SYS-OSS/FRET-qemu
4
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

bitmaps patches for 2020-07-27

Browse Source 
- Improve handling of various post-copy bitmap migration scenarios. A lost
 bitmap should merely mean that the next backup must be full rather than
 incremental, rather than abruptly breaking the entire guest migration.
 - Associated iotest improvements
 -----BEGIN PGP SIGNATURE-----
 
 iQEzBAABCAAdFiEEccLMIrHEYCkn0vOqp6FrSiUnQ2oFAl8fPRkACgkQp6FrSiUn
 Q2qanQf/dRTrqZ7/hs8aENySf44o0dBzOLZr+FBcrqEj2sd0c6jPzV2X5CVtnA1v
 gBgKJJGLpti3mSeNQDbaXZIQrsesBAuxvJsc6vZ9npDCdMYnK/qPE3Zfw1bx12qR
 cb39ba28P4izgs216h92ZACtUewnvjkxyJgN7zfmCJdNcwZINMUItAS183tSbQjn
 n39Wb7a+umsRgV9HQv/6cXlQIPqFMyAOl5kkzV3evuw7EBoHFnNq4cjPrUnjkqiD
 xf2pcSomaedYd37SpvoH57JxfL3z/90OBcuXhFvbqFk4FgQ63rJ32nRve2ZbIDI0
 XPbohnYjYoFv6Xs/jtTzctZCbZ+jTg==
 =1dmz
 -----END PGP SIGNATURE-----

Merge remote-tracking branch 'remotes/ericb/tags/pull-bitmaps-2020-07-27' into staging

bitmaps patches for 2020-07-27

- Improve handling of various post-copy bitmap migration scenarios. A lost
bitmap should merely mean that the next backup must be full rather than
incremental, rather than abruptly breaking the entire guest migration.
- Associated iotest improvements

# gpg: Signature made Mon 27 Jul 2020 21:46:17 BST
# gpg:                using RSA key 71C2CC22B1C4602927D2F3AAA7A16B4A2527436A
# gpg: Good signature from "Eric Blake <eblake@redhat.com>" [full]
# gpg:                 aka "Eric Blake (Free Software Programmer) <ebb9@byu.net>" [full]
# gpg:                 aka "[jpeg image of size 6874]" [full]
# Primary key fingerprint: 71C2 CC22 B1C4 6029 27D2  F3AA A7A1 6B4A 2527 436A

* remotes/ericb/tags/pull-bitmaps-2020-07-27: (24 commits)
  migration: Fix typos in bitmap migration comments
  iotests: Adjust which migration tests are quick
  qemu-iotests/199: add source-killed case to bitmaps postcopy
  qemu-iotests/199: add early shutdown case to bitmaps postcopy
  qemu-iotests/199: check persistent bitmaps
  qemu-iotests/199: prepare for new test-cases addition
  migration/savevm: don't worry if bitmap migration postcopy failed
  migration/block-dirty-bitmap: cancel migration on shutdown
  migration/block-dirty-bitmap: relax error handling in incoming part
  migration/block-dirty-bitmap: keep bitmap state for all bitmaps
  migration/block-dirty-bitmap: simplify dirty_bitmap_load_complete
  migration/block-dirty-bitmap: rename finish_lock to just lock
  migration/block-dirty-bitmap: refactor state global variables
  migration/block-dirty-bitmap: move mutex init to dirty_bitmap_mig_init
  migration/block-dirty-bitmap: rename dirty_bitmap_mig_cleanup
  migration/block-dirty-bitmap: rename state structure types
  migration/block-dirty-bitmap: fix dirty_bitmap_mig_before_vm_start
  qemu-iotests/199: increase postcopy period
  qemu-iotests/199: change discard patterns
  qemu-iotests/199: improve performance: set bitmap by discard
  ...

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
This commit is contained in:
Peter Maydell 2020-07-28 14:38:17 +01:00
commit 2649915121
9 changed files with 555 additions and 244 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
block/qcow2.c
View File

@ -66,7 +66,7 @@ typedef struct {
} QEMU_PACKED QCowExtension; } QEMU_PACKED QCowExtension;
#define QCOW2_EXT_MAGIC_END 0 #define QCOW2_EXT_MAGIC_END 0
#define QCOW2_EXT_MAGIC_BACKING_FORMAT 0xE2792ACA #define QCOW2_EXT_MAGIC_BACKING_FORMAT 0xe2792aca
#define QCOW2_EXT_MAGIC_FEATURE_TABLE 0x6803f857 #define QCOW2_EXT_MAGIC_FEATURE_TABLE 0x6803f857
#define QCOW2_EXT_MAGIC_CRYPTO_HEADER 0x0537be77 #define QCOW2_EXT_MAGIC_CRYPTO_HEADER 0x0537be77
#define QCOW2_EXT_MAGIC_BITMAPS 0x23852875 #define QCOW2_EXT_MAGIC_BITMAPS 0x23852875

2
docs/interop/qcow2.txt
View File

@ -231,7 +231,7 @@ be stored. Each extension has a structure like the following:
Byte 0 - 3: Header extension type: Byte 0 - 3: Header extension type:
0x00000000 - End of the header extension area 0x00000000 - End of the header extension area
0xE2792ACA - Backing file format name string 0xe2792aca - Backing file format name string
0x6803f857 - Feature name table 0x6803f857 - Feature name table
0x23852875 - Bitmaps extension 0x23852875 - Bitmaps extension
0x0537be77 - Full disk encryption header pointer 0x0537be77 - Full disk encryption header pointer

