qemu-e2k/block/backup.c

722 lines
23 KiB
C
Raw Normal View History

block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
/*
* QEMU backup
*
* Copyright (C) 2013 Proxmox Server Solutions
*
* Authors:
* Dietmar Maurer (dietmar@proxmox.com)
*
* This work is licensed under the terms of the GNU GPL, version 2 or later.
* See the COPYING file in the top-level directory.
*
*/
#include "qemu/osdep.h"
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
#include "trace.h"
#include "block/block.h"
#include "block/block_int.h"
#include "block/blockjob_int.h"
#include "block/block_backup.h"
2016-03-14 09:01:28 +01:00
#include "qapi/error.h"
#include "qapi/qmp/qerror.h"
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
#include "qemu/ratelimit.h"
#include "qemu/cutils.h"
#include "sysemu/block-backend.h"
#include "qemu/bitmap.h"
#include "qemu/error-report.h"
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
#define BACKUP_CLUSTER_SIZE_DEFAULT (1 << 16)
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
typedef struct CowRequest {
int64_t start_byte;
int64_t end_byte;
QLIST_ENTRY(CowRequest) list;
CoQueue wait_queue; /* coroutines blocked on this request */
} CowRequest;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
typedef struct BackupBlockJob {
BlockJob common;
BlockBackend *target;
/* bitmap for sync=incremental */
BdrvDirtyBitmap *sync_bitmap;
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
MirrorSyncMode sync_mode;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
BlockdevOnError on_source_error;
BlockdevOnError on_target_error;
CoRwlock flush_rwlock;
uint64_t len;
uint64_t bytes_read;
int64_t cluster_size;
bool compress;
NotifierWithReturn before_write;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
QLIST_HEAD(, CowRequest) inflight_reqs;
HBitmap *copy_bitmap;
bool use_copy_range;
int64_t copy_range_size;
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
bool serialize_target_writes;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
} BackupBlockJob;
static const BlockJobDriver backup_job_driver;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
/* See if in-flight requests overlap and wait for them to complete */
static void coroutine_fn wait_for_overlapping_requests(BackupBlockJob *job,
int64_t start,
int64_t end)
{
CowRequest *req;
bool retry;
do {
retry = false;
QLIST_FOREACH(req, &job->inflight_reqs, list) {
if (end > req->start_byte && start < req->end_byte) {
qemu_co_queue_wait(&req->wait_queue, NULL);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
retry = true;
break;
}
}
} while (retry);
}
/* Keep track of an in-flight request */
static void cow_request_begin(CowRequest *req, BackupBlockJob *job,
int64_t start, int64_t end)
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
{
req->start_byte = start;
req->end_byte = end;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
qemu_co_queue_init(&req->wait_queue);
QLIST_INSERT_HEAD(&job->inflight_reqs, req, list);
}
/* Forget about a completed request */
static void cow_request_end(CowRequest *req)
{
QLIST_REMOVE(req, list);
qemu_co_queue_restart_all(&req->wait_queue);
}
/* Copy range to target with a bounce buffer and return the bytes copied. If
* error occurred, return a negative error number */
static int coroutine_fn backup_cow_with_bounce_buffer(BackupBlockJob *job,
int64_t start,
int64_t end,
bool is_write_notifier,
bool *error_is_read,
void **bounce_buffer)
{
int ret;
QEMUIOVector qiov;
BlockBackend *blk = job->common.blk;
int nbytes;
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
int read_flags = is_write_notifier ? BDRV_REQ_NO_SERIALISING : 0;
int write_flags = job->serialize_target_writes ? BDRV_REQ_SERIALISING : 0;
hbitmap_reset(job->copy_bitmap, start / job->cluster_size, 1);
