num, nb_sectors, opaque);
|
|
|
|
|
2011-10-13 14:08:25 +02:00
|
|
|
return bdrv_co_aio_rw_vector(bs, sector_num, qiov, nb_sectors,
|
2011-10-13 22:09:28 +02:00
|
|
|
cb, opaque, true);
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2009-09-09 17:53:37 +02:00
|
|
|
|
|
|
|
typedef struct MultiwriteCB {
|
|
|
|
int error;
|
|
|
|
int num_requests;
|
|
|
|
int num_callbacks;
|
|
|
|
struct {
|
|
|
|
BlockDriverCompletionFunc *cb;
|
|
|
|
void *opaque;
|
|
|
|
QEMUIOVector *free_qiov;
|
|
|
|
} callbacks[];
|
|
|
|
} MultiwriteCB;
|
|
|
|
|
|
|
|
static void multiwrite_user_cb(MultiwriteCB *mcb)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < mcb->num_callbacks; i++) {
|
|
|
|
mcb->callbacks[i].cb(mcb->callbacks[i].opaque, mcb->error);
|
2010-04-21 21:35:45 +02:00
|
|
|
if (mcb->callbacks[i].free_qiov) {
|
|
|
|
qemu_iovec_destroy(mcb->callbacks[i].free_qiov);
|
|
|
|
}
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(mcb->callbacks[i].free_qiov);
|
2009-09-09 17:53:37 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static void multiwrite_cb(void *opaque, int ret)
|
|
|
|
{
|
|
|
|
MultiwriteCB *mcb = opaque;
|
|
|
|
|
2010-05-22 19:15:08 +02:00
|
|
|
trace_multiwrite_cb(mcb, ret);
|
|
|
|
|
2010-04-01 22:48:44 +02:00
|
|
|
if (ret < 0 && !mcb->error) {
|
2009-09-09 17:53:37 +02:00
|
|
|
mcb->error = ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
mcb->num_requests--;
|
|
|
|
if (mcb->num_requests == 0) {
|
2010-07-01 16:08:51 +02:00
|
|
|
multiwrite_user_cb(mcb);
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(mcb);
|
2009-09-09 17:53:37 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static int multiwrite_req_compare(const void *a, const void *b)
|
|
|
|
{
|
2010-05-19 20:53:10 +02:00
|
|
|
const BlockRequest *req1 = a, *req2 = b;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Note that we can't simply subtract req2->sector from req1->sector
|
|
|
|
* here as that could overflow the return value.
|
|
|
|
*/
|
|
|
|
if (req1->sector > req2->sector) {
|
|
|
|
return 1;
|
|
|
|
} else if (req1->sector < req2->sector) {
|
|
|
|
return -1;
|
|
|
|
} else {
|
|
|
|
return 0;
|
|
|
|
}
|
2009-09-09 17:53:37 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Takes a bunch of requests and tries to merge them. Returns the number of
|
|
|
|
* requests that remain after merging.
|
|
|
|
*/
|
|
|
|
static int multiwrite_merge(BlockDriverState *bs, BlockRequest *reqs,
|
|
|
|
int num_reqs, MultiwriteCB *mcb)
|
|
|
|
{
|
|
|
|
int i, outidx;
|
|
|
|
|
|
|
|
// Sort requests by start sector
|
|
|
|
qsort(reqs, num_reqs, sizeof(*reqs), &multiwrite_req_compare);
|
|
|
|
|
|
|
|
// Check if adjacent requests touch the same clusters. If so, combine them,
|
|
|
|
// filling up gaps with zero sectors.
|
|
|
|
outidx = 0;
|
|
|
|
for (i = 1; i < num_reqs; i++) {
|
|
|
|
int merge = 0;
|
|
|
|
int64_t oldreq_last = reqs[outidx].sector + reqs[outidx].nb_sectors;
|
|
|
|
|
2012-02-21 16:43:52 +01:00
|
|
|
// Handle exactly sequential writes and overlapping writes.
|
2009-09-09 17:53:37 +02:00
|
|
|
if (reqs[i].sector <= oldreq_last) {
|
|
|
|
merge = 1;
|
|
|
|
}
|
|
|
|
|
2010-01-26 14:49:08 +01:00
|
|
|
if (reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1 > IOV_MAX) {
|
|
|
|
merge = 0;
|
|
|
|
}
|
|
|
|
|
2009-09-09 17:53:37 +02:00
|
|
|
if (merge) {
|
|
|
|
size_t size;
|
2011-08-21 05:09:37 +02:00
|
|
|
QEMUIOVector *qiov = g_malloc0(sizeof(*qiov));
|
2009-09-09 17:53:37 +02:00
|
|
|
qemu_iovec_init(qiov,
|
|
|
|
reqs[outidx].qiov->niov + reqs[i].qiov->niov + 1);
|
|
|
|
|
|
|
|
// Add the first request to the merged one. If the requests are
|
|
|
|
// overlapping, drop the last sectors of the first request.
|
|
|
|
size = (reqs[i].sector - reqs[outidx].sector) << 9;
|
consolidate qemu_iovec_copy() and qemu_iovec_concat() and make them consistent
qemu_iovec_concat() is currently a wrapper for
qemu_iovec_copy(), use the former (with extra
"0" arg) in a few places where it is used.
Change skip argument of qemu_iovec_copy() from
uint64_t to size_t, since size of qiov itself
is size_t, so there's no way to skip larger
sizes. Rename it to soffset, to make it clear
that the offset is applied to src.
Also change the only usage of uint64_t in
hw/9pfs/virtio-9p.c, in v9fs_init_qiov_from_pdu() -
all callers of it actually uses size_t too,
not uint64_t.
One added restriction: as for all other iovec-related
functions, soffset must point inside src.
Order of argumens is already good:
qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
int c, size_t bytes)
vs:
qemu_iovec_concat(QEMUIOVector *dst,
QEMUIOVector *src,
size_t soffset, size_t sbytes)
(note soffset is after _src_ not dst, since it applies to src;
for memset it applies to qiov).
Note that in many places where this function is used,
the previous call is qemu_iovec_reset(), which means
many callers actually want copy (replacing dst content),
not concat. So we may want to add a wrapper like
qemu_iovec_copy() with the same arguments but which
calls qemu_iovec_reset() before _concat().
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-03-12 18:28:06 +01:00
|
|
|
qemu_iovec_concat(qiov, reqs[outidx].qiov, 0, size);
|
2009-09-09 17:53:37 +02:00
|
|
|
|
2012-02-21 16:43:52 +01:00
|
|
|
// We should need to add any zeros between the two requests
|
|
|
|
assert (reqs[i].sector <= oldreq_last);
|
2009-09-09 17:53:37 +02:00
|
|
|
|
|
|
|
// Add the second request
|
consolidate qemu_iovec_copy() and qemu_iovec_concat() and make them consistent
qemu_iovec_concat() is currently a wrapper for
qemu_iovec_copy(), use the former (with extra
"0" arg) in a few places where it is used.
Change skip argument of qemu_iovec_copy() from
uint64_t to size_t, since size of qiov itself
is size_t, so there's no way to skip larger
sizes. Rename it to soffset, to make it clear
that the offset is applied to src.
Also change the only usage of uint64_t in
hw/9pfs/virtio-9p.c, in v9fs_init_qiov_from_pdu() -
all callers of it actually uses size_t too,
not uint64_t.
One added restriction: as for all other iovec-related
functions, soffset must point inside src.
Order of argumens is already good:
qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
int c, size_t bytes)
vs:
qemu_iovec_concat(QEMUIOVector *dst,
QEMUIOVector *src,
size_t soffset, size_t sbytes)
(note soffset is after _src_ not dst, since it applies to src;
for memset it applies to qiov).
Note that in many places where this function is used,
the previous call is qemu_iovec_reset(), which means
many callers actually want copy (replacing dst content),
not concat. So we may want to add a wrapper like
qemu_iovec_copy() with the same arguments but which
calls qemu_iovec_reset() before _concat().