456
migration/block-dirty-bitmap.c
View File

@ -97,26 +97,28 @@
#define DIRTY_BITMAP_MIG_START_FLAG_ENABLED 0x01 #define DIRTY_BITMAP_MIG_START_FLAG_ENABLED 0x01
#define DIRTY_BITMAP_MIG_START_FLAG_PERSISTENT 0x02 #define DIRTY_BITMAP_MIG_START_FLAG_PERSISTENT 0x02
/* 0x04 was "AUTOLOAD" flags on elder versions, no it is ignored */ /* 0x04 was "AUTOLOAD" flags on older versions, now it is ignored */
#define DIRTY_BITMAP_MIG_START_FLAG_RESERVED_MASK 0xf8 #define DIRTY_BITMAP_MIG_START_FLAG_RESERVED_MASK 0xf8
typedef struct DirtyBitmapMigBitmapState { /* State of one bitmap during save process */
typedef struct SaveBitmapState {
/* Written during setup phase. */ /* Written during setup phase. */
BlockDriverState *bs; BlockDriverState *bs;
const char *node_name; const char *node_name;
BdrvDirtyBitmap *bitmap; BdrvDirtyBitmap *bitmap;
uint64_t total_sectors; uint64_t total_sectors;
uint64_t sectors_per_chunk; uint64_t sectors_per_chunk;
QSIMPLEQ_ENTRY(DirtyBitmapMigBitmapState) entry; QSIMPLEQ_ENTRY(SaveBitmapState) entry;
uint8_t flags; uint8_t flags;
/* For bulk phase. */ /* For bulk phase. */
bool bulk_completed; bool bulk_completed;
uint64_t cur_sector; uint64_t cur_sector;
} DirtyBitmapMigBitmapState; } SaveBitmapState;
typedef struct DirtyBitmapMigState { /* State of the dirty bitmap migration (DBM) during save process */
QSIMPLEQ_HEAD(, DirtyBitmapMigBitmapState) dbms_list; typedef struct DBMSaveState {
QSIMPLEQ_HEAD(, SaveBitmapState) dbms_list;
bool bulk_completed; bool bulk_completed;
bool no_bitmaps; bool no_bitmaps;
@ -124,30 +126,44 @@ typedef struct DirtyBitmapMigState {
/* for send_bitmap_bits() */ /* for send_bitmap_bits() */
BlockDriverState *prev_bs; BlockDriverState *prev_bs;
BdrvDirtyBitmap *prev_bitmap; BdrvDirtyBitmap *prev_bitmap;
} DirtyBitmapMigState; } DBMSaveState;
typedef struct DirtyBitmapLoadState { typedef struct LoadBitmapState {
BlockDriverState *bs;
BdrvDirtyBitmap *bitmap;
bool migrated;
bool enabled;
} LoadBitmapState;
/* State of the dirty bitmap migration (DBM) during load process */
typedef struct DBMLoadState {
uint32_t flags; uint32_t flags;
char node_name[256]; char node_name[256];
char bitmap_name[256]; char bitmap_name[256];
BlockDriverState *bs; BlockDriverState *bs;
BdrvDirtyBitmap *bitmap; BdrvDirtyBitmap *bitmap;
} DirtyBitmapLoadState;
static DirtyBitmapMigState dirty_bitmap_mig_state; bool before_vm_start_handled; /* set in dirty_bitmap_mig_before_vm_start */
typedef struct DirtyBitmapLoadBitmapState { /*
BlockDriverState *bs; * cancelled
BdrvDirtyBitmap *bitmap; * Incoming migration is cancelled for some reason. That means that we
bool migrated; * still should read our chunks from migration stream, to not affect other
} DirtyBitmapLoadBitmapState; * migration objects (like RAM), but just ignore them and do not touch any
static GSList *enabled_bitmaps; * bitmaps or nodes.
QemuMutex finish_lock; */
bool cancelled;
void init_dirty_bitmap_incoming_migration(void) GSList *bitmaps;
{ QemuMutex lock; /* protect bitmaps */
qemu_mutex_init(&finish_lock); } DBMLoadState;
}
typedef struct DBMState {
DBMSaveState save;
DBMLoadState load;
} DBMState;
static DBMState dbm_state;
static uint32_t qemu_get_bitmap_flags(QEMUFile *f) static uint32_t qemu_get_bitmap_flags(QEMUFile *f)
{ {
@ -164,27 +180,27 @@ static uint32_t qemu_get_bitmap_flags(QEMUFile *f)
static void qemu_put_bitmap_flags(QEMUFile *f, uint32_t flags) static void qemu_put_bitmap_flags(QEMUFile *f, uint32_t flags)
{ {
/* The code currently do not send flags more than one byte */ /* The code currently does not send flags as more than one byte */
assert(!(flags & (0xffffff00 | DIRTY_BITMAP_MIG_EXTRA_FLAGS))); assert(!(flags & (0xffffff00 | DIRTY_BITMAP_MIG_EXTRA_FLAGS)));
qemu_put_byte(f, flags); qemu_put_byte(f, flags);
} }
static void send_bitmap_header(QEMUFile *f, DirtyBitmapMigBitmapState *dbms, static void send_bitmap_header(QEMUFile *f, DBMSaveState *s,
uint32_t additional_flags) SaveBitmapState *dbms, uint32_t additional_flags)
{ {
BlockDriverState *bs = dbms->bs; BlockDriverState *bs = dbms->bs;
BdrvDirtyBitmap *bitmap = dbms->bitmap; BdrvDirtyBitmap *bitmap = dbms->bitmap;
uint32_t flags = additional_flags; uint32_t flags = additional_flags;
trace_send_bitmap_header_enter(); trace_send_bitmap_header_enter();
if (bs != dirty_bitmap_mig_state.prev_bs) { if (bs != s->prev_bs) {
dirty_bitmap_mig_state.prev_bs = bs; s->prev_bs = bs;
flags |= DIRTY_BITMAP_MIG_FLAG_DEVICE_NAME; flags |= DIRTY_BITMAP_MIG_FLAG_DEVICE_NAME;
} }
if (bitmap != dirty_bitmap_mig_state.prev_bitmap) { if (bitmap != s->prev_bitmap) {
dirty_bitmap_mig_state.prev_bitmap = bitmap; s->prev_bitmap = bitmap;
flags |= DIRTY_BITMAP_MIG_FLAG_BITMAP_NAME; flags |= DIRTY_BITMAP_MIG_FLAG_BITMAP_NAME;
} }
@ -199,19 +215,22 @@ static void send_bitmap_header(QEMUFile *f, DirtyBitmapMigBitmapState *dbms,
} }
} }
static void send_bitmap_start(QEMUFile *f, DirtyBitmapMigBitmapState *dbms) static void send_bitmap_start(QEMUFile *f, DBMSaveState *s,
SaveBitmapState *dbms)
{ {
send_bitmap_header(f, dbms, DIRTY_BITMAP_MIG_FLAG_START); send_bitmap_header(f, s, dbms, DIRTY_BITMAP_MIG_FLAG_START);
qemu_put_be32(f, bdrv_dirty_bitmap_granularity(dbms->bitmap)); qemu_put_be32(f, bdrv_dirty_bitmap_granularity(dbms->bitmap));
qemu_put_byte(f, dbms->flags); qemu_put_byte(f, dbms->flags);
} }
static void send_bitmap_complete(QEMUFile *f, DirtyBitmapMigBitmapState *dbms) static void send_bitmap_complete(QEMUFile *f, DBMSaveState *s,
SaveBitmapState *dbms)
{ {
send_bitmap_header(f, dbms, DIRTY_BITMAP_MIG_FLAG_COMPLETE); send_bitmap_header(f, s, dbms, DIRTY_BITMAP_MIG_FLAG_COMPLETE);
} }
static void send_bitmap_bits(QEMUFile *f, DirtyBitmapMigBitmapState *dbms, static void send_bitmap_bits(QEMUFile *f, DBMSaveState *s,
SaveBitmapState *dbms,
uint64_t start_sector, uint32_t nr_sectors) uint64_t start_sector, uint32_t nr_sectors)
{ {
/* align for buffer_is_zero() */ /* align for buffer_is_zero() */
@ -236,7 +255,7 @@ static void send_bitmap_bits(QEMUFile *f, DirtyBitmapMigBitmapState *dbms,
trace_send_bitmap_bits(flags, start_sector, nr_sectors, buf_size); trace_send_bitmap_bits(flags, start_sector, nr_sectors, buf_size);
send_bitmap_header(f, dbms, flags); send_bitmap_header(f, s, dbms, flags);
qemu_put_be64(f, start_sector); qemu_put_be64(f, start_sector);
qemu_put_be32(f, nr_sectors); qemu_put_be32(f, nr_sectors);
@ -255,12 +274,12 @@ static void send_bitmap_bits(QEMUFile *f, DirtyBitmapMigBitmapState *dbms,
} }
/* Called with iothread lock taken. */ /* Called with iothread lock taken. */
static void dirty_bitmap_mig_cleanup(void) static void dirty_bitmap_do_save_cleanup(DBMSaveState *s)
{ {
DirtyBitmapMigBitmapState *dbms; SaveBitmapState *dbms;
while ((dbms = QSIMPLEQ_FIRST(&dirty_bitmap_mig_state.dbms_list)) != NULL) { while ((dbms = QSIMPLEQ_FIRST(&s->dbms_list)) != NULL) {
QSIMPLEQ_REMOVE_HEAD(&dirty_bitmap_mig_state.dbms_list, entry); QSIMPLEQ_REMOVE_HEAD(&s->dbms_list, entry);
bdrv_dirty_bitmap_set_busy(dbms->bitmap, false); bdrv_dirty_bitmap_set_busy(dbms->bitmap, false);
bdrv_unref(dbms->bs); bdrv_unref(dbms->bs);
g_free(dbms); g_free(dbms);
@ -268,10 +287,11 @@ static void dirty_bitmap_mig_cleanup(void)
} }
/* Called with iothread lock taken. */ /* Called with iothread lock taken. */