nbytes = MIN(job->cluster_size, job->len - start);
if (!*bounce_buffer) {
*bounce_buffer = blk_blockalign(blk, job->cluster_size);
}
qemu_iovec_init_buf(&qiov, *bounce_buffer, nbytes);
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
ret = blk_co_preadv(blk, start, qiov.size, &qiov, read_flags);
if (ret < 0) {
trace_backup_do_cow_read_fail(job, start, ret);
if (error_is_read) {
*error_is_read = true;
}
goto fail;
}
if (qemu_iovec_is_zero(&qiov)) {
ret = blk_co_pwrite_zeroes(job->target, start,
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
qiov.size, write_flags | BDRV_REQ_MAY_UNMAP);
} else {
ret = blk_co_pwritev(job->target, start,
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
qiov.size, &qiov, write_flags |
(job->compress ? BDRV_REQ_WRITE_COMPRESSED : 0));
}
if (ret < 0) {
trace_backup_do_cow_write_fail(job, start, ret);
if (error_is_read) {
*error_is_read = false;
}
goto fail;
}
return nbytes;
fail:
hbitmap_set(job->copy_bitmap, start / job->cluster_size, 1);
return ret;
}
/* Copy range to target and return the bytes copied. If error occurred, return a
* negative error number. */
static int coroutine_fn backup_cow_with_offload(BackupBlockJob *job,
int64_t start,
int64_t end,
bool is_write_notifier)
{
int ret;
int nr_clusters;
BlockBackend *blk = job->common.blk;
int nbytes;
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
int read_flags = is_write_notifier ? BDRV_REQ_NO_SERIALISING : 0;
int write_flags = job->serialize_target_writes ? BDRV_REQ_SERIALISING : 0;
assert(QEMU_IS_ALIGNED(job->copy_range_size, job->cluster_size));
nbytes = MIN(job->copy_range_size, end - start);
nr_clusters = DIV_ROUND_UP(nbytes, job->cluster_size);
hbitmap_reset(job->copy_bitmap, start / job->cluster_size,
nr_clusters);
ret = blk_co_copy_range(blk, start, job->target, start, nbytes,
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
read_flags, write_flags);
if (ret < 0) {
trace_backup_do_cow_copy_range_fail(job, start, ret);
hbitmap_set(job->copy_bitmap, start / job->cluster_size,
nr_clusters);
return ret;
}
return nbytes;
}
static int coroutine_fn backup_do_cow(BackupBlockJob *job,
int64_t offset, uint64_t bytes,
bool *error_is_read,
bool is_write_notifier)
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
{
CowRequest cow_request;
int ret = 0;
int64_t start, end; /* bytes */
void *bounce_buffer = NULL;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
qemu_co_rwlock_rdlock(&job->flush_rwlock);
start = QEMU_ALIGN_DOWN(offset, job->cluster_size);
end = QEMU_ALIGN_UP(bytes + offset, job->cluster_size);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
trace_backup_do_cow_enter(job, start, offset, bytes);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
wait_for_overlapping_requests(job, start, end);
cow_request_begin(&cow_request, job, start, end);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
while (start < end) {
if (!hbitmap_get(job->copy_bitmap, start / job->cluster_size)) {
trace_backup_do_cow_skip(job, start);
start += job->cluster_size;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
continue; /* already copied */
}
trace_backup_do_cow_process(job, start);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
if (job->use_copy_range) {
ret = backup_cow_with_offload(job, start, end, is_write_notifier);
if (ret < 0) {
job->use_copy_range = false;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
}
if (!job->use_copy_range) {
ret = backup_cow_with_bounce_buffer(job, start, end, is_write_notifier,
error_is_read, &bounce_buffer);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
if (ret < 0) {
break;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
/* Publish progress, guest I/O counts as progress too. Note that the
* offset field is an opaque progress value, it is not a disk offset.