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-03-12 18:28:06 +01:00
|
|
|
qemu_iovec_concat(qiov, reqs[i].qiov, 0, reqs[i].qiov->size);
|
2009-09-09 17:53:37 +02:00
|
|
|
|
2010-05-21 11:09:42 +02:00
|
|
|
reqs[outidx].nb_sectors = qiov->size >> 9;
|
2009-09-09 17:53:37 +02:00
|
|
|
reqs[outidx].qiov = qiov;
|
|
|
|
|
|
|
|
mcb->callbacks[i].free_qiov = reqs[outidx].qiov;
|
|
|
|
} else {
|
|
|
|
outidx++;
|
|
|
|
reqs[outidx].sector = reqs[i].sector;
|
|
|
|
reqs[outidx].nb_sectors = reqs[i].nb_sectors;
|
|
|
|
reqs[outidx].qiov = reqs[i].qiov;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return outidx + 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Submit multiple AIO write requests at once.
|
|
|
|
*
|
|
|
|
* On success, the function returns 0 and all requests in the reqs array have
|
|
|
|
* been submitted. In error case this function returns -1, and any of the
|
|
|
|
* requests may or may not be submitted yet. In particular, this means that the
|
|
|
|
* callback will be called for some of the requests, for others it won't. The
|
|
|
|
* caller must check the error field of the BlockRequest to wait for the right
|
|
|
|
* callbacks (if error != 0, no callback will be called).
|
|
|
|
*
|
|
|
|
* The implementation may modify the contents of the reqs array, e.g. to merge
|
|
|
|
* requests. However, the fields opaque and error are left unmodified as they
|
|
|
|
* are used to signal failure for a single request to the caller.
|
|
|
|
*/
|
|
|
|
int bdrv_aio_multiwrite(BlockDriverState *bs, BlockRequest *reqs, int num_reqs)
|
|
|
|
{
|
|
|
|
MultiwriteCB *mcb;
|
|
|
|
int i;
|
|
|
|
|
2011-03-07 17:01:04 +01:00
|
|
|
/* don't submit writes if we don't have a medium */
|
|
|
|
if (bs->drv == NULL) {
|
|
|
|
for (i = 0; i < num_reqs; i++) {
|
|
|
|
reqs[i].error = -ENOMEDIUM;
|
|
|
|
}
|
|
|
|
return -1;
|
|
|
|
}
|
|
|
|
|
2009-09-09 17:53:37 +02:00
|
|
|
if (num_reqs == 0) {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Create MultiwriteCB structure
|
2011-08-21 05:09:37 +02:00
|
|
|
mcb = g_malloc0(sizeof(*mcb) + num_reqs * sizeof(*mcb->callbacks));
|
2009-09-09 17:53:37 +02:00
|
|
|
mcb->num_requests = 0;
|
|
|
|
mcb->num_callbacks = num_reqs;
|
|
|
|
|
|
|
|
for (i = 0; i < num_reqs; i++) {
|
|
|
|
mcb->callbacks[i].cb = reqs[i].cb;
|
|
|
|
mcb->callbacks[i].opaque = reqs[i].opaque;
|
|
|
|
}
|
|
|
|
|
|
|
|
// Check for mergable requests
|
|
|
|
num_reqs = multiwrite_merge(bs, reqs, num_reqs, mcb);
|
|
|
|
|
2010-05-22 19:15:08 +02:00
|
|
|
trace_bdrv_aio_multiwrite(mcb, mcb->num_callbacks, num_reqs);
|
|
|
|
|
2011-11-14 17:50:50 +01:00
|
|
|
/* Run the aio requests. */
|
|
|
|
mcb->num_requests = num_reqs;
|
2009-09-09 17:53:37 +02:00
|
|
|
for (i = 0; i < num_reqs; i++) {
|
2011-11-30 09:12:30 +01:00
|
|
|
bdrv_aio_writev(bs, reqs[i].sector, reqs[i].qiov,
|
2009-09-09 17:53:37 +02:00
|
|
|
reqs[i].nb_sectors, multiwrite_cb, mcb);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
void bdrv_aio_cancel(BlockDriverAIOCB *acb)
|
|
|
|
{
|
2012-10-31 16:34:37 +01:00
|
|
|
acb->aiocb_info->cancel(acb);
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
|
|
|
|
2011-11-08 06:00:14 +01:00
|
|
|
/* block I/O throttling */
|
|
|
|
static bool bdrv_exceed_bps_limits(BlockDriverState *bs, int nb_sectors,
|
|
|
|
bool is_write, double elapsed_time, uint64_t *wait)
|
|
|
|
{
|
|
|
|
uint64_t bps_limit = 0;
|
|
|
|
double bytes_limit, bytes_base, bytes_res;
|
|
|
|
double slice_time, wait_time;
|
|
|
|
|
|
|
|
if (bs->io_limits.bps[BLOCK_IO_LIMIT_TOTAL]) {
|
|
|
|
bps_limit = bs->io_limits.bps[BLOCK_IO_LIMIT_TOTAL];
|
|
|
|
} else if (bs->io_limits.bps[is_write]) {
|
|
|
|
bps_limit = bs->io_limits.bps[is_write];
|
|
|
|
} else {
|
|
|
|
if (wait) {
|
|
|
|
*wait = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
slice_time = bs->slice_end - bs->slice_start;
|
|
|
|
slice_time /= (NANOSECONDS_PER_SECOND);
|
|
|
|
bytes_limit = bps_limit * slice_time;
|
|
|
|
bytes_base = bs->nr_bytes[is_write] - bs->io_base.bytes[is_write];
|
|
|
|
if (bs->io_limits.bps[BLOCK_IO_LIMIT_TOTAL]) {
|
|
|
|
bytes_base += bs->nr_bytes[!is_write] - bs->io_base.bytes[!is_write];
|
|
|
|
}
|
|
|
|
|
|
|
|
/* bytes_base: the bytes of data which have been read/written; and
|
|
|
|
* it is obtained from the history statistic info.
|
|
|
|
* bytes_res: the remaining bytes of data which need to be read/written.
|
|
|
|
* (bytes_base + bytes_res) / bps_limit: used to calcuate
|
|
|
|
* the total time for completing reading/writting all data.
|
|
|
|
*/
|
|
|
|
bytes_res = (unsigned) nb_sectors * BDRV_SECTOR_SIZE;
|
|
|
|
|
|
|
|
if (bytes_base + bytes_res <= bytes_limit) {
|
|
|
|
if (wait) {
|
|
|
|
*wait = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Calc approx time to dispatch */
|
|
|
|
wait_time = (bytes_base + bytes_res) / bps_limit - elapsed_time;
|
|
|
|
|
|
|
|
/* When the I/O rate at runtime exceeds the limits,
|
|
|
|
* bs->slice_end need to be extended in order that the current statistic
|
|
|
|
* info can be kept until the timer fire, so it is increased and tuned
|
|
|
|
* based on the result of experiment.