static int add_bitmaps_to_list(BlockDriverState *bs, const char *bs_name) static int add_bitmaps_to_list(DBMSaveState *s, BlockDriverState *bs,
const char *bs_name)
{ {
BdrvDirtyBitmap *bitmap; BdrvDirtyBitmap *bitmap;
DirtyBitmapMigBitmapState *dbms; SaveBitmapState *dbms;
Error *local_err = NULL; Error *local_err = NULL;
FOR_EACH_DIRTY_BITMAP(bs, bitmap) { FOR_EACH_DIRTY_BITMAP(bs, bitmap) {
@ -309,7 +329,7 @@ static int add_bitmaps_to_list(BlockDriverState *bs, const char *bs_name)
bdrv_ref(bs); bdrv_ref(bs);
bdrv_dirty_bitmap_set_busy(bitmap, true); bdrv_dirty_bitmap_set_busy(bitmap, true);
dbms = g_new0(DirtyBitmapMigBitmapState, 1); dbms = g_new0(SaveBitmapState, 1);
dbms->bs = bs; dbms->bs = bs;
dbms->node_name = bs_name; dbms->node_name = bs_name;
dbms->bitmap = bitmap; dbms->bitmap = bitmap;
@ -323,25 +343,24 @@ static int add_bitmaps_to_list(BlockDriverState *bs, const char *bs_name)
dbms->flags |= DIRTY_BITMAP_MIG_START_FLAG_PERSISTENT; dbms->flags |= DIRTY_BITMAP_MIG_START_FLAG_PERSISTENT;
} }
QSIMPLEQ_INSERT_TAIL(&dirty_bitmap_mig_state.dbms_list, QSIMPLEQ_INSERT_TAIL(&s->dbms_list, dbms, entry);
dbms, entry);
} }
return 0; return 0;
} }
/* Called with iothread lock taken. */ /* Called with iothread lock taken. */
static int init_dirty_bitmap_migration(void) static int init_dirty_bitmap_migration(DBMSaveState *s)
{ {
BlockDriverState *bs; BlockDriverState *bs;
DirtyBitmapMigBitmapState *dbms; SaveBitmapState *dbms;
GHashTable *handled_by_blk = g_hash_table_new(NULL, NULL); GHashTable *handled_by_blk = g_hash_table_new(NULL, NULL);
BlockBackend *blk; BlockBackend *blk;
dirty_bitmap_mig_state.bulk_completed = false; s->bulk_completed = false;
dirty_bitmap_mig_state.prev_bs = NULL; s->prev_bs = NULL;
dirty_bitmap_mig_state.prev_bitmap = NULL; s->prev_bitmap = NULL;
dirty_bitmap_mig_state.no_bitmaps = false; s->no_bitmaps = false;
/* /*
* Use blockdevice name for direct (or filtered) children of named block * Use blockdevice name for direct (or filtered) children of named block
@ -370,7 +389,7 @@ static int init_dirty_bitmap_migration(void)
} }
if (bs && bs->drv && !bs->drv->is_filter) { if (bs && bs->drv && !bs->drv->is_filter) {
if (add_bitmaps_to_list(bs, name)) { if (add_bitmaps_to_list(s, bs, name)) {
goto fail; goto fail;
} }
g_hash_table_add(handled_by_blk, bs); g_hash_table_add(handled_by_blk, bs);
@ -382,18 +401,18 @@ static int init_dirty_bitmap_migration(void)
continue; continue;
} }
if (add_bitmaps_to_list(bs, bdrv_get_node_name(bs))) { if (add_bitmaps_to_list(s, bs, bdrv_get_node_name(bs))) {
goto fail; goto fail;
} }
} }
/* unset migration flags here, to not roll back it */ /* unset migration flags here, to not roll back it */
QSIMPLEQ_FOREACH(dbms, &dirty_bitmap_mig_state.dbms_list, entry) { QSIMPLEQ_FOREACH(dbms, &s->dbms_list, entry) {
bdrv_dirty_bitmap_skip_store(dbms->bitmap, true); bdrv_dirty_bitmap_skip_store(dbms->bitmap, true);
} }
if (QSIMPLEQ_EMPTY(&dirty_bitmap_mig_state.dbms_list)) { if (QSIMPLEQ_EMPTY(&s->dbms_list)) {
dirty_bitmap_mig_state.no_bitmaps = true; s->no_bitmaps = true;
} }
g_hash_table_destroy(handled_by_blk); g_hash_table_destroy(handled_by_blk);
@ -402,18 +421,19 @@ static int init_dirty_bitmap_migration(void)
fail: fail:
g_hash_table_destroy(handled_by_blk); g_hash_table_destroy(handled_by_blk);
dirty_bitmap_mig_cleanup(); dirty_bitmap_do_save_cleanup(s);
return -1; return -1;
} }
/* Called with no lock taken. */ /* Called with no lock taken. */
static void bulk_phase_send_chunk(QEMUFile *f, DirtyBitmapMigBitmapState *dbms) static void bulk_phase_send_chunk(QEMUFile *f, DBMSaveState *s,
SaveBitmapState *dbms)
{ {
uint32_t nr_sectors = MIN(dbms->total_sectors - dbms->cur_sector, uint32_t nr_sectors = MIN(dbms->total_sectors - dbms->cur_sector,
dbms->sectors_per_chunk); dbms->sectors_per_chunk);
send_bitmap_bits(f, dbms, dbms->cur_sector, nr_sectors); send_bitmap_bits(f, s, dbms, dbms->cur_sector, nr_sectors);
dbms->cur_sector += nr_sectors; dbms->cur_sector += nr_sectors;
if (dbms->cur_sector >= dbms->total_sectors) { if (dbms->cur_sector >= dbms->total_sectors) {
@ -422,61 +442,66 @@ static void bulk_phase_send_chunk(QEMUFile *f, DirtyBitmapMigBitmapState *dbms)
} }
/* Called with no lock taken. */ /* Called with no lock taken. */
static void bulk_phase(QEMUFile *f, bool limit) static void bulk_phase(QEMUFile *f, DBMSaveState *s, bool limit)
{ {
DirtyBitmapMigBitmapState *dbms; SaveBitmapState *dbms;
QSIMPLEQ_FOREACH(dbms, &dirty_bitmap_mig_state.dbms_list, entry) { QSIMPLEQ_FOREACH(dbms, &s->dbms_list, entry) {
while (!dbms->bulk_completed) { while (!dbms->bulk_completed) {
bulk_phase_send_chunk(f, dbms); bulk_phase_send_chunk(f, s, dbms);
if (limit && qemu_file_rate_limit(f)) { if (limit && qemu_file_rate_limit(f)) {
return; return;
} }
} }
} }
dirty_bitmap_mig_state.bulk_completed = true; s->bulk_completed = true;
} }
/* for SaveVMHandlers */ /* for SaveVMHandlers */
static void dirty_bitmap_save_cleanup(void *opaque) static void dirty_bitmap_save_cleanup(void *opaque)
{ {
dirty_bitmap_mig_cleanup(); DBMSaveState *s = &((DBMState *)opaque)->save;
dirty_bitmap_do_save_cleanup(s);
} }
static int dirty_bitmap_save_iterate(QEMUFile *f, void *opaque) static int dirty_bitmap_save_iterate(QEMUFile *f, void *opaque)
{ {
DBMSaveState *s = &((DBMState *)opaque)->save;
trace_dirty_bitmap_save_iterate(migration_in_postcopy()); trace_dirty_bitmap_save_iterate(migration_in_postcopy());
if (migration_in_postcopy() && !dirty_bitmap_mig_state.bulk_completed) { if (migration_in_postcopy() && !s->bulk_completed) {
bulk_phase(f, true); bulk_phase(f, s, true);
} }
qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS); qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS);
return dirty_bitmap_mig_state.bulk_completed; return s->bulk_completed;
} }
/* Called with iothread lock taken. */ /* Called with iothread lock taken. */
static int dirty_bitmap_save_complete(QEMUFile *f, void *opaque) static int dirty_bitmap_save_complete(QEMUFile *f, void *opaque)
{ {
DirtyBitmapMigBitmapState *dbms; DBMSaveState *s = &((DBMState *)opaque)->save;
SaveBitmapState *dbms;
trace_dirty_bitmap_save_complete_enter(); trace_dirty_bitmap_save_complete_enter();
if (!dirty_bitmap_mig_state.bulk_completed) { if (!s->bulk_completed) {
bulk_phase(f, false); bulk_phase(f, s, false);
} }
QSIMPLEQ_FOREACH(dbms, &dirty_bitmap_mig_state.dbms_list, entry) { QSIMPLEQ_FOREACH(dbms, &s->dbms_list, entry) {
send_bitmap_complete(f, dbms); send_bitmap_complete(f, s, dbms);
} }
qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS); qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS);
trace_dirty_bitmap_save_complete_finish(); trace_dirty_bitmap_save_complete_finish();
dirty_bitmap_mig_cleanup(); dirty_bitmap_save_cleanup(opaque);
return 0; return 0;
} }
@ -486,12 +511,13 @@ static void dirty_bitmap_save_pending(QEMUFile *f, void *opaque,
uint64_t *res_compatible, uint64_t *res_compatible,
uint64_t *res_postcopy_only) uint64_t *res_postcopy_only)
{ {
DirtyBitmapMigBitmapState *dbms; DBMSaveState *s = &((DBMState *)opaque)->save;
SaveBitmapState *dbms;
uint64_t pending = 0; uint64_t pending = 0;
qemu_mutex_lock_iothread(); qemu_mutex_lock_iothread();
QSIMPLEQ_FOREACH(dbms, &dirty_bitmap_mig_state.dbms_list, entry) { QSIMPLEQ_FOREACH(dbms, &s->dbms_list, entry) {