*/
start += ret;
job->bytes_read += ret;
job_progress_update(&job->common.job, ret);
ret = 0;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
if (bounce_buffer) {
qemu_vfree(bounce_buffer);
}
cow_request_end(&cow_request);
trace_backup_do_cow_return(job, offset, bytes, ret);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
qemu_co_rwlock_unlock(&job->flush_rwlock);
return ret;
}
static int coroutine_fn backup_before_write_notify(
NotifierWithReturn *notifier,
void *opaque)
{
BackupBlockJob *job = container_of(notifier, BackupBlockJob, before_write);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
BdrvTrackedRequest *req = opaque;
assert(req->bs == blk_bs(job->common.blk));
assert(QEMU_IS_ALIGNED(req->offset, BDRV_SECTOR_SIZE));
assert(QEMU_IS_ALIGNED(req->bytes, BDRV_SECTOR_SIZE));
return backup_do_cow(job, req->offset, req->bytes, NULL, true);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
static void backup_cleanup_sync_bitmap(BackupBlockJob *job, int ret)
{
BdrvDirtyBitmap *bm;
BlockDriverState *bs = blk_bs(job->common.blk);
2018-03-10 09:27:37 +01:00
if (ret < 0) {
/* Merge the successor back into the parent, delete nothing. */
bm = bdrv_reclaim_dirty_bitmap(bs, job->sync_bitmap, NULL);
assert(bm);
} else {
/* Everything is fine, delete this bitmap and install the backup. */
bm = bdrv_dirty_bitmap_abdicate(bs, job->sync_bitmap, NULL);
assert(bm);
}
}
static void backup_commit(Job *job)
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common.job);
if (s->sync_bitmap) {
backup_cleanup_sync_bitmap(s, 0);
}
}
static void backup_abort(Job *job)
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common.job);
if (s->sync_bitmap) {
backup_cleanup_sync_bitmap(s, -1);
}
}
static void backup_clean(Job *job)
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common.job);
assert(s->target);
blk_unref(s->target);
s->target = NULL;
}
static void backup_attached_aio_context(BlockJob *job, AioContext *aio_context)
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common);
blk_set_aio_context(s->target, aio_context);
}
void backup_do_checkpoint(BlockJob *job, Error **errp)
{
BackupBlockJob *backup_job = container_of(job, BackupBlockJob, common);
int64_t len;
assert(block_job_driver(job) == &backup_job_driver);
if (backup_job->sync_mode != MIRROR_SYNC_MODE_NONE) {
error_setg(errp, "The backup job only supports block checkpoint in"
" sync=none mode");
return;
}
len = DIV_ROUND_UP(backup_job->len, backup_job->cluster_size);
hbitmap_set(backup_job->copy_bitmap, 0, len);
}
static void backup_drain(BlockJob *job)
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common);
/* Need to keep a reference in case blk_drain triggers execution
* of backup_complete...
*/
if (s->target) {
BlockBackend *target = s->target;
blk_ref(target);
blk_drain(target);
blk_unref(target);
}
}
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
static BlockErrorAction backup_error_action(BackupBlockJob *job,
bool read, int error)
{
if (read) {
return block_job_error_action(&job->common, job->on_source_error,
true, error);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
} else {
return block_job_error_action(&job->common, job->on_target_error,
false, error);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
}
static bool coroutine_fn yield_and_check(BackupBlockJob *job)
{
uint64_t delay_ns;
if (job_is_cancelled(&job->common.job)) {
return true;
}
/* We need to yield even for delay_ns = 0 so that bdrv_drain_all() can
* return. Without a yield, the VM would not reboot. */
delay_ns = block_job_ratelimit_get_delay(&job->common, job->bytes_read);
job->bytes_read = 0;
job_sleep_ns(&job->common.job, delay_ns);
if (job_is_cancelled(&job->common.job)) {
return true;
}
return false;
}
static int coroutine_fn backup_run_incremental(BackupBlockJob *job)
{
int ret;
bool error_is_read;
int64_t cluster;
HBitmapIter hbi;
hbitmap_iter_init(&hbi, job->copy_bitmap, 0);
while ((cluster = hbitmap_iter_next(&hbi)) != -1) {
do {
if (yield_and_check(job)) {
return 0;
}
ret = backup_do_cow(job, cluster * job->cluster_size,
job->cluster_size, &error_is_read, false);
if (ret < 0 && backup_error_action(job, error_is_read, -ret) ==
BLOCK_ERROR_ACTION_REPORT)
{
return ret;
}
} while (ret < 0);
}
return 0;
}
/* init copy_bitmap from sync_bitmap */
static void backup_incremental_init_copy_bitmap(BackupBlockJob *job)
{
BdrvDirtyBitmapIter *dbi;
int64_t offset;
int64_t end = DIV_ROUND_UP(bdrv_dirty_bitmap_size(job->sync_bitmap),
job->cluster_size);
dbi = bdrv_dirty_iter_new(job->sync_bitmap);
while ((offset = bdrv_dirty_iter_next(dbi)) != -1) {
int64_t cluster = offset / job->cluster_size;
int64_t next_cluster;
offset += bdrv_dirty_bitmap_granularity(job->sync_bitmap);