|
|
|
|
*/
|
|
|
|
bs->slice_time = wait_time * BLOCK_IO_SLICE_TIME * 10;
|
|
|
|
bs->slice_end += bs->slice_time - 3 * BLOCK_IO_SLICE_TIME;
|
|
|
|
if (wait) {
|
|
|
|
*wait = wait_time * BLOCK_IO_SLICE_TIME * 10;
|
|
|
|
}
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool bdrv_exceed_iops_limits(BlockDriverState *bs, bool is_write,
|
|
|
|
double elapsed_time, uint64_t *wait)
|
|
|
|
{
|
|
|
|
uint64_t iops_limit = 0;
|
|
|
|
double ios_limit, ios_base;
|
|
|
|
double slice_time, wait_time;
|
|
|
|
|
|
|
|
if (bs->io_limits.iops[BLOCK_IO_LIMIT_TOTAL]) {
|
|
|
|
iops_limit = bs->io_limits.iops[BLOCK_IO_LIMIT_TOTAL];
|
|
|
|
} else if (bs->io_limits.iops[is_write]) {
|
|
|
|
iops_limit = bs->io_limits.iops[is_write];
|
|
|
|
} else {
|
|
|
|
if (wait) {
|
|
|
|
*wait = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
slice_time = bs->slice_end - bs->slice_start;
|
|
|
|
slice_time /= (NANOSECONDS_PER_SECOND);
|
|
|
|
ios_limit = iops_limit * slice_time;
|
|
|
|
ios_base = bs->nr_ops[is_write] - bs->io_base.ios[is_write];
|
|
|
|
if (bs->io_limits.iops[BLOCK_IO_LIMIT_TOTAL]) {
|
|
|
|
ios_base += bs->nr_ops[!is_write] - bs->io_base.ios[!is_write];
|
|
|
|
}
|
|
|
|
|
|
|
|
if (ios_base + 1 <= ios_limit) {
|
|
|
|
if (wait) {
|
|
|
|
*wait = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Calc approx time to dispatch */
|
|
|
|
wait_time = (ios_base + 1) / iops_limit;
|
|
|
|
if (wait_time > elapsed_time) {
|
|
|
|
wait_time = wait_time - elapsed_time;
|
|
|
|
} else {
|
|
|
|
wait_time = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
bs->slice_time = wait_time * BLOCK_IO_SLICE_TIME * 10;
|
|
|
|
bs->slice_end += bs->slice_time - 3 * BLOCK_IO_SLICE_TIME;
|
|
|
|
if (wait) {
|
|
|
|
*wait = wait_time * BLOCK_IO_SLICE_TIME * 10;
|
|
|
|
}
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
static bool bdrv_exceed_io_limits(BlockDriverState *bs, int nb_sectors,
|
|
|
|
bool is_write, int64_t *wait)
|
|
|
|
{
|
|
|
|
int64_t now, max_wait;
|
|
|
|
uint64_t bps_wait = 0, iops_wait = 0;
|
|
|
|
double elapsed_time;
|
|
|
|
int bps_ret, iops_ret;
|
|
|
|
|
|
|
|
now = qemu_get_clock_ns(vm_clock);
|
|
|
|
if ((bs->slice_start < now)
|
|
|
|
&& (bs->slice_end > now)) {
|
|
|
|
bs->slice_end = now + bs->slice_time;
|
|
|
|
} else {
|
|
|
|
bs->slice_time = 5 * BLOCK_IO_SLICE_TIME;
|
|
|
|
bs->slice_start = now;
|
|
|
|
bs->slice_end = now + bs->slice_time;
|
|
|
|
|
|
|
|
bs->io_base.bytes[is_write] = bs->nr_bytes[is_write];
|
|
|
|
bs->io_base.bytes[!is_write] = bs->nr_bytes[!is_write];
|
|
|
|
|
|
|
|
bs->io_base.ios[is_write] = bs->nr_ops[is_write];
|
|
|
|
bs->io_base.ios[!is_write] = bs->nr_ops[!is_write];
|
|
|
|
}
|
|
|
|
|
|
|
|
elapsed_time = now - bs->slice_start;
|
|
|
|
elapsed_time /= (NANOSECONDS_PER_SECOND);
|
|
|
|
|
|
|
|
bps_ret = bdrv_exceed_bps_limits(bs, nb_sectors,
|
|
|
|
is_write, elapsed_time, &bps_wait);
|
|
|
|
iops_ret = bdrv_exceed_iops_limits(bs, is_write,
|
|
|
|
elapsed_time, &iops_wait);
|
|
|
|
if (bps_ret || iops_ret) {
|
|
|
|
max_wait = bps_wait > iops_wait ? bps_wait : iops_wait;
|
|
|
|
if (wait) {
|
|
|
|
*wait = max_wait;
|
|
|
|
}
|
|
|
|
|
|
|
|
now = qemu_get_clock_ns(vm_clock);
|
|
|
|
if (bs->slice_end < now + max_wait) {
|
|
|
|
bs->slice_end = now + max_wait;
|
|
|
|
}
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (wait) {
|
|
|
|
*wait = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
return false;
|
|
|
|
}
|
2006-08-07 04:38:06 +02:00
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
/**************************************************************/
|
|
|
|
/* async block device emulation */
|
|
|
|
|
2009-05-25 12:37:32 +02:00
|
|
|
typedef struct BlockDriverAIOCBSync {
|
|
|
|
BlockDriverAIOCB common;
|
|
|
|
QEMUBH *bh;
|
|
|
|
int ret;
|
|
|
|
/* vector translation state */
|
|
|
|
QEMUIOVector *qiov;
|
|
|
|
uint8_t *bounce;
|
|
|
|
int is_write;
|
|
|
|
} BlockDriverAIOCBSync;
|
|
|
|
|
|
|
|
static void bdrv_aio_cancel_em(BlockDriverAIOCB *blockacb)
|
|
|
|
{
|
2010-05-05 11:44:39 +02:00
|
|
|
BlockDriverAIOCBSync *acb =
|
|
|
|
container_of(blockacb, BlockDriverAIOCBSync, common);
|
2009-06-01 11:07:23 +02:00
|
|
|
qemu_bh_delete(acb->bh);
|
2009-06-23 15:20:36 +02:00
|
|
|
acb->bh = NULL;
|
2009-05-25 12:37:32 +02:00
|
|
|
qemu_aio_release(acb);
|
|
|
|
}
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
static const AIOCBInfo bdrv_em_aiocb_info = {
|
2009-05-25 12:37:32 +02:00
|
|
|
.aiocb_size = sizeof(BlockDriverAIOCBSync),
|
|
|
|
.cancel = bdrv_aio_cancel_em,
|
|
|
|
};
|
|
|
|
|
2006-08-07 04:38:06 +02:00
|
|
|
static void bdrv_aio_bh_cb(void *opaque)
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
2006-08-07 04:38:06 +02:00
|
|
|
BlockDriverAIOCBSync *acb = opaque;
|
2009-04-07 20:43:24 +02:00
|
|
|
|
|
|
|
if (!acb->is_write)
|
allow qemu_iovec_from_buffer() to specify offset from which to start copying
Similar to
qemu_iovec_memset(QEMUIOVector *qiov, size_t offset,
int c, size_t bytes);
the new prototype is:
qemu_iovec_from_buf(QEMUIOVector *qiov, size_t offset,
const void *buf, size_t bytes);
The processing starts at offset bytes within qiov.
This way, we may copy a bounce buffer directly to
a middle of qiov.
This is exactly the same function as iov_from_buf() from
iov.c, so use the existing implementation and rename it
to qemu_iovec_from_buf() to be shorter and to match the
utility function.
As with utility implementation, we now assert that the
offset is inside actual iovec. Nothing changed for
current callers, because `offset' parameter is new.
While at it, stop using "bounce-qiov" in block/qcow2.c
and copy decrypted data directly from cluster_data
instead of recreating a temp qiov for doing that.
Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>
2012-06-07 18:17:55 +02:00
|
|
|
qemu_iovec_from_buf(acb->qiov, 0, acb->bounce, acb->qiov->size);
|
2009-04-07 20:43:28 +02:00
|
|
|
qemu_vfree(acb->bounce);
|
2006-08-07 04:38:06 +02:00
|
|
|
acb->common.cb(acb->common.opaque, acb->ret);
|
2009-06-01 11:07:23 +02:00
|
|
|
qemu_bh_delete(acb->bh);
|
2009-06-23 15:20:36 +02:00
|
|
|
acb->bh = NULL;
|
2006-08-07 04:38:06 +02:00
|
|
|
qemu_aio_release(acb);
|
2006-08-01 18:21:11 +02:00
|
|
|
}
|
2006-06-26 22:08:57 +02:00
|
|
|
|
2009-04-07 20:43:24 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_aio_rw_vector(BlockDriverState *bs,
|
|
|
|
int64_t sector_num,
|
|
|
|
QEMUIOVector *qiov,
|
|
|
|
int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb,
|
|
|
|
void *opaque,
|
|
|
|
int is_write)
|
|
|
|
|
2006-08-01 18:21:11 +02:00
|
|
|
{
|
2006-08-07 04:38:06 +02:00
|
|
|