uint64_t gran = bdrv_dirty_bitmap_granularity(dbms->bitmap); uint64_t gran = bdrv_dirty_bitmap_granularity(dbms->bitmap);
uint64_t sectors = dbms->bulk_completed ? 0 : uint64_t sectors = dbms->bulk_completed ? 0 :
dbms->total_sectors - dbms->cur_sector; dbms->total_sectors - dbms->cur_sector;
@ -507,11 +533,16 @@ static void dirty_bitmap_save_pending(QEMUFile *f, void *opaque,
} }
/* First occurrence of this bitmap. It should be created if doesn't exist */ /* First occurrence of this bitmap. It should be created if doesn't exist */
static int dirty_bitmap_load_start(QEMUFile *f, DirtyBitmapLoadState *s) static int dirty_bitmap_load_start(QEMUFile *f, DBMLoadState *s)
{ {
Error *local_err = NULL; Error *local_err = NULL;
uint32_t granularity = qemu_get_be32(f); uint32_t granularity = qemu_get_be32(f);
uint8_t flags = qemu_get_byte(f); uint8_t flags = qemu_get_byte(f);
LoadBitmapState *b;
if (s->cancelled) {
return 0;
}
if (s->bitmap) { if (s->bitmap) {
error_report("Bitmap with the same name ('%s') already exists on " error_report("Bitmap with the same name ('%s') already exists on "
@ -538,90 +569,140 @@ static int dirty_bitmap_load_start(QEMUFile *f, DirtyBitmapLoadState *s)
bdrv_disable_dirty_bitmap(s->bitmap); bdrv_disable_dirty_bitmap(s->bitmap);
if (flags & DIRTY_BITMAP_MIG_START_FLAG_ENABLED) { if (flags & DIRTY_BITMAP_MIG_START_FLAG_ENABLED) {
DirtyBitmapLoadBitmapState *b;
bdrv_dirty_bitmap_create_successor(s->bitmap, &local_err); bdrv_dirty_bitmap_create_successor(s->bitmap, &local_err);
if (local_err) { if (local_err) {
error_report_err(local_err); error_report_err(local_err);
return -EINVAL; return -EINVAL;
} }
}
b = g_new(DirtyBitmapLoadBitmapState, 1); b = g_new(LoadBitmapState, 1);
b->bs = s->bs; b->bs = s->bs;
b->bitmap = s->bitmap; b->bitmap = s->bitmap;
b->migrated = false; b->migrated = false;
enabled_bitmaps = g_slist_prepend(enabled_bitmaps, b); b->enabled = flags & DIRTY_BITMAP_MIG_START_FLAG_ENABLED;
}
s->bitmaps = g_slist_prepend(s->bitmaps, b);
return 0; return 0;
} }
void dirty_bitmap_mig_before_vm_start(void) /*
* before_vm_start_handle_item
*
* g_slist_foreach helper
*
* item is LoadBitmapState*
* opaque is DBMLoadState*
*/
static void before_vm_start_handle_item(void *item, void *opaque)
{ {
GSList *item; DBMLoadState *s = opaque;
LoadBitmapState *b = item;
qemu_mutex_lock(&finish_lock);
for (item = enabled_bitmaps; item; item = g_slist_next(item)) {
DirtyBitmapLoadBitmapState *b = item->data;
if (b->enabled) {
if (b->migrated) { if (b->migrated) {
bdrv_enable_dirty_bitmap_locked(b->bitmap); bdrv_enable_dirty_bitmap(b->bitmap);
} else { } else {
bdrv_dirty_bitmap_enable_successor(b->bitmap); bdrv_dirty_bitmap_enable_successor(b->bitmap);
} }
g_free(b);
} }
g_slist_free(enabled_bitmaps); if (b->migrated) {
enabled_bitmaps = NULL; s->bitmaps = g_slist_remove(s->bitmaps, b);
g_free(b);
qemu_mutex_unlock(&finish_lock); }
} }
static void dirty_bitmap_load_complete(QEMUFile *f, DirtyBitmapLoadState *s) void dirty_bitmap_mig_before_vm_start(void)
{
DBMLoadState *s = &dbm_state.load;
qemu_mutex_lock(&s->lock);
assert(!s->before_vm_start_handled);
g_slist_foreach(s->bitmaps, before_vm_start_handle_item, s);
s->before_vm_start_handled = true;
qemu_mutex_unlock(&s->lock);
}
static void cancel_incoming_locked(DBMLoadState *s)
{
GSList *item;
if (s->cancelled) {
return;
}
s->cancelled = true;
s->bs = NULL;
s->bitmap = NULL;
/* Drop all unfinished bitmaps */
for (item = s->bitmaps; item; item = g_slist_next(item)) {
LoadBitmapState *b = item->data;
/*
* Bitmap must be unfinished, as finished bitmaps should already be
* removed from the list.
*/
assert(!s->before_vm_start_handled || !b->migrated);
if (bdrv_dirty_bitmap_has_successor(b->bitmap)) {
bdrv_reclaim_dirty_bitmap(b->bitmap, &error_abort);
}
bdrv_release_dirty_bitmap(b->bitmap);
}
g_slist_free_full(s->bitmaps, g_free);
s->bitmaps = NULL;
}
void dirty_bitmap_mig_cancel_outgoing(void)
{
dirty_bitmap_do_save_cleanup(&dbm_state.save);
}
void dirty_bitmap_mig_cancel_incoming(void)
{
DBMLoadState *s = &dbm_state.load;
qemu_mutex_lock(&s->lock);
cancel_incoming_locked(s);
qemu_mutex_unlock(&s->lock);
}
static void dirty_bitmap_load_complete(QEMUFile *f, DBMLoadState *s)
{ {
GSList *item; GSList *item;
trace_dirty_bitmap_load_complete(); trace_dirty_bitmap_load_complete();
if (s->cancelled) {
return;
}
bdrv_dirty_bitmap_deserialize_finish(s->bitmap); bdrv_dirty_bitmap_deserialize_finish(s->bitmap);
qemu_mutex_lock(&finish_lock); if (bdrv_dirty_bitmap_has_successor(s->bitmap)) {
bdrv_reclaim_dirty_bitmap(s->bitmap, &error_abort);
}
for (item = enabled_bitmaps; item; item = g_slist_next(item)) { for (item = s->bitmaps; item; item = g_slist_next(item)) {
DirtyBitmapLoadBitmapState *b = item->data; LoadBitmapState *b = item->data;
if (b->bitmap == s->bitmap) { if (b->bitmap == s->bitmap) {
b->migrated = true; b->migrated = true;
if (s->before_vm_start_handled) {
s->bitmaps = g_slist_remove(s->bitmaps, b);
g_free(b);
}
break; break;
} }
} }
if (bdrv_dirty_bitmap_has_successor(s->bitmap)) {
bdrv_dirty_bitmap_lock(s->bitmap);
if (enabled_bitmaps == NULL) {
/* in postcopy */
bdrv_reclaim_dirty_bitmap_locked(s->bitmap, &error_abort);
bdrv_enable_dirty_bitmap_locked(s->bitmap);
} else {
/* target not started, successor must be empty */
int64_t count = bdrv_get_dirty_count(s->bitmap);
BdrvDirtyBitmap *ret = bdrv_reclaim_dirty_bitmap_locked(s->bitmap,
NULL);
/* bdrv_reclaim_dirty_bitmap can fail only on no successor (it
* must be) or on merge fail, but merge can't fail when second
* bitmap is empty
*/
assert(ret == s->bitmap &&
count == bdrv_get_dirty_count(s->bitmap));
}
bdrv_dirty_bitmap_unlock(s->bitmap);
}
qemu_mutex_unlock(&finish_lock);
} }
static int dirty_bitmap_load_bits(QEMUFile *f, DirtyBitmapLoadState *s) static int dirty_bitmap_load_bits(QEMUFile *f, DBMLoadState *s)
{ {
uint64_t first_byte = qemu_get_be64(f) << BDRV_SECTOR_BITS; uint64_t first_byte = qemu_get_be64(f) << BDRV_SECTOR_BITS;
uint64_t nr_bytes = (uint64_t)qemu_get_be32(f) << BDRV_SECTOR_BITS; uint64_t nr_bytes = (uint64_t)qemu_get_be32(f) << BDRV_SECTOR_BITS;
@ -630,15 +711,46 @@ static int dirty_bitmap_load_bits(QEMUFile *f, DirtyBitmapLoadState *s)
if (s->flags & DIRTY_BITMAP_MIG_FLAG_ZEROES) { if (s->flags & DIRTY_BITMAP_MIG_FLAG_ZEROES) {
trace_dirty_bitmap_load_bits_zeroes(); trace_dirty_bitmap_load_bits_zeroes();
bdrv_dirty_bitmap_deserialize_zeroes(s->bitmap, first_byte, nr_bytes, if (!s->cancelled) {
false); bdrv_dirty_bitmap_deserialize_zeroes(s->bitmap, first_byte,
nr_bytes, false);
}
} else { } else {
size_t ret; size_t ret;
uint8_t *buf; g_autofree uint8_t *buf = NULL;
uint64_t buf_size = qemu_get_be64(f); uint64_t buf_size = qemu_get_be64(f);
uint64_t needed_size = uint64_t needed_size;
bdrv_dirty_bitmap_serialization_size(s->bitmap,
first_byte, nr_bytes); /*
* The actual check for buf_size is done a bit later. We can't do it in
* cancelled mode as we don't have the bitmap to check the constraints
* (so, we allocate a buffer and read prior to the check). On the other
* hand, we shouldn't blindly g_malloc the number from the stream.
* Actually one chunk should not be larger than CHUNK_SIZE. Let's allow
* a bit larger (which means that bitmap migration will fail anyway and
* the whole migration will most probably fail soon due to broken