if (offset >= bdrv_dirty_bitmap_size(job->sync_bitmap)) {
hbitmap_set(job->copy_bitmap, cluster, end - cluster);
break;
}
offset = bdrv_dirty_bitmap_next_zero(job->sync_bitmap, offset,
UINT64_MAX);
if (offset == -1) {
hbitmap_set(job->copy_bitmap, cluster, end - cluster);
break;
}
next_cluster = DIV_ROUND_UP(offset, job->cluster_size);
hbitmap_set(job->copy_bitmap, cluster, next_cluster - cluster);
if (next_cluster >= end) {
break;
}
bdrv_set_dirty_iter(dbi, next_cluster * job->cluster_size);
}
/* TODO job_progress_set_remaining() would make more sense */
job_progress_update(&job->common.job,
job->len - hbitmap_count(job->copy_bitmap) * job->cluster_size);
bdrv_dirty_iter_free(dbi);
}
static int coroutine_fn backup_run(Job *job, Error **errp)
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
{
BackupBlockJob *s = container_of(job, BackupBlockJob, common.job);
BlockDriverState *bs = blk_bs(s->common.blk);
int64_t offset, nb_clusters;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
int ret = 0;
QLIST_INIT(&s->inflight_reqs);
qemu_co_rwlock_init(&s->flush_rwlock);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
nb_clusters = DIV_ROUND_UP(s->len, s->cluster_size);
job_progress_set_remaining(job, s->len);
s->copy_bitmap = hbitmap_alloc(nb_clusters, 0);
if (s->sync_mode == MIRROR_SYNC_MODE_INCREMENTAL) {
backup_incremental_init_copy_bitmap(s);
} else {
hbitmap_set(s->copy_bitmap, 0, nb_clusters);
}
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
s->before_write.notify = backup_before_write_notify;
bdrv_add_before_write_notifier(bs, &s->before_write);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
if (s->sync_mode == MIRROR_SYNC_MODE_NONE) {
/* All bits are set in copy_bitmap to allow any cluster to be copied.
* This does not actually require them to be copied. */
while (!job_is_cancelled(job)) {
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
/* Yield until the job is cancelled. We just let our before_write
* notify callback service CoW requests. */
job_yield(job);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
} else if (s->sync_mode == MIRROR_SYNC_MODE_INCREMENTAL) {
ret = backup_run_incremental(s);
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
} else {
/* Both FULL and TOP SYNC_MODE's require copying.. */
for (offset = 0; offset < s->len;
offset += s->cluster_size) {
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
bool error_is_read;
int alloced = 0;
if (yield_and_check(s)) {
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
break;
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
}
if (s->sync_mode == MIRROR_SYNC_MODE_TOP) {
block: Make bdrv_is_allocated() byte-based We are gradually moving away from sector-based interfaces, towards byte-based. In the common case, allocation is unlikely to ever use values that are not naturally sector-aligned, but it is possible that byte-based values will let us be more precise about allocation at the end of an unaligned file that can do byte-based access. Changing the signature of the function to use int64_t *pnum ensures that the compiler enforces that all callers are updated. For now, the io.c layer still assert()s that all callers are sector-aligned on input and that *pnum is sector-aligned on return to the caller, but that can be relaxed when a later patch implements byte-based block status. Therefore, this code adds usages like DIV_ROUND_UP(,BDRV_SECTOR_SIZE) to callers that still want aligned values, where the call might reasonbly give non-aligned results in the future; on the other hand, no rounding is needed for callers that should just continue to work with byte alignment. For the most part this patch is just the addition of scaling at the callers followed by inverse scaling at bdrv_is_allocated(). But some code, particularly bdrv_commit(), gets a lot simpler because it no longer has to mess with sectors; also, it is now possible to pass NULL if the caller does not care how much of the image is allocated beyond the initial offset. Leave comments where we can further simplify once a later patch eliminates the need for sector-aligned requests through bdrv_is_allocated(). For ease of review, bdrv_is_allocated_above() will be tackled separately. Signed-off-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2017-07-07 14:44:57 +02:00
int i;
int64_t n;
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
/* Check to see if these blocks are already in the
* backing file. */
for (i = 0; i < s->cluster_size;) {
/* bdrv_is_allocated() only returns true/false based
* on the first set of sectors it comes across that
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
* are are all in the same state.