BlockDriverAIOCBSync *acb;
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
acb = qemu_aio_get(&bdrv_em_aiocb_info, bs, cb, opaque);
|
2009-04-07 20:43:24 +02:00
|
|
|
acb->is_write = is_write;
|
|
|
|
acb->qiov = qiov;
|
2009-04-22 22:20:00 +02:00
|
|
|
acb->bounce = qemu_blockalign(bs, qiov->size);
|
2011-11-14 17:50:54 +01:00
|
|
|
acb->bh = qemu_bh_new(bdrv_aio_bh_cb, acb);
|
2009-04-07 20:43:24 +02:00
|
|
|
|
|
|
|
if (is_write) {
|
2012-06-07 18:21:06 +02:00
|
|
|
qemu_iovec_to_buf(acb->qiov, 0, acb->bounce, qiov->size);
|
2011-10-13 14:08:21 +02:00
|
|
|
acb->ret = bs->drv->bdrv_write(bs, sector_num, acb->bounce, nb_sectors);
|
2009-04-07 20:43:24 +02:00
|
|
|
} else {
|
2011-10-13 14:08:21 +02:00
|
|
|
acb->ret = bs->drv->bdrv_read(bs, sector_num, acb->bounce, nb_sectors);
|
2009-04-07 20:43:24 +02:00
|
|
|
}
|
|
|
|
|
2006-08-07 04:38:06 +02:00
|
|
|
qemu_bh_schedule(acb->bh);
|
2009-04-07 20:43:24 +02:00
|
|
|
|
2006-08-07 04:38:06 +02:00
|
|
|
return &acb->common;
|
2006-06-26 22:08:57 +02:00
|
|
|
}
|
|
|
|
|
2009-04-07 20:43:24 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_aio_readv_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
|
2006-08-07 04:38:06 +02:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
2006-06-26 22:08:57 +02:00
|
|
|
{
|
2009-04-07 20:43:24 +02:00
|
|
|
return bdrv_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 0);
|
|
|
|
}
|
2006-08-01 18:21:11 +02:00
|
|
|
|
2009-04-07 20:43:24 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_aio_writev_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, QEMUIOVector *qiov, int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
|
|
|
{
|
|
|
|
return bdrv_aio_rw_vector(bs, sector_num, qiov, nb_sectors, cb, opaque, 1);
|
2006-06-26 22:08:57 +02:00
|
|
|
}
|
|
|
|
|
2011-06-30 10:05:46 +02:00
|
|
|
|
|
|
|
typedef struct BlockDriverAIOCBCoroutine {
|
|
|
|
BlockDriverAIOCB common;
|
|
|
|
BlockRequest req;
|
|
|
|
bool is_write;
|
2012-11-13 16:35:08 +01:00
|
|
|
bool *done;
|
2011-06-30 10:05:46 +02:00
|
|
|
QEMUBH* bh;
|
|
|
|
} BlockDriverAIOCBCoroutine;
|
|
|
|
|
|
|
|
static void bdrv_aio_co_cancel_em(BlockDriverAIOCB *blockacb)
|
|
|
|
{
|
2012-11-13 16:35:08 +01:00
|
|
|
BlockDriverAIOCBCoroutine *acb =
|
|
|
|
container_of(blockacb, BlockDriverAIOCBCoroutine, common);
|
|
|
|
bool done = false;
|
|
|
|
|
|
|
|
acb->done = &done;
|
|
|
|
while (!done) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
2011-06-30 10:05:46 +02:00
|
|
|
}
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
static const AIOCBInfo bdrv_em_co_aiocb_info = {
|
2011-06-30 10:05:46 +02:00
|
|
|
.aiocb_size = sizeof(BlockDriverAIOCBCoroutine),
|
|
|
|
.cancel = bdrv_aio_co_cancel_em,
|
|
|
|
};
|
|
|
|
|
2011-10-14 10:41:29 +02:00
|
|
|
static void bdrv_co_em_bh(void *opaque)
|
2011-06-30 10:05:46 +02:00
|
|
|
{
|
|
|
|
BlockDriverAIOCBCoroutine *acb = opaque;
|
|
|
|
|
|
|
|
acb->common.cb(acb->common.opaque, acb->req.error);
|
2012-11-13 16:35:08 +01:00
|
|
|
|
|
|
|
if (acb->done) {
|
|
|
|
*acb->done = true;
|
|
|
|
}
|
|
|
|
|
2011-06-30 10:05:46 +02:00
|
|
|
qemu_bh_delete(acb->bh);
|
|
|
|
qemu_aio_release(acb);
|
|
|
|
}
|
|
|
|
|
2011-10-13 14:08:23 +02:00
|
|
|
/* Invoke bdrv_co_do_readv/bdrv_co_do_writev */
|
|
|
|
static void coroutine_fn bdrv_co_do_rw(void *opaque)
|
|
|
|
{
|
|
|
|
BlockDriverAIOCBCoroutine *acb = opaque;
|
|
|
|
BlockDriverState *bs = acb->common.bs;
|
|
|
|
|
|
|
|
if (!acb->is_write) {
|
|
|
|
acb->req.error = bdrv_co_do_readv(bs, acb->req.sector,
|
2012-01-18 15:40:42 +01:00
|
|
|
acb->req.nb_sectors, acb->req.qiov, 0);
|
2011-10-13 14:08:23 +02:00
|
|
|
} else {
|
|
|
|
acb->req.error = bdrv_co_do_writev(bs, acb->req.sector,
|
2012-02-07 14:27:25 +01:00
|
|
|
acb->req.nb_sectors, acb->req.qiov, 0);
|
2011-10-13 14:08:23 +02:00
|
|
|
}
|
|
|
|
|
2011-10-14 10:41:29 +02:00
|
|
|
acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
|
2011-10-13 14:08:23 +02:00
|
|
|
qemu_bh_schedule(acb->bh);
|
|
|
|
}
|
|
|
|
|
2011-06-30 10:05:46 +02:00
|
|
|
static BlockDriverAIOCB *bdrv_co_aio_rw_vector(BlockDriverState *bs,
|
|
|
|
int64_t sector_num,
|
|
|
|
QEMUIOVector *qiov,
|
|
|
|
int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb,
|
|
|
|
void *opaque,
|
2011-10-13 22:09:28 +02:00
|
|
|
bool is_write)
|
2011-06-30 10:05:46 +02:00
|
|
|
{
|
|
|
|
Coroutine *co;
|
|
|
|
BlockDriverAIOCBCoroutine *acb;
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
acb = qemu_aio_get(&bdrv_em_co_aiocb_info, bs, cb, opaque);
|
2011-06-30 10:05:46 +02:00
|
|
|
acb->req.sector = sector_num;
|
|
|
|
acb->req.nb_sectors = nb_sectors;
|
|
|
|
acb->req.qiov = qiov;
|
|
|
|
acb->is_write = is_write;
|
2012-11-13 16:35:08 +01:00
|
|
|
acb->done = NULL;
|
2011-06-30 10:05:46 +02:00
|
|
|
|
2011-10-13 22:09:28 +02:00
|
|
|
co = qemu_coroutine_create(bdrv_co_do_rw);
|
2011-06-30 10:05:46 +02:00
|
|
|
qemu_coroutine_enter(co, acb);
|
|
|
|
|
|
|
|
return &acb->common;
|
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
static void coroutine_fn bdrv_aio_flush_co_entry(void *opaque)
|
2009-09-04 19:01:49 +02:00
|
|
|
{
|
2011-10-17 12:32:12 +02:00
|
|
|
BlockDriverAIOCBCoroutine *acb = opaque;
|
|
|
|
BlockDriverState *bs = acb->common.bs;
|
2009-09-04 19:01:49 +02:00
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
acb->req.error = bdrv_co_flush(bs);
|
|
|
|
acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
|
2009-09-04 19:01:49 +02:00
|
|
|
qemu_bh_schedule(acb->bh);
|
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
BlockDriverAIOCB *bdrv_aio_flush(BlockDriverState *bs,
|
2010-05-26 17:51:49 +02:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
|
|
|
{
|
2011-10-17 12:32:12 +02:00
|
|
|
trace_bdrv_aio_flush(bs, opaque);
|
2010-05-26 17:51:49 +02:00
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
Coroutine *co;
|
|
|
|
BlockDriverAIOCBCoroutine *acb;
|
2010-05-26 17:51:49 +02:00
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
acb = qemu_aio_get(&bdrv_em_co_aiocb_info, bs, cb, opaque);
|
2012-11-13 16:35:08 +01:00
|
|
|
acb->done = NULL;
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
co = qemu_coroutine_create(bdrv_aio_flush_co_entry);
|
|
|
|
qemu_coroutine_enter(co, acb);
|
2010-05-26 17:51:49 +02:00
|
|
|
|
|
|
|
return &acb->common;
|
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:14 +02:00
|
|
|
static void coroutine_fn bdrv_aio_discard_co_entry(void *opaque)
|
|
|
|
{
|
|
|
|
BlockDriverAIOCBCoroutine *acb = opaque;
|
|
|
|
BlockDriverState *bs = acb->common.bs;
|
|
|
|
|
|
|
|
acb->req.error = bdrv_co_discard(bs, acb->req.sector, acb->req.nb_sectors);