* stream).
*/
if (buf_size > 10 * CHUNK_SIZE) {
error_report("Bitmap migration stream buffer allocation request "
"is too large");
return -EIO;
}
buf = g_malloc(buf_size);
ret = qemu_get_buffer(f, buf, buf_size);
if (ret != buf_size) {
error_report("Failed to read bitmap bits");
return -EIO;
}
if (s->cancelled) {
return 0;
}
needed_size = bdrv_dirty_bitmap_serialization_size(s->bitmap,
first_byte,
nr_bytes);
if (needed_size > buf_size || if (needed_size > buf_size ||
buf_size > QEMU_ALIGN_UP(needed_size, 4 * sizeof(long)) buf_size > QEMU_ALIGN_UP(needed_size, 4 * sizeof(long))
@ -647,26 +759,18 @@ static int dirty_bitmap_load_bits(QEMUFile *f, DirtyBitmapLoadState *s)
error_report("Migrated bitmap granularity doesn't " error_report("Migrated bitmap granularity doesn't "
"match the destination bitmap '%s' granularity", "match the destination bitmap '%s' granularity",
bdrv_dirty_bitmap_name(s->bitmap)); bdrv_dirty_bitmap_name(s->bitmap));
return -EINVAL; cancel_incoming_locked(s);
} return 0;
buf = g_malloc(buf_size);
ret = qemu_get_buffer(f, buf, buf_size);
if (ret != buf_size) {
error_report("Failed to read bitmap bits");
g_free(buf);
return -EIO;
} }
bdrv_dirty_bitmap_deserialize_part(s->bitmap, buf, first_byte, nr_bytes, bdrv_dirty_bitmap_deserialize_part(s->bitmap, buf, first_byte, nr_bytes,
false); false);
g_free(buf);
} }
return 0; return 0;
} }
static int dirty_bitmap_load_header(QEMUFile *f, DirtyBitmapLoadState *s) static int dirty_bitmap_load_header(QEMUFile *f, DBMLoadState *s)
{ {
Error *local_err = NULL; Error *local_err = NULL;
bool nothing; bool nothing;
@ -680,14 +784,16 @@ static int dirty_bitmap_load_header(QEMUFile *f, DirtyBitmapLoadState *s)
error_report("Unable to read node name string"); error_report("Unable to read node name string");
return -EINVAL; return -EINVAL;
} }
if (!s->cancelled) {
s->bs = bdrv_lookup_bs(s->node_name, s->node_name, &local_err); s->bs = bdrv_lookup_bs(s->node_name, s->node_name, &local_err);
if (!s->bs) { if (!s->bs) {
error_report_err(local_err); error_report_err(local_err);
return -EINVAL; cancel_incoming_locked(s);
} }
} else if (!s->bs && !nothing) { }
} else if (!s->bs && !nothing && !s->cancelled) {
error_report("Error: block device name is not set"); error_report("Error: block device name is not set");
return -EINVAL; cancel_incoming_locked(s);
} }
if (s->flags & DIRTY_BITMAP_MIG_FLAG_BITMAP_NAME) { if (s->flags & DIRTY_BITMAP_MIG_FLAG_BITMAP_NAME) {
@ -695,47 +801,66 @@ static int dirty_bitmap_load_header(QEMUFile *f, DirtyBitmapLoadState *s)
error_report("Unable to read bitmap name string"); error_report("Unable to read bitmap name string");
return -EINVAL; return -EINVAL;
} }
if (!s->cancelled) {
s->bitmap = bdrv_find_dirty_bitmap(s->bs, s->bitmap_name); s->bitmap = bdrv_find_dirty_bitmap(s->bs, s->bitmap_name);
/* bitmap may be NULL here, it wouldn't be an error if it is the /*
* first occurrence of the bitmap */ * bitmap may be NULL here, it wouldn't be an error if it is the
* first occurrence of the bitmap
*/
if (!s->bitmap && !(s->flags & DIRTY_BITMAP_MIG_FLAG_START)) { if (!s->bitmap && !(s->flags & DIRTY_BITMAP_MIG_FLAG_START)) {
error_report("Error: unknown dirty bitmap " error_report("Error: unknown dirty bitmap "
"'%s' for block device '%s'", "'%s' for block device '%s'",
s->bitmap_name, s->node_name); s->bitmap_name, s->node_name);
return -EINVAL; cancel_incoming_locked(s);
} }
} else if (!s->bitmap && !nothing) { }
} else if (!s->bitmap && !nothing && !s->cancelled) {
error_report("Error: block device name is not set"); error_report("Error: block device name is not set");
return -EINVAL; cancel_incoming_locked(s);
} }
return 0; return 0;
} }
/*
* dirty_bitmap_load
*
* Load sequence of dirty bitmap chunks. Return error only on fatal io stream
* violations. On other errors just cancel bitmaps incoming migration and return
* 0.
*
* Note, than when incoming bitmap migration is canceled, we still must read all
* our chunks (and just ignore them), to not affect other migration objects.
*/
static int dirty_bitmap_load(QEMUFile *f, void *opaque, int version_id) static int dirty_bitmap_load(QEMUFile *f, void *opaque, int version_id)
{ {
static DirtyBitmapLoadState s; DBMLoadState *s = &((DBMState *)opaque)->load;
int ret = 0; int ret = 0;
trace_dirty_bitmap_load_enter(); trace_dirty_bitmap_load_enter();
if (version_id != 1) { if (version_id != 1) {
QEMU_LOCK_GUARD(&s->lock);
cancel_incoming_locked(s);
return -EINVAL; return -EINVAL;
} }
do { do {
ret = dirty_bitmap_load_header(f, &s); QEMU_LOCK_GUARD(&s->lock);
ret = dirty_bitmap_load_header(f, s);
if (ret < 0) { if (ret < 0) {
cancel_incoming_locked(s);
return ret; return ret;
} }
if (s.flags & DIRTY_BITMAP_MIG_FLAG_START) { if (s->flags & DIRTY_BITMAP_MIG_FLAG_START) {
ret = dirty_bitmap_load_start(f, &s); ret = dirty_bitmap_load_start(f, s);
} else if (s.flags & DIRTY_BITMAP_MIG_FLAG_COMPLETE) { } else if (s->flags & DIRTY_BITMAP_MIG_FLAG_COMPLETE) {
dirty_bitmap_load_complete(f, &s); dirty_bitmap_load_complete(f, s);
} else if (s.flags & DIRTY_BITMAP_MIG_FLAG_BITS) { } else if (s->flags & DIRTY_BITMAP_MIG_FLAG_BITS) {
ret = dirty_bitmap_load_bits(f, &s); ret = dirty_bitmap_load_bits(f, s);
} }
if (!ret) { if (!ret) {
@ -743,9 +868,10 @@ static int dirty_bitmap_load(QEMUFile *f, void *opaque, int version_id)
} }
if (ret) { if (ret) {
cancel_incoming_locked(s);
return ret; return ret;
} }
} while (!(s.flags & DIRTY_BITMAP_MIG_FLAG_EOS)); } while (!(s->flags & DIRTY_BITMAP_MIG_FLAG_EOS));
trace_dirty_bitmap_load_success(); trace_dirty_bitmap_load_success();
return 0; return 0;
@ -753,13 +879,14 @@ static int dirty_bitmap_load(QEMUFile *f, void *opaque, int version_id)
static int dirty_bitmap_save_setup(QEMUFile *f, void *opaque) static int dirty_bitmap_save_setup(QEMUFile *f, void *opaque)
{ {
DirtyBitmapMigBitmapState *dbms = NULL; DBMSaveState *s = &((DBMState *)opaque)->save;
if (init_dirty_bitmap_migration() < 0) { SaveBitmapState *dbms = NULL;
if (init_dirty_bitmap_migration(s) < 0) {
return -1; return -1;
} }
QSIMPLEQ_FOREACH(dbms, &dirty_bitmap_mig_state.dbms_list, entry) { QSIMPLEQ_FOREACH(dbms, &s->dbms_list, entry) {
send_bitmap_start(f, dbms); send_bitmap_start(f, s, dbms);
} }
qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS); qemu_put_bitmap_flags(f, DIRTY_BITMAP_MIG_FLAG_EOS);
@ -768,7 +895,9 @@ static int dirty_bitmap_save_setup(QEMUFile *f, void *opaque)
static bool dirty_bitmap_is_active(void *opaque) static bool dirty_bitmap_is_active(void *opaque)
{ {
return migrate_dirty_bitmaps() && !dirty_bitmap_mig_state.no_bitmaps; DBMSaveState *s = &((DBMState *)opaque)->save;
return migrate_dirty_bitmaps() && !s->no_bitmaps;
} }
static bool dirty_bitmap_is_active_iterate(void *opaque) static bool dirty_bitmap_is_active_iterate(void *opaque)
@ -796,9 +925,10 @@ static SaveVMHandlers savevm_dirty_bitmap_handlers = {
void dirty_bitmap_mig_init(void) void dirty_bitmap_mig_init(void)
{ {
QSIMPLEQ_INIT(&dirty_bitmap_mig_state.dbms_list); QSIMPLEQ_INIT(&dbm_state.save.dbms_list);
qemu_mutex_init(&dbm_state.load.lock);
register_savevm_live("dirty-bitmap", 0, 1, register_savevm_live("dirty-bitmap", 0, 1,
&savevm_dirty_bitmap_handlers, &savevm_dirty_bitmap_handlers,
&dirty_bitmap_mig_state); &dbm_state);
} }