* For that reason we must verify each sector in the
* backup cluster length. We end up copying more than
* needed but at some point that is always the case. */
alloced =
block: Make bdrv_is_allocated() byte-based We are gradually moving away from sector-based interfaces, towards byte-based. In the common case, allocation is unlikely to ever use values that are not naturally sector-aligned, but it is possible that byte-based values will let us be more precise about allocation at the end of an unaligned file that can do byte-based access. Changing the signature of the function to use int64_t *pnum ensures that the compiler enforces that all callers are updated. For now, the io.c layer still assert()s that all callers are sector-aligned on input and that *pnum is sector-aligned on return to the caller, but that can be relaxed when a later patch implements byte-based block status. Therefore, this code adds usages like DIV_ROUND_UP(,BDRV_SECTOR_SIZE) to callers that still want aligned values, where the call might reasonbly give non-aligned results in the future; on the other hand, no rounding is needed for callers that should just continue to work with byte alignment. For the most part this patch is just the addition of scaling at the callers followed by inverse scaling at bdrv_is_allocated(). But some code, particularly bdrv_commit(), gets a lot simpler because it no longer has to mess with sectors; also, it is now possible to pass NULL if the caller does not care how much of the image is allocated beyond the initial offset. Leave comments where we can further simplify once a later patch eliminates the need for sector-aligned requests through bdrv_is_allocated(). For ease of review, bdrv_is_allocated_above() will be tackled separately. Signed-off-by: Eric Blake <eblake@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2017-07-07 14:44:57 +02:00
bdrv_is_allocated(bs, offset + i,
s->cluster_size - i, &n);
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
i += n;
if (alloced || n == 0) {
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
break;
}
}
/* If the above loop never found any sectors that are in
* the topmost image, skip this backup. */
if (alloced == 0) {
continue;
}
}
/* FULL sync mode we copy the whole drive. */
if (alloced < 0) {
ret = alloced;
} else {
ret = backup_do_cow(s, offset, s->cluster_size,
&error_is_read, false);
}
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
if (ret < 0) {
/* Depending on error action, fail now or retry cluster */
BlockErrorAction action =
backup_error_action(s, error_is_read, -ret);
if (action == BLOCK_ERROR_ACTION_REPORT) {
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
break;
} else {
offset -= s->cluster_size;
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
continue;
}
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
}
}
notifier_with_return_remove(&s->before_write);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
/* wait until pending backup_do_cow() calls have completed */
qemu_co_rwlock_wrlock(&s->flush_rwlock);
qemu_co_rwlock_unlock(&s->flush_rwlock);
hbitmap_free(s->copy_bitmap);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
return ret;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
static const BlockJobDriver backup_job_driver = {
.job_driver = {
.instance_size = sizeof(BackupBlockJob),
.job_type = JOB_TYPE_BACKUP,
.free = block_job_free,
.user_resume = block_job_user_resume,
.drain = block_job_drain,
.run = backup_run,
.commit = backup_commit,
.abort = backup_abort,
.clean = backup_clean,
},
.attached_aio_context = backup_attached_aio_context,
.drain = backup_drain,
};
BlockJob *backup_job_create(const char *job_id, BlockDriverState *bs,
BlockDriverState *target, int64_t speed,
MirrorSyncMode sync_mode, BdrvDirtyBitmap *sync_bitmap,
bool compress,
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
BlockdevOnError on_source_error,
BlockdevOnError on_target_error,
int creation_flags,
BlockCompletionFunc *cb, void *opaque,
JobTxn *txn, Error **errp)
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
{
int64_t len;
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
BlockDriverInfo bdi;
BackupBlockJob *job = NULL;
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
int ret;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
assert(bs);
assert(target);