|
|
|
|
acb->bh = qemu_bh_new(bdrv_co_em_bh, acb);
|
|
|
|
qemu_bh_schedule(acb->bh);
|
|
|
|
}
|
|
|
|
|
|
|
|
BlockDriverAIOCB *bdrv_aio_discard(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
|
|
|
{
|
|
|
|
Coroutine *co;
|
|
|
|
BlockDriverAIOCBCoroutine *acb;
|
|
|
|
|
|
|
|
trace_bdrv_aio_discard(bs, sector_num, nb_sectors, opaque);
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
acb = qemu_aio_get(&bdrv_em_co_aiocb_info, bs, cb, opaque);
|
2011-10-17 12:32:14 +02:00
|
|
|
acb->req.sector = sector_num;
|
|
|
|
acb->req.nb_sectors = nb_sectors;
|
2012-11-13 16:35:08 +01:00
|
|
|
acb->done = NULL;
|
2011-10-17 12:32:14 +02:00
|
|
|
co = qemu_coroutine_create(bdrv_aio_discard_co_entry);
|
|
|
|
qemu_coroutine_enter(co, acb);
|
|
|
|
|
|
|
|
return &acb->common;
|
|
|
|
}
|
|
|
|
|
2004-08-01 23:59:26 +02:00
|
|
|
void bdrv_init(void)
|
|
|
|
{
|
2009-05-10 00:03:42 +02:00
|
|
|
module_call_init(MODULE_INIT_BLOCK);
|
2004-08-01 23:59:26 +02:00
|
|
|
}
|
2006-08-07 04:38:06 +02:00
|
|
|
|
2009-10-27 18:41:44 +01:00
|
|
|
void bdrv_init_with_whitelist(void)
|
|
|
|
{
|
|
|
|
use_bdrv_whitelist = 1;
|
|
|
|
bdrv_init();
|
|
|
|
}
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
void *qemu_aio_get(const AIOCBInfo *aiocb_info, BlockDriverState *bs,
|
2009-05-25 12:37:32 +02:00
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
2006-08-07 04:38:06 +02:00
|
|
|
{
|
|
|
|
BlockDriverAIOCB *acb;
|
|
|
|
|
2012-10-31 16:34:37 +01:00
|
|
|
acb = g_slice_alloc(aiocb_info->aiocb_size);
|
|
|
|
acb->aiocb_info = aiocb_info;
|
2006-08-07 04:38:06 +02:00
|
|
|
acb->bs = bs;
|
|
|
|
acb->cb = cb;
|
|
|
|
acb->opaque = opaque;
|
|
|
|
return acb;
|
|
|
|
}
|
|
|
|
|
|
|
|
void qemu_aio_release(void *p)
|
|
|
|
{
|
2012-10-31 16:34:36 +01:00
|
|
|
BlockDriverAIOCB *acb = p;
|
2012-10-31 16:34:37 +01:00
|
|
|
g_slice_free1(acb->aiocb_info->aiocb_size, acb);
|
2006-08-07 04:38:06 +02:00
|
|
|
}
|
2006-08-19 13:45:59 +02:00
|
|
|
|
2011-07-15 13:50:26 +02:00
|
|
|
/**************************************************************/
|
|
|
|
/* Coroutine block device emulation */
|
|
|
|
|
|
|
|
typedef struct CoroutineIOCompletion {
|
|
|
|
Coroutine *coroutine;
|
|
|
|
int ret;
|
|
|
|
} CoroutineIOCompletion;
|
|
|
|
|
|
|
|
static void bdrv_co_io_em_complete(void *opaque, int ret)
|
|
|
|
{
|
|
|
|
CoroutineIOCompletion *co = opaque;
|
|
|
|
|
|
|
|
co->ret = ret;
|
|
|
|
qemu_coroutine_enter(co->coroutine, NULL);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int coroutine_fn bdrv_co_io_em(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
int nb_sectors, QEMUIOVector *iov,
|
|
|
|
bool is_write)
|
|
|
|
{
|
|
|
|
CoroutineIOCompletion co = {
|
|
|
|
.coroutine = qemu_coroutine_self(),
|
|
|
|
};
|
|
|
|
BlockDriverAIOCB *acb;
|
|
|
|
|
|
|
|
if (is_write) {
|
2011-10-05 18:17:02 +02:00
|
|
|
acb = bs->drv->bdrv_aio_writev(bs, sector_num, iov, nb_sectors,
|
|
|
|
bdrv_co_io_em_complete, &co);
|
2011-07-15 13:50:26 +02:00
|
|
|
} else {
|
2011-10-05 18:17:02 +02:00
|
|
|
acb = bs->drv->bdrv_aio_readv(bs, sector_num, iov, nb_sectors,
|
|
|
|
bdrv_co_io_em_complete, &co);
|
2011-07-15 13:50:26 +02:00
|
|
|
}
|
|
|
|
|
2011-09-30 18:34:58 +02:00
|
|
|
trace_bdrv_co_io_em(bs, sector_num, nb_sectors, is_write, acb);
|
2011-07-15 13:50:26 +02:00
|
|
|
if (!acb) {
|
|
|
|
return -EIO;
|
|
|
|
}
|
|
|
|
qemu_coroutine_yield();
|
|
|
|
|
|
|
|
return co.ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int coroutine_fn bdrv_co_readv_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
QEMUIOVector *iov)
|
|
|
|
{
|
|
|
|
return bdrv_co_io_em(bs, sector_num, nb_sectors, iov, false);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int coroutine_fn bdrv_co_writev_em(BlockDriverState *bs,
|
|
|
|
int64_t sector_num, int nb_sectors,
|
|
|
|
QEMUIOVector *iov)
|
|
|
|
{
|
|
|
|
return bdrv_co_io_em(bs, sector_num, nb_sectors, iov, true);
|
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
static void coroutine_fn bdrv_flush_co_entry(void *opaque)
|
2011-07-15 16:05:00 +02:00
|
|
|
{
|
2011-10-17 12:32:12 +02:00
|
|
|
RwCo *rwco = opaque;
|
|
|
|
|
|
|
|
rwco->ret = bdrv_co_flush(rwco->bs);
|
|
|
|
}
|
|
|
|
|
|
|
|
int coroutine_fn bdrv_co_flush(BlockDriverState *bs)
|
|
|
|
{
|
2011-11-10 18:10:11 +01:00
|
|
|
int ret;
|
|
|
|
|
2012-03-12 18:26:01 +01:00
|
|
|
if (!bs || !bdrv_is_inserted(bs) || bdrv_is_read_only(bs)) {
|
2011-10-17 12:32:12 +02:00
|
|
|
return 0;
|
2011-11-10 18:10:11 +01:00
|
|
|
}
|
|
|
|
|
2011-11-10 18:13:59 +01:00
|
|
|
/* Write back cached data to the OS even with cache=unsafe */
|
2011-11-10 18:10:11 +01:00
|
|
|
if (bs->drv->bdrv_co_flush_to_os) {
|
|
|
|
ret = bs->drv->bdrv_co_flush_to_os(bs);
|
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-11-10 18:13:59 +01:00
|
|
|
/* But don't actually force it to the disk with cache=unsafe */
|
|
|
|
if (bs->open_flags & BDRV_O_NO_FLUSH) {
|
2012-08-15 12:52:45 +02:00
|
|
|
goto flush_parent;
|
2011-11-10 18:13:59 +01:00
|
|
|
}
|
|
|
|
|
2011-11-10 18:10:11 +01:00
|
|
|
if (bs->drv->bdrv_co_flush_to_disk) {
|
2012-03-12 18:26:01 +01:00
|
|
|
ret = bs->drv->bdrv_co_flush_to_disk(bs);
|
2011-10-17 12:32:12 +02:00
|
|
|
} else if (bs->drv->bdrv_aio_flush) {
|
|
|
|
BlockDriverAIOCB *acb;
|
|
|
|
CoroutineIOCompletion co = {
|
|
|
|
.coroutine = qemu_coroutine_self(),
|
|
|
|
};
|
|
|
|
|
|
|
|
acb = bs->drv->bdrv_aio_flush(bs, bdrv_co_io_em_complete, &co);
|
|
|
|
if (acb == NULL) {
|
2012-03-12 18:26:01 +01:00
|
|
|
ret = -EIO;
|
2011-10-17 12:32:12 +02:00
|
|
|
} else {
|
|
|
|
qemu_coroutine_yield();
|
2012-03-12 18:26:01 +01:00
|
|
|
ret = co.ret;
|
2011-10-17 12:32:12 +02:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
/*
|
|
|
|
* Some block drivers always operate in either writethrough or unsafe
|
|
|
|
* mode and don't support bdrv_flush therefore. Usually qemu doesn't
|
|
|
|
* know how the server works (because the behaviour is hardcoded or
|
|
|
|
* depends on server-side configuration), so we can't ensure that
|
|
|
|
* everything is safe on disk. Returning an error doesn't work because
|
|
|
|
* that would break guests even if the server operates in writethrough
|
|
|
|
* mode.
|
|
|
|
*
|
|
|
|
* Let's hope the user knows what he's doing.
|
|
|
|
*/
|
2012-03-12 18:26:01 +01:00
|
|
|
ret = 0;
|
2011-10-17 12:32:12 +02:00
|
|
|
}
|
2012-03-12 18:26:01 +01:00
|
|
|
if (ret < 0) {
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Now flush the underlying protocol. It will also have BDRV_O_NO_FLUSH
|
|
|
|
* in the case of cache=unsafe, so there are no useless flushes.