15
migration/migration.c
View File

@ -165,8 +165,6 @@ void migration_object_init(void)
qemu_sem_init(&current_incoming->postcopy_pause_sem_dst, 0); qemu_sem_init(&current_incoming->postcopy_pause_sem_dst, 0);
qemu_sem_init(&current_incoming->postcopy_pause_sem_fault, 0); qemu_sem_init(&current_incoming->postcopy_pause_sem_fault, 0);
init_dirty_bitmap_incoming_migration();
if (!migration_object_check(current_migration, &err)) { if (!migration_object_check(current_migration, &err)) {
error_report_err(err); error_report_err(err);
exit(1); exit(1);
@ -190,6 +188,19 @@ void migration_shutdown(void)
*/ */
migrate_fd_cancel(current_migration); migrate_fd_cancel(current_migration);
object_unref(OBJECT(current_migration)); object_unref(OBJECT(current_migration));
/*
* Cancel outgoing migration of dirty bitmaps. It should
* at least unref used block nodes.
*/
dirty_bitmap_mig_cancel_outgoing();
/*
* Cancel incoming migration of dirty bitmaps. Dirty bitmaps
* are non-critical data, and their loss never considered as
* something serious.
*/
dirty_bitmap_mig_cancel_incoming();
} }
/* For outgoing */ /* For outgoing */