if (bs == target) {
error_setg(errp, "Source and target cannot be the same");
return NULL;
}
if (!bdrv_is_inserted(bs)) {
error_setg(errp, "Device is not inserted: %s",
bdrv_get_device_name(bs));
return NULL;
}
if (!bdrv_is_inserted(target)) {
error_setg(errp, "Device is not inserted: %s",
bdrv_get_device_name(target));
return NULL;
}
if (compress && target->drv->bdrv_co_pwritev_compressed == NULL) {
error_setg(errp, "Compression is not supported for this drive %s",
bdrv_get_device_name(target));
return NULL;
}
if (bdrv_op_is_blocked(bs, BLOCK_OP_TYPE_BACKUP_SOURCE, errp)) {
return NULL;
}
if (bdrv_op_is_blocked(target, BLOCK_OP_TYPE_BACKUP_TARGET, errp)) {
return NULL;
}
if (sync_mode == MIRROR_SYNC_MODE_INCREMENTAL) {
if (!sync_bitmap) {
error_setg(errp, "must provide a valid bitmap name for "
"\"incremental\" sync mode");
return NULL;
}
/* Create a new bitmap, and freeze/disable this one. */
if (bdrv_dirty_bitmap_create_successor(bs, sync_bitmap, errp) < 0) {
return NULL;
}
} else if (sync_bitmap) {
error_setg(errp,
"a sync_bitmap was provided to backup_run, "
"but received an incompatible sync_mode (%s)",
MirrorSyncMode_str(sync_mode));
return NULL;
}
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
len = bdrv_getlength(bs);
if (len < 0) {
error_setg_errno(errp, -len, "unable to get length for '%s'",
bdrv_get_device_name(bs));
goto error;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
/* job->len is fixed, so we can't allow resize */
job = block_job_create(job_id, &backup_job_driver, txn, bs,
BLK_PERM_CONSISTENT_READ,
BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE |
BLK_PERM_WRITE_UNCHANGED | BLK_PERM_GRAPH_MOD,
speed, creation_flags, cb, opaque, errp);
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
if (!job) {
goto error;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}
/* The target must match the source in size, so no resize here either */
job->target = blk_new(BLK_PERM_WRITE,
BLK_PERM_CONSISTENT_READ | BLK_PERM_WRITE |
BLK_PERM_WRITE_UNCHANGED | BLK_PERM_GRAPH_MOD);
ret = blk_insert_bs(job->target, target, errp);
if (ret < 0) {
goto error;
}
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
job->on_source_error = on_source_error;
job->on_target_error = on_target_error;
Implement sync modes for drive-backup. This patch adds sync-modes to the drive-backup interface and implements the FULL, NONE and TOP modes of synchronization. FULL performs as before copying the entire contents of the drive while preserving the point-in-time using CoW. NONE only copies new writes to the target drive. TOP copies changes to the topmost drive image and preserves the point-in-time using CoW. For sync mode TOP are creating a new target image using the same backing file as the original disk image. Then any new data that has been laid on top of it since creation is copied in the main backup_run() loop. There is an extra check in the 'TOP' case so that we don't bother to copy all the data of the backing file as it already exists in the target. This is where the bdrv_co_is_allocated() is used to determine if the data exists in the topmost layer or below. Also any new data being written is intercepted via the write_notifier hook which ends up calling backup_do_cow() to copy old data out before it gets overwritten. For mode 'NONE' we create the new target image and only copy in the original data from the disk image starting from the time the call was made. This preserves the point in time data by only copying the parts that are *going to change* to the target image. This way we can reconstruct the final image by checking to see if the given block exists in the new target image first, and if it does not, you can get it from the original image. This is basically an optimization allowing you to do point-in-time snapshots with low overhead vs the 'FULL' version. Since there is no old data to copy out the loop in backup_run() for the NONE case just calls qemu_coroutine_yield() which only wakes up after an event (usually cancel in this case). The rest is handled by the before_write notifier which again calls backup_do_cow() to write out the old data so it can be preserved. Signed-off-by: Ian Main <imain@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-07-26 20:39:04 +02:00
job->sync_mode = sync_mode;
job->sync_bitmap = sync_mode == MIRROR_SYNC_MODE_INCREMENTAL ?