|
|
|
|
*/
|
2012-08-15 12:52:45 +02:00
|
|
|
flush_parent:
|
2012-03-12 18:26:01 +01:00
|
|
|
return bdrv_co_flush(bs->file);
|
2011-10-17 12:32:12 +02:00
|
|
|
}
|
|
|
|
|
2011-11-14 22:09:45 +01:00
|
|
|
void bdrv_invalidate_cache(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bs->drv && bs->drv->bdrv_invalidate_cache) {
|
|
|
|
bs->drv->bdrv_invalidate_cache(bs);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_invalidate_cache_all(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
bdrv_invalidate_cache(bs);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-03-23 08:36:49 +01:00
|
|
|
void bdrv_clear_incoming_migration_all(void)
|
|
|
|
{
|
|
|
|
BlockDriverState *bs;
|
|
|
|
|
|
|
|
QTAILQ_FOREACH(bs, &bdrv_states, list) {
|
|
|
|
bs->open_flags = bs->open_flags & ~(BDRV_O_INCOMING);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
int bdrv_flush(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
Coroutine *co;
|
|
|
|
RwCo rwco = {
|
|
|
|
.bs = bs,
|
|
|
|
.ret = NOT_DONE,
|
2011-07-15 16:05:00 +02:00
|
|
|
};
|
|
|
|
|
2011-10-17 12:32:12 +02:00
|
|
|
if (qemu_in_coroutine()) {
|
|
|
|
/* Fast-path if already in coroutine context */
|
|
|
|
bdrv_flush_co_entry(&rwco);
|
|
|
|
} else {
|
|
|
|
co = qemu_coroutine_create(bdrv_flush_co_entry);
|
|
|
|
qemu_coroutine_enter(co, &rwco);
|
|
|
|
while (rwco.ret == NOT_DONE) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
2011-07-15 16:05:00 +02:00
|
|
|
}
|
2011-10-17 12:32:12 +02:00
|
|
|
|
|
|
|
return rwco.ret;
|
2011-07-15 16:05:00 +02:00
|
|
|
}
|
|
|
|
|
2011-10-17 12:32:14 +02:00
|
|
|
static void coroutine_fn bdrv_discard_co_entry(void *opaque)
|
|
|
|
{
|
|
|
|
RwCo *rwco = opaque;
|
|
|
|
|
|
|
|
rwco->ret = bdrv_co_discard(rwco->bs, rwco->sector_num, rwco->nb_sectors);
|
|
|
|
}
|
|
|
|
|
|
|
|
int coroutine_fn bdrv_co_discard(BlockDriverState *bs, int64_t sector_num,
|
|
|
|
int nb_sectors)
|
|
|
|
{
|
|
|
|
if (!bs->drv) {
|
|
|
|
return -ENOMEDIUM;
|
|
|
|
} else if (bdrv_check_request(bs, sector_num, nb_sectors)) {
|
|
|
|
return -EIO;
|
|
|
|
} else if (bs->read_only) {
|
|
|
|
return -EROFS;
|
|
|
|
} else if (bs->drv->bdrv_co_discard) {
|
|
|
|
return bs->drv->bdrv_co_discard(bs, sector_num, nb_sectors);
|
|
|
|
} else if (bs->drv->bdrv_aio_discard) {
|
|
|
|
BlockDriverAIOCB *acb;
|
|
|
|
CoroutineIOCompletion co = {
|
|
|
|
.coroutine = qemu_coroutine_self(),
|
|
|
|
};
|
|
|
|
|
|
|
|
acb = bs->drv->bdrv_aio_discard(bs, sector_num, nb_sectors,
|
|
|
|
bdrv_co_io_em_complete, &co);
|
|
|
|
if (acb == NULL) {
|
|
|
|
return -EIO;
|
|
|
|
} else {
|
|
|
|
qemu_coroutine_yield();
|
|
|
|
return co.ret;
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_discard(BlockDriverState *bs, int64_t sector_num, int nb_sectors)
|
|
|
|
{
|
|
|
|
Coroutine *co;
|
|
|
|
RwCo rwco = {
|
|
|
|
.bs = bs,
|
|
|
|
.sector_num = sector_num,
|
|
|
|
.nb_sectors = nb_sectors,
|
|
|
|
.ret = NOT_DONE,
|
|
|
|
};
|
|
|
|
|
|
|
|
if (qemu_in_coroutine()) {
|
|
|
|
/* Fast-path if already in coroutine context */
|
|
|
|
bdrv_discard_co_entry(&rwco);
|
|
|
|
} else {
|
|
|
|
co = qemu_coroutine_create(bdrv_discard_co_entry);
|
|
|
|
qemu_coroutine_enter(co, &rwco);
|
|
|
|
while (rwco.ret == NOT_DONE) {
|
|
|
|
qemu_aio_wait();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return rwco.ret;
|
|
|
|
}
|
|
|
|
|
2006-08-19 13:45:59 +02:00
|
|
|
/**************************************************************/
|
|
|
|
/* removable device support */
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Return TRUE if the media is present
|
|
|
|
*/
|
|
|
|
int bdrv_is_inserted(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
2011-09-06 18:58:41 +02:00
|
|
|
|
2006-08-19 13:45:59 +02:00
|
|
|
if (!drv)
|
|
|
|
return 0;
|
|
|
|
if (!drv->bdrv_is_inserted)
|
2011-09-06 18:58:41 +02:00
|
|
|
return 1;
|
|
|
|
return drv->bdrv_is_inserted(bs);
|
2006-08-19 13:45:59 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
2011-08-03 15:08:08 +02:00
|
|
|
* Return whether the media changed since the last call to this
|
|
|
|
* function, or -ENOTSUP if we don't know. Most drivers don't know.
|
2006-08-19 13:45:59 +02:00
|
|
|
*/
|
|
|
|
int bdrv_media_changed(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
2011-08-03 15:08:08 +02:00
|
|
|
if (drv && drv->bdrv_media_changed) {
|
|
|
|
return drv->bdrv_media_changed(bs);
|
|
|
|
}
|
|
|
|
return -ENOTSUP;
|
2006-08-19 13:45:59 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* If eject_flag is TRUE, eject the media. Otherwise, close the tray
|
|
|
|
*/
|
2012-02-03 19:24:53 +01:00
|
|
|
void bdrv_eject(BlockDriverState *bs, bool eject_flag)
|
2006-08-19 13:45:59 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
2011-07-20 18:23:42 +02:00
|
|
|
if (drv && drv->bdrv_eject) {
|
|
|
|
drv->bdrv_eject(bs, eject_flag);
|
2006-08-19 13:45:59 +02:00
|
|
|
}
|
2012-02-14 16:41:13 +01:00
|
|
|
|
|
|
|
if (bs->device_name[0] != '\0') {
|
|
|
|
bdrv_emit_qmp_eject_event(bs, eject_flag);
|
|
|
|
}
|
2006-08-19 13:45:59 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Lock or unlock the media (if it is locked, the user won't be able
|
|
|
|
* to eject it manually).
|
|
|
|
*/
|
2011-09-06 18:58:47 +02:00
|
|
|
void bdrv_lock_medium(BlockDriverState *bs, bool locked)
|
2006-08-19 13:45:59 +02:00
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
2011-09-06 18:58:47 +02:00
|
|
|
trace_bdrv_lock_medium(bs, locked);
|
2011-03-29 21:04:40 +02:00
|
|
|
|
2011-09-06 18:58:47 +02:00
|
|
|
if (drv && drv->bdrv_lock_medium) {
|
|
|
|
drv->bdrv_lock_medium(bs, locked);
|
2006-08-19 13:45:59 +02:00
|
|
|
}
|
|
|
|
}
|
2007-12-24 17:10:43 +01:00
|
|
|
|
|
|
|
/* needed for generic scsi interface */
|
|
|
|
|
|
|
|
int bdrv_ioctl(BlockDriverState *bs, unsigned long int req, void *buf)
|
|
|
|
{
|
|
|
|
BlockDriver *drv = bs->drv;
|
|
|
|
|
|
|
|
if (drv && drv->bdrv_ioctl)
|
|
|
|
return drv->bdrv_ioctl(bs, req, buf);
|
|
|
|
return -ENOTSUP;
|
|
|
|
}
|
2009-03-12 20:57:08 +01:00
|
|
|
|
2009-03-28 18:28:41 +01:00
|
|
|
BlockDriverAIOCB *bdrv_aio_ioctl(BlockDriverState *bs,
|
|
|
|
unsigned long int req, void *buf,
|
|
|
|
BlockDriverCompletionFunc *cb, void *opaque)
|
2009-03-12 20:57:08 +01:00
|
|
|
{
|
2009-03-28 18:28:41 +01:00
|
|
|
BlockDriver *drv = bs->drv;
|
2009-03-12 20:57:08 +01:00
|
|
|
|
2009-03-28 18:28:41 +01:00
|
|
|
if (drv && drv->bdrv_aio_ioctl)
|
|
|
|
return drv->bdrv_aio_ioctl(bs, req, buf, cb, opaque);
|
|
|
|
return NULL;
|
2009-03-12 20:57:08 +01:00
|
|
|
}
|
2009-04-22 22:20:00 +02:00
|
|
|
|
2011-09-06 18:58:56 +02:00
|
|
|
void bdrv_set_buffer_alignment(BlockDriverState *bs, int align)
|
|
|
|
{
|
|
|
|
bs->buffer_alignment = align;
|
|
|
|
}
|
2009-11-02 14:40:41 +01:00
|
|
|
|
2009-04-22 22:20:00 +02:00
|
|
|
void *qemu_blockalign(BlockDriverState *bs, size_t size)
|
|
|
|
{
|
|
|
|
return qemu_memalign((bs && bs->buffer_alignment) ? bs->buffer_alignment : 512, size);
|
|
|
|
}
|
2009-11-02 14:40:41 +01:00
|
|
|
|
2013-01-11 16:41:27 +01:00
|
|
|
/*
|
|
|
|
* Check if all memory in this vector is sector aligned.