3
migration/migration.h
View File

@ -335,7 +335,8 @@ void migrate_send_rp_recv_bitmap(MigrationIncomingState *mis,
void migrate_send_rp_resume_ack(MigrationIncomingState *mis, uint32_t value); void migrate_send_rp_resume_ack(MigrationIncomingState *mis, uint32_t value);
void dirty_bitmap_mig_before_vm_start(void); void dirty_bitmap_mig_before_vm_start(void);
void init_dirty_bitmap_incoming_migration(void); void dirty_bitmap_mig_cancel_outgoing(void);
void dirty_bitmap_mig_cancel_incoming(void);
void migrate_add_address(SocketAddress *address); void migrate_add_address(SocketAddress *address);
int foreach_not_ignored_block(RAMBlockIterFunc func, void *opaque); int foreach_not_ignored_block(RAMBlockIterFunc func, void *opaque);

33
migration/savevm.c
View File

@ -1813,6 +1813,9 @@ static void *postcopy_ram_listen_thread(void *opaque)
MigrationIncomingState *mis = migration_incoming_get_current(); MigrationIncomingState *mis = migration_incoming_get_current();
QEMUFile *f = mis->from_src_file; QEMUFile *f = mis->from_src_file;
int load_res; int load_res;
MigrationState *migr = migrate_get_current();
object_ref(OBJECT(migr));
migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE, migrate_set_state(&mis->state, MIGRATION_STATUS_ACTIVE,
MIGRATION_STATUS_POSTCOPY_ACTIVE); MIGRATION_STATUS_POSTCOPY_ACTIVE);
@ -1839,11 +1842,24 @@ static void *postcopy_ram_listen_thread(void *opaque)
trace_postcopy_ram_listen_thread_exit(); trace_postcopy_ram_listen_thread_exit();
if (load_res < 0) { if (load_res < 0) {
error_report("%s: loadvm failed: %d", __func__, load_res);
qemu_file_set_error(f, load_res); qemu_file_set_error(f, load_res);
dirty_bitmap_mig_cancel_incoming();
if (postcopy_state_get() == POSTCOPY_INCOMING_RUNNING &&
!migrate_postcopy_ram() && migrate_dirty_bitmaps())
{
error_report("%s: loadvm failed during postcopy: %d. All states "
"are migrated except dirty bitmaps. Some dirty "
"bitmaps may be lost, and present migrated dirty "
"bitmaps are correctly migrated and valid.",
__func__, load_res);
load_res = 0; /* prevent further exit() */
} else {
error_report("%s: loadvm failed: %d", __func__, load_res);
migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE, migrate_set_state(&mis->state, MIGRATION_STATUS_POSTCOPY_ACTIVE,
MIGRATION_STATUS_FAILED); MIGRATION_STATUS_FAILED);
} else { }
}
if (load_res >= 0) {
/* /*
* This looks good, but it's possible that the device loading in the * This looks good, but it's possible that the device loading in the
* main thread hasn't finished yet, and so we might not be in 'RUN' * main thread hasn't finished yet, and so we might not be in 'RUN'
@ -1879,6 +1895,8 @@ static void *postcopy_ram_listen_thread(void *opaque)
mis->have_listen_thread = false; mis->have_listen_thread = false;
postcopy_state_set(POSTCOPY_INCOMING_END); postcopy_state_set(POSTCOPY_INCOMING_END);
object_unref(OBJECT(migr));
return NULL; return NULL;
} }
@ -2437,6 +2455,8 @@ static bool postcopy_pause_incoming(MigrationIncomingState *mis)
{ {
trace_postcopy_pause_incoming(); trace_postcopy_pause_incoming();
assert(migrate_postcopy_ram());
/* Clear the triggered bit to allow one recovery */ /* Clear the triggered bit to allow one recovery */
mis->postcopy_recover_triggered = false; mis->postcopy_recover_triggered = false;
@ -2521,15 +2541,22 @@ out:
if (ret < 0) { if (ret < 0) {
qemu_file_set_error(f, ret); qemu_file_set_error(f, ret);
/* Cancel bitmaps incoming regardless of recovery */
dirty_bitmap_mig_cancel_incoming();
/* /*
* If we are during an active postcopy, then we pause instead * If we are during an active postcopy, then we pause instead
* of bail out to at least keep the VM's dirty data. Note * of bail out to at least keep the VM's dirty data. Note
* that POSTCOPY_INCOMING_LISTENING stage is still not enough, * that POSTCOPY_INCOMING_LISTENING stage is still not enough,
* during which we're still receiving device states and we * during which we're still receiving device states and we
* still haven't yet started the VM on destination. * still haven't yet started the VM on destination.
*
* Only RAM postcopy supports recovery. Still, if RAM postcopy is
* enabled, canceled bitmaps postcopy will not affect RAM postcopy
* recovering.
*/ */
if (postcopy_state_get() == POSTCOPY_INCOMING_RUNNING && if (postcopy_state_get() == POSTCOPY_INCOMING_RUNNING &&
postcopy_pause_incoming(mis)) { migrate_postcopy_ram() && postcopy_pause_incoming(mis)) {
/* Reset f to point to the newly created channel */ /* Reset f to point to the newly created channel */
f = mis->from_src_file; f = mis->from_src_file;
goto retry; goto retry;