sync_bitmap : NULL;
job->compress = compress;
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
block/backup: fix fleecing scheme: use serialized writes Fleecing scheme works as follows: we want a kind of temporary snapshot of active drive A. We create temporary image B, with B->backing = A. Then we start backup(sync=none) from A to B. From this point, B reads as point-in-time snapshot of A (A continues to be active drive, accepting guest IO). This scheme needs some additional synchronization between reads from B and backup COW operations, otherwise, the following situation is theoretically possible: (assume B is qcow2, client is NBD client, reading from B) 1. client starts reading and take qcow2 mutex in qcow2_co_preadv, and goes up to l2 table loading (assume cache miss) 2) guest write => backup COW => qcow2 write => try to take qcow2 mutex => waiting 3. l2 table loaded, we see that cluster is UNALLOCATED, go to "case QCOW2_CLUSTER_UNALLOCATED" and unlock mutex before bdrv_co_preadv(bs->backing, ...) 4) aha, mutex unlocked, backup COW continues, and we finally finish guest write and change cluster in our active disk A 5. actually, do bdrv_co_preadv(bs->backing, ...) and read _new updated_ data. To avoid this, let's make backup writes serializing, to not intersect with reads from B. Note: we expand range of handled cases from (sync=none and B->backing = A) to just (A in backing chain of B), to finally allow safe reading from B during backup for all cases when A in backing chain of B, i.e. B formally looks like point-in-time snapshot of A. Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Reviewed-by: Fam Zheng <famz@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2018-07-09 18:37:19 +02:00
/* Detect image-fleecing (and similar) schemes */
job->serialize_target_writes = bdrv_chain_contains(target, bs);
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
/* If there is no backing file on the target, we cannot rely on COW if our
* backup cluster size is smaller than the target cluster size. Even for
* targets with a backing file, try to avoid COW if possible. */
ret = bdrv_get_info(target, &bdi);
if (ret == -ENOTSUP && !target->backing) {
/* Cluster size is not defined */
Convert error_report() to warn_report() Convert all uses of error_report("warning:"... to use warn_report() instead. This helps standardise on a single method of printing warnings to the user. All of the warnings were changed using these two commands: find ./* -type f -exec sed -i \ 's|error_report(".*warning[,:] |warn_report("|Ig' {} + Indentation fixed up manually afterwards. The test-qdev-global-props test case was manually updated to ensure that this patch passes make check (as the test cases are case sensitive). Signed-off-by: Alistair Francis <alistair.francis@xilinx.com> Suggested-by: Thomas Huth <thuth@redhat.com> Cc: Jeff Cody <jcody@redhat.com> Cc: Kevin Wolf <kwolf@redhat.com> Cc: Max Reitz <mreitz@redhat.com> Cc: Ronnie Sahlberg <ronniesahlberg@gmail.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Peter Lieven <pl@kamp.de> Cc: Josh Durgin <jdurgin@redhat.com> Cc: "Richard W.M. Jones" <rjones@redhat.com> Cc: Markus Armbruster <armbru@redhat.com> Cc: Peter Crosthwaite <crosthwaite.peter@gmail.com> Cc: Richard Henderson <rth@twiddle.net> Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com> Cc: Greg Kurz <groug@kaod.org> Cc: Rob Herring <robh@kernel.org> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Peter Chubb <peter.chubb@nicta.com.au> Cc: Eduardo Habkost <ehabkost@redhat.com> Cc: Marcel Apfelbaum <marcel@redhat.com> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Igor Mammedov <imammedo@redhat.com> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: Alexander Graf <agraf@suse.de> Cc: Gerd Hoffmann <kraxel@redhat.com> Cc: Jason Wang <jasowang@redhat.com> Cc: Marcelo Tosatti <mtosatti@redhat.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Cornelia Huck <cohuck@redhat.com> Cc: Stefan Hajnoczi <stefanha@redhat.com> Acked-by: David Gibson <david@gibson.dropbear.id.au> Acked-by: Greg Kurz <groug@kaod.org> Acked-by: Cornelia Huck <cohuck@redhat.com> Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com> Reviewed by: Peter Chubb <peter.chubb@data61.csiro.au> Acked-by: Max Reitz <mreitz@redhat.com> Acked-by: Marcel Apfelbaum <marcel@redhat.com> Message-Id: <e1cfa2cd47087c248dd24caca9c33d9af0c499b0.1499866456.git.alistair.francis@xilinx.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> Signed-off-by: Markus Armbruster <armbru@redhat.com>
2017-07-12 15:57:41 +02:00
warn_report("The target block device doesn't provide "
"information about the block size and it doesn't have a "