|
|
|
|
*/
|
|
|
|
bool bdrv_qiov_is_aligned(BlockDriverState *bs, QEMUIOVector *qiov)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < qiov->niov; i++) {
|
|
|
|
if ((uintptr_t) qiov->iov[i].iov_base % bs->buffer_alignment) {
|
|
|
|
return false;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
2009-11-02 14:40:41 +01:00
|
|
|
void bdrv_set_dirty_tracking(BlockDriverState *bs, int enable)
|
|
|
|
{
|
|
|
|
int64_t bitmap_size;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2010-01-26 09:31:48 +01:00
|
|
|
bs->dirty_count = 0;
|
2009-11-30 18:21:19 +01:00
|
|
|
if (enable) {
|
2009-11-30 18:21:20 +01:00
|
|
|
if (!bs->dirty_bitmap) {
|
|
|
|
bitmap_size = (bdrv_getlength(bs) >> BDRV_SECTOR_BITS) +
|
2012-04-12 14:01:04 +02:00
|
|
|
BDRV_SECTORS_PER_DIRTY_CHUNK * BITS_PER_LONG - 1;
|
|
|
|
bitmap_size /= BDRV_SECTORS_PER_DIRTY_CHUNK * BITS_PER_LONG;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2012-04-12 14:01:04 +02:00
|
|
|
bs->dirty_bitmap = g_new0(unsigned long, bitmap_size);
|
2009-11-30 18:21:19 +01:00
|
|
|
}
|
2009-11-02 14:40:41 +01:00
|
|
|
} else {
|
2009-11-30 18:21:20 +01:00
|
|
|
if (bs->dirty_bitmap) {
|
2011-08-21 05:09:37 +02:00
|
|
|
g_free(bs->dirty_bitmap);
|
2009-11-30 18:21:20 +01:00
|
|
|
bs->dirty_bitmap = NULL;
|
2009-11-30 18:21:19 +01:00
|
|
|
}
|
2009-11-02 14:40:41 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_get_dirty(BlockDriverState *bs, int64_t sector)
|
|
|
|
{
|
2009-11-30 18:21:19 +01:00
|
|
|
int64_t chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;
|
2009-11-30 18:21:19 +01:00
|
|
|
|
2009-11-30 18:21:20 +01:00
|
|
|
if (bs->dirty_bitmap &&
|
|
|
|
(sector << BDRV_SECTOR_BITS) < bdrv_getlength(bs)) {
|
2012-10-18 16:49:18 +02:00
|
|
|
return !!(bs->dirty_bitmap[chunk / BITS_PER_LONG] &
|
|
|
|
(1UL << (chunk % BITS_PER_LONG)));
|
2009-11-02 14:40:41 +01:00
|
|
|
} else {
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2012-10-18 16:49:18 +02:00
|
|
|
int64_t bdrv_get_next_dirty(BlockDriverState *bs, int64_t sector)
|
|
|
|
{
|
|
|
|
int64_t chunk;
|
|
|
|
int bit, elem;
|
|
|
|
|
|
|
|
/* Avoid an infinite loop. */
|
|
|
|
assert(bs->dirty_count > 0);
|
|
|
|
|
|
|
|
sector = (sector | (BDRV_SECTORS_PER_DIRTY_CHUNK - 1)) + 1;
|
|
|
|
chunk = sector / (int64_t)BDRV_SECTORS_PER_DIRTY_CHUNK;
|
|
|
|
|
|
|
|
QEMU_BUILD_BUG_ON(sizeof(bs->dirty_bitmap[0]) * 8 != BITS_PER_LONG);
|
|
|
|
elem = chunk / BITS_PER_LONG;
|
|
|
|
bit = chunk % BITS_PER_LONG;
|
|
|
|
for (;;) {
|
|
|
|
if (sector >= bs->total_sectors) {
|
|
|
|
sector = 0;
|
|
|
|
bit = elem = 0;
|
|
|
|
}
|
|
|
|
if (bit == 0 && bs->dirty_bitmap[elem] == 0) {
|
|
|
|
sector += BDRV_SECTORS_PER_DIRTY_CHUNK * BITS_PER_LONG;
|
|
|
|
elem++;
|
|
|
|
} else {
|
|
|
|
if (bs->dirty_bitmap[elem] & (1UL << bit)) {
|
|
|
|
return sector;
|
|
|
|
}
|
|
|
|
sector += BDRV_SECTORS_PER_DIRTY_CHUNK;
|
|
|
|
if (++bit == BITS_PER_LONG) {
|
|
|
|
bit = 0;
|
|
|
|
elem++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_set_dirty(BlockDriverState *bs, int64_t cur_sector,
|
|
|
|
int nr_sectors)
|
|
|
|
{
|
|
|
|
set_dirty_bitmap(bs, cur_sector, nr_sectors, 1);
|
|
|
|
}
|
|
|
|
|
2009-11-30 18:21:19 +01:00
|
|
|
void bdrv_reset_dirty(BlockDriverState *bs, int64_t cur_sector,
|
|
|
|
int nr_sectors)
|
2009-11-02 14:40:41 +01:00
|
|
|
{
|
|
|
|
set_dirty_bitmap(bs, cur_sector, nr_sectors, 0);
|
|
|
|
}
|
2010-01-26 09:31:48 +01:00
|
|
|
|
|
|
|
int64_t bdrv_get_dirty_count(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->dirty_count;
|
|
|
|
}
|
2010-12-16 13:52:15 +01:00
|
|
|
|
2011-01-26 15:12:34 +01:00
|
|
|
void bdrv_set_in_use(BlockDriverState *bs, int in_use)
|
|
|
|
{
|
|
|
|
assert(bs->in_use != in_use);
|
|
|
|
bs->in_use = in_use;
|
|
|
|
}
|
|
|
|
|
|
|
|
int bdrv_in_use(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
return bs->in_use;
|
|
|
|
}
|
|
|
|
|
2011-09-26 22:43:50 +02:00
|
|
|
void bdrv_iostatus_enable(BlockDriverState *bs)
|
|
|
|
{
|
2011-10-14 22:11:23 +02:00
|
|
|
bs->iostatus_enabled = true;
|
2011-10-14 22:22:24 +02:00
|
|
|
bs->iostatus = BLOCK_DEVICE_IO_STATUS_OK;
|
2011-09-26 22:43:50 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
/* The I/O status is only enabled if the drive explicitly
|
|
|
|
* enables it _and_ the VM is configured to stop on errors */
|
|
|
|
bool bdrv_iostatus_is_enabled(const BlockDriverState *bs)
|
|
|
|
{
|
2011-10-14 22:11:23 +02:00
|
|
|
return (bs->iostatus_enabled &&
|
2012-09-28 17:22:55 +02:00
|
|
|
(bs->on_write_error == BLOCKDEV_ON_ERROR_ENOSPC ||
|
|
|
|
bs->on_write_error == BLOCKDEV_ON_ERROR_STOP ||
|
|
|
|
bs->on_read_error == BLOCKDEV_ON_ERROR_STOP));
|
2011-09-26 22:43:50 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_iostatus_disable(BlockDriverState *bs)
|
|
|
|
{
|
2011-10-14 22:11:23 +02:00
|
|
|
bs->iostatus_enabled = false;
|
2011-09-26 22:43:50 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_iostatus_reset(BlockDriverState *bs)
|
|
|
|
{
|
|
|
|
if (bdrv_iostatus_is_enabled(bs)) {
|
2011-10-14 22:22:24 +02:00
|
|
|
bs->iostatus = BLOCK_DEVICE_IO_STATUS_OK;
|
2012-10-18 16:49:27 +02:00
|
|
|
if (bs->job) {
|
|
|
|
block_job_iostatus_reset(bs->job);
|
|
|
|
}
|
2011-09-26 22:43:50 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
void bdrv_iostatus_set_err(BlockDriverState *bs, int error)
|
|
|
|
{
|
2012-09-28 17:22:57 +02:00
|
|
|
assert(bdrv_iostatus_is_enabled(bs));
|
|
|
|
if (bs->iostatus == BLOCK_DEVICE_IO_STATUS_OK) {
|
2011-10-14 22:22:24 +02:00
|
|
|
bs->iostatus = error == ENOSPC ? BLOCK_DEVICE_IO_STATUS_NOSPACE :
|
|
|
|
BLOCK_DEVICE_IO_STATUS_FAILED;
|
2011-09-26 22:43:50 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-08-25 08:26:01 +02:00
|
|
|
void
|
|
|
|
bdrv_acct_start(BlockDriverState *bs, BlockAcctCookie *cookie, int64_t bytes,
|
|
|
|
enum BlockAcctType type)
|
|
|
|
{
|
|
|
|
assert(type < BDRV_MAX_IOTYPE);
|
|
|
|
|
|
|
|
cookie->bytes = bytes;
|
2011-08-25 08:26:10 +02:00
|
|
|
cookie->start_time_ns = get_clock();
|
2011-08-25 08:26:01 +02:00
|
|
|
cookie->type = type;