252
tests/qemu-iotests/199
View File

@ -20,17 +20,76 @@
import os import os
import iotests import iotests
import time
from iotests import qemu_img from iotests import qemu_img
debug = False
disk_a = os.path.join(iotests.test_dir, 'disk_a') disk_a = os.path.join(iotests.test_dir, 'disk_a')
disk_b = os.path.join(iotests.test_dir, 'disk_b') disk_b = os.path.join(iotests.test_dir, 'disk_b')
size = '256G' size = '256G'
fifo = os.path.join(iotests.test_dir, 'mig_fifo') fifo = os.path.join(iotests.test_dir, 'mig_fifo')
class TestDirtyBitmapPostcopyMigration(iotests.QMPTestCase): granularity = 512
nb_bitmaps = 15
GiB = 1024 * 1024 * 1024
discards1 = (
(0, GiB),
(2 * GiB + 512 * 5, 512),
(3 * GiB + 512 * 5, 512),
(100 * GiB, GiB)
)
discards2 = (
(3 * GiB + 512 * 8, 512),
(4 * GiB + 512 * 8, 512),
(50 * GiB, GiB),
(100 * GiB + GiB // 2, GiB)
)
def apply_discards(vm, discards):
for d in discards:
vm.hmp_qemu_io('drive0', 'discard {} {}'.format(*d))
def event_seconds(event):
return event['timestamp']['seconds'] + \
event['timestamp']['microseconds'] / 1000000.0
def event_dist(e1, e2):
return event_seconds(e2) - event_seconds(e1)
def check_bitmaps(vm, count):
result = vm.qmp('query-block')
if count == 0:
assert 'dirty-bitmaps' not in result['return'][0]
else:
assert len(result['return'][0]['dirty-bitmaps']) == count
class TestDirtyBitmapPostcopyMigration(iotests.QMPTestCase):
def tearDown(self): def tearDown(self):
if debug:
self.vm_a_events += self.vm_a.get_qmp_events()
self.vm_b_events += self.vm_b.get_qmp_events()
for e in self.vm_a_events:
e['vm'] = 'SRC'
for e in self.vm_b_events:
e['vm'] = 'DST'
events = (self.vm_a_events + self.vm_b_events)
events = [(e['timestamp']['seconds'],
e['timestamp']['microseconds'],
e['vm'],
e['event'],
e.get('data', '')) for e in events]
for e in sorted(events):
print('{}.{:06} {} {} {}'.format(*e))
self.vm_a.shutdown() self.vm_a.shutdown()
self.vm_b.shutdown() self.vm_b.shutdown()
os.remove(disk_a) os.remove(disk_a)
@ -41,51 +100,66 @@ class TestDirtyBitmapPostcopyMigration(iotests.QMPTestCase):
os.mkfifo(fifo) os.mkfifo(fifo)
qemu_img('create', '-f', iotests.imgfmt, disk_a, size) qemu_img('create', '-f', iotests.imgfmt, disk_a, size)
qemu_img('create', '-f', iotests.imgfmt, disk_b, size) qemu_img('create', '-f', iotests.imgfmt, disk_b, size)
self.vm_a = iotests.VM(path_suffix='a').add_drive(disk_a) self.vm_a = iotests.VM(path_suffix='a').add_drive(disk_a,
self.vm_b = iotests.VM(path_suffix='b').add_drive(disk_b) 'discard=unmap')
self.vm_b = iotests.VM(path_suffix='b').add_drive(disk_b,
'discard=unmap')
self.vm_b.add_incoming("exec: cat '" + fifo + "'") self.vm_b.add_incoming("exec: cat '" + fifo + "'")
self.vm_a.launch() self.vm_a.launch()
self.vm_b.launch() self.vm_b.launch()
def test_postcopy(self): # collect received events for debug
write_size = 0x40000000 self.vm_a_events = []
granularity = 512 self.vm_b_events = []
chunk = 4096
def start_postcopy(self):
""" Run migration until RESUME event on target. Return this event. """
for i in range(nb_bitmaps):
result = self.vm_a.qmp('block-dirty-bitmap-add', node='drive0', result = self.vm_a.qmp('block-dirty-bitmap-add', node='drive0',
name='bitmap', granularity=granularity) name='bitmap{}'.format(i),
self.assert_qmp(result, 'return', {}); granularity=granularity,
persistent=True)
s = 0
while s < write_size:
self.vm_a.hmp_qemu_io('drive0', 'write %d %d' % (s, chunk))
s += 0x10000
s = 0x8000
while s < write_size:
self.vm_a.hmp_qemu_io('drive0', 'write %d %d' % (s, chunk))
s += 0x10000
result = self.vm_a.qmp('x-debug-block-dirty-bitmap-sha256',
node='drive0', name='bitmap')
sha256 = result['return']['sha256']
result = self.vm_a.qmp('block-dirty-bitmap-clear', node='drive0',
name='bitmap')
self.assert_qmp(result, 'return', {});
s = 0
while s < write_size:
self.vm_a.hmp_qemu_io('drive0', 'write %d %d' % (s, chunk))
s += 0x10000
bitmaps_cap = {'capability': 'dirty-bitmaps', 'state': True}
events_cap = {'capability': 'events', 'state': True}
result = self.vm_a.qmp('migrate-set-capabilities',
capabilities=[bitmaps_cap, events_cap])
self.assert_qmp(result, 'return', {}) self.assert_qmp(result, 'return', {})
result = self.vm_b.qmp('migrate-set-capabilities', result = self.vm_a.qmp('x-debug-block-dirty-bitmap-sha256',
capabilities=[bitmaps_cap]) node='drive0', name='bitmap0')
empty_sha256 = result['return']['sha256']
apply_discards(self.vm_a, discards1)
result = self.vm_a.qmp('x-debug-block-dirty-bitmap-sha256',
node='drive0', name='bitmap0')
self.discards1_sha256 = result['return']['sha256']
# Check, that updating the bitmap by discards works
assert self.discards1_sha256 != empty_sha256
# We want to calculate resulting sha256. Do it in bitmap0, so, disable
# other bitmaps
for i in range(1, nb_bitmaps):
result = self.vm_a.qmp('block-dirty-bitmap-disable', node='drive0',
name='bitmap{}'.format(i))
self.assert_qmp(result, 'return', {})
apply_discards(self.vm_a, discards2)
result = self.vm_a.qmp('x-debug-block-dirty-bitmap-sha256',
node='drive0', name='bitmap0')
self.all_discards_sha256 = result['return']['sha256']
# Now, enable some bitmaps, to be updated during migration
for i in range(2, nb_bitmaps, 2):
result = self.vm_a.qmp('block-dirty-bitmap-enable', node='drive0',
name='bitmap{}'.format(i))
self.assert_qmp(result, 'return', {})
caps = [{'capability': 'dirty-bitmaps', 'state': True},
{'capability': 'events', 'state': True}]
result = self.vm_a.qmp('migrate-set-capabilities', capabilities=caps)
self.assert_qmp(result, 'return', {})
result = self.vm_b.qmp('migrate-set-capabilities', capabilities=caps)
self.assert_qmp(result, 'return', {}) self.assert_qmp(result, 'return', {})
result = self.vm_a.qmp('migrate', uri='exec:cat>' + fifo) result = self.vm_a.qmp('migrate', uri='exec:cat>' + fifo)
@ -94,26 +168,94 @@ class TestDirtyBitmapPostcopyMigration(iotests.QMPTestCase):
result = self.vm_a.qmp('migrate-start-postcopy') result = self.vm_a.qmp('migrate-start-postcopy')
self.assert_qmp(result, 'return', {}) self.assert_qmp(result, 'return', {})
while True: event_resume = self.vm_b.event_wait('RESUME')
event = self.vm_a.event_wait('MIGRATION') self.vm_b_events.append(event_resume)
if event['data']['status'] == 'completed': return event_resume
break
s = 0x8000 def test_postcopy_success(self):
while s < write_size: event_resume = self.start_postcopy()
self.vm_b.hmp_qemu_io('drive0', 'write %d %d' % (s, chunk))
s += 0x10000
result = self.vm_b.qmp('query-block'); # enabled bitmaps should be updated
while len(result['return'][0]['dirty-bitmaps']) > 1: apply_discards(self.vm_b, discards2)
time.sleep(2)
result = self.vm_b.qmp('query-block');
match = {'data': {'status': 'completed'}}
event_complete = self.vm_b.event_wait('MIGRATION', match=match)
self.vm_b_events.append(event_complete)
# take queued event, should already been happened
event_stop = self.vm_a.event_wait('STOP')
self.vm_a_events.append(event_stop)
downtime = event_dist(event_stop, event_resume)
postcopy_time = event_dist(event_resume, event_complete)
assert downtime * 10 < postcopy_time
if debug:
print('downtime:', downtime)
print('postcopy_time:', postcopy_time)
# check that there are no bitmaps stored on source
self.vm_a_events += self.vm_a.get_qmp_events()
self.vm_a.shutdown