"backing file. The default block size of %u bytes is "
"used. If the actual block size of the target exceeds "
"this default, the backup may be unusable",
BACKUP_CLUSTER_SIZE_DEFAULT);
job->cluster_size = BACKUP_CLUSTER_SIZE_DEFAULT;
} else if (ret < 0 && !target->backing) {
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
error_setg_errno(errp, -ret,
"Couldn't determine the cluster size of the target image, "
"which has no backing file");
error_append_hint(errp,
"Aborting, since this may create an unusable destination image\n");
goto error;
} else if (ret < 0 && target->backing) {
/* Not fatal; just trudge on ahead. */
job->cluster_size = BACKUP_CLUSTER_SIZE_DEFAULT;
} else {
job->cluster_size = MAX(BACKUP_CLUSTER_SIZE_DEFAULT, bdi.cluster_size);
}
job->use_copy_range = true;
job->copy_range_size = MIN_NON_ZERO(blk_get_max_transfer(job->common.blk),
blk_get_max_transfer(job->target));
job->copy_range_size = MAX(job->cluster_size,
QEMU_ALIGN_UP(job->copy_range_size,
job->cluster_size));
block/backup: avoid copying less than full target clusters During incremental backups, if the target has a cluster size that is larger than the backup cluster size and we are backing up to a target that cannot (for whichever reason) pull clusters up from a backing image, we may inadvertantly create unusable incremental backup images. For example: If the bitmap tracks changes at a 64KB granularity and we transmit 64KB of data at a time but the target uses a 128KB cluster size, it is possible that only half of a target cluster will be recognized as dirty by the backup block job. When the cluster is allocated on the target image but only half populated with data, we lose the ability to distinguish between zero padding and uninitialized data. This does not happen if the target image has a backing file that points to the last known good backup. Even if we have a backing file, though, it's likely going to be faster to just buffer the redundant data ourselves from the live image than fetching it from the backing file, so let's just always round up to the target granularity. The same logic applies to backup modes top, none, and full. Copying fractional clusters without the guarantee of COW is dangerous, but even if we can rely on COW, it's likely better to just re-copy the data. Reported-by: Fam Zheng <famz@redhat.com> Signed-off-by: John Snow <jsnow@redhat.com> Reviewed-by: Fam Zheng <famz@redhat.com> Message-id: 1456433911-24718-3-git-send-email-jsnow@redhat.com Signed-off-by: Jeff Cody <jcody@redhat.com>
2016-02-25 21:58:30 +01:00
/* Required permissions are already taken with target's blk_new() */
block_job_add_bdrv(&job->common, "target", target, 0, BLK_PERM_ALL,
&error_abort);
job->len = len;
return &job->common;
error:
if (sync_bitmap) {
bdrv_reclaim_dirty_bitmap(bs, sync_bitmap, NULL);
}
if (job) {
backup_clean(&job->common.job);
job_early_fail(&job->common.job);
}
return NULL;
block: add basic backup support to block driver backup_start() creates a block job that copies a point-in-time snapshot of a block device to a target block device. We call backup_do_cow() for each write during backup. That function reads the original data from the block device before it gets overwritten. The data is then written to the target device. Currently backup cluster size is hardcoded to 65536 bytes. [I made a number of changes to Dietmar's original patch and folded them in to make code review easy. Here is the full list: * Drop BackupDumpFunc interface in favor of a target block device * Detect zero clusters with buffer_is_zero() and use bdrv_co_write_zeroes() * Use 0 delay instead of 1us, like other block jobs * Unify creation/start functions into backup_start() * Simplify cleanup, free bitmap in backup_run() instead of cb * function * Use HBitmap to avoid duplicating bitmap code * Use bdrv_getlength() instead of accessing ->total_sectors * directly * Delete the backup.h header file, it is no longer necessary * Move ./backup.c to block/backup.c * Remove #ifdefed out code * Coding style and whitespace cleanups * Use bdrv_add_before_write_notifier() instead of blockjob-specific hooks * Keep our own in-flight CowRequest list instead of using block.c tracked requests. This means a little code duplication but is much simpler than trying to share the tracked requests list and use the backup block size. * Add on_source_error and on_target_error error handling. * Use trace events instead of DPRINTF() -- stefanha] Signed-off-by: Dietmar Maurer <dietmar@proxmox.com> Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com> Signed-off-by: Kevin Wolf <kwolf@redhat.com>
2013-06-24 17:13:11 +02:00
}