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
bdrv_acct_done(BlockDriverState *bs, BlockAcctCookie *cookie)
|
|
|
|
{
|
|
|
|
assert(cookie->type < BDRV_MAX_IOTYPE);
|
|
|
|
|
|
|
|
bs->nr_bytes[cookie->type] += cookie->bytes;
|
|
|
|
bs->nr_ops[cookie->type]++;
|
2011-08-25 08:26:10 +02:00
|
|
|
bs->total_time_ns[cookie->type] += get_clock() - cookie->start_time_ns;
|
2011-08-25 08:26:01 +02:00
|
|
|
}
|
|
|
|
|
2012-11-30 13:52:09 +01:00
|
|
|
void bdrv_img_create(const char *filename, const char *fmt,
|
|
|
|
const char *base_filename, const char *base_fmt,
|
|
|
|
char *options, uint64_t img_size, int flags, Error **errp)
|
2010-12-16 13:52:15 +01:00
|
|
|
{
|
|
|
|
QEMUOptionParameter *param = NULL, *create_options = NULL;
|
2011-06-01 14:03:31 +02:00
|
|
|
QEMUOptionParameter *backing_fmt, *backing_file, *size;
|
2010-12-16 13:52:15 +01:00
|
|
|
BlockDriverState *bs = NULL;
|
|
|
|
BlockDriver *drv, *proto_drv;
|
2011-01-24 10:32:20 +01:00
|
|
|
BlockDriver *backing_drv = NULL;
|
2010-12-16 13:52:15 +01:00
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* Find driver and parse its options */
|
|
|
|
drv = bdrv_find_format(fmt);
|
|
|
|
if (!drv) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Unknown file format '%s'", fmt);
|
2012-11-30 13:52:09 +01:00
|
|
|
return;
|
2010-12-16 13:52:15 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
proto_drv = bdrv_find_protocol(filename);
|
|
|
|
if (!proto_drv) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Unknown protocol '%s'", filename);
|
2012-11-30 13:52:09 +01:00
|
|
|
return;
|
2010-12-16 13:52:15 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
create_options = append_option_parameters(create_options,
|
|
|
|
drv->create_options);
|
|
|
|
create_options = append_option_parameters(create_options,
|
|
|
|
proto_drv->create_options);
|
|
|
|
|
|
|
|
/* Create parameter list with default values */
|
|
|
|
param = parse_option_parameters("", create_options, param);
|
|
|
|
|
|
|
|
set_option_parameter_int(param, BLOCK_OPT_SIZE, img_size);
|
|
|
|
|
|
|
|
/* Parse -o options */
|
|
|
|
if (options) {
|
|
|
|
param = parse_option_parameters(options, create_options, param);
|
|
|
|
if (param == NULL) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Invalid options for file format '%s'.", fmt);
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (base_filename) {
|
|
|
|
if (set_option_parameter(param, BLOCK_OPT_BACKING_FILE,
|
|
|
|
base_filename)) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Backing file not supported for file format '%s'",
|
|
|
|
fmt);
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (base_fmt) {
|
|
|
|
if (set_option_parameter(param, BLOCK_OPT_BACKING_FMT, base_fmt)) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Backing file format not supported for file "
|
|
|
|
"format '%s'", fmt);
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-12-16 13:52:17 +01:00
|
|
|
backing_file = get_option_parameter(param, BLOCK_OPT_BACKING_FILE);
|
|
|
|
if (backing_file && backing_file->value.s) {
|
|
|
|
if (!strcmp(filename, backing_file->value.s)) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Error: Trying to create an image with the "
|
|
|
|
"same filename as the backing file");
|
2010-12-16 13:52:17 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-12-16 13:52:15 +01:00
|
|
|
backing_fmt = get_option_parameter(param, BLOCK_OPT_BACKING_FMT);
|
|
|
|
if (backing_fmt && backing_fmt->value.s) {
|
2011-01-24 10:32:20 +01:00
|
|
|
backing_drv = bdrv_find_format(backing_fmt->value.s);
|
|
|
|
if (!backing_drv) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Unknown backing file format '%s'",
|
|
|
|
backing_fmt->value.s);
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// The size for the image must always be specified, with one exception:
|
|
|
|
// If we are using a backing file, we can obtain the size from there
|
2011-06-01 14:03:31 +02:00
|
|
|
size = get_option_parameter(param, BLOCK_OPT_SIZE);
|
|
|
|
if (size && size->value.n == -1) {
|
2010-12-16 13:52:15 +01:00
|
|
|
if (backing_file && backing_file->value.s) {
|
|
|
|
uint64_t size;
|
|
|
|
char buf[32];
|
2012-04-12 14:01:03 +02:00
|
|
|
int back_flags;
|
|
|
|
|
|
|
|
/* backing files always opened read-only */
|
|
|
|
back_flags =
|
|
|
|
flags & ~(BDRV_O_RDWR | BDRV_O_SNAPSHOT | BDRV_O_NO_BACKING);
|
2010-12-16 13:52:15 +01:00
|
|
|
|
|
|
|
bs = bdrv_new("");
|
|
|
|
|
2012-04-12 14:01:03 +02:00
|
|
|
ret = bdrv_open(bs, backing_file->value.s, back_flags, backing_drv);
|
2010-12-16 13:52:15 +01:00
|
|
|
if (ret < 0) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg_errno(errp, -ret, "Could not open '%s'",
|
|
|
|
backing_file->value.s);
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
bdrv_get_geometry(bs, &size);
|
|
|
|
size *= 512;
|
|
|
|
|
|
|
|
snprintf(buf, sizeof(buf), "%" PRId64, size);
|
|
|
|
set_option_parameter(param, BLOCK_OPT_SIZE, buf);
|
|
|
|
} else {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "Image creation needs a size parameter");
|
2010-12-16 13:52:15 +01:00
|
|
|
goto out;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
printf("Formatting '%s', fmt=%s ", filename, fmt);
|
|
|
|
print_option_parameters(param);
|
|
|
|
puts("");
|
|
|
|
|
|
|
|
ret = bdrv_create(drv, filename, param);
|
|
|
|
if (ret < 0) {
|
|
|
|
if (ret == -ENOTSUP) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp,"Formatting or formatting option not supported for "
|
|
|
|
"file format '%s'", fmt);
|
2010-12-16 13:52:15 +01:00
|
|
|
} else if (ret == -EFBIG) {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "The image size is too large for file format '%s'",
|
|
|
|
fmt);
|
2010-12-16 13:52:15 +01:00
|
|
|
} else {
|
2012-11-30 13:52:04 +01:00
|
|
|
error_setg(errp, "%s: error while creating %s: %s", filename, fmt,
|
|
|
|
strerror(-ret));
|
2010-12-16 13:52:15 +01:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
free_option_parameters(create_options);
|
|
|
|
free_option_parameters(param);
|
|
|
|
|
|
|
|
if (bs) {
|
|
|
|
bdrv_delete(bs);
|
|
|
|
}
|
|
|
|
}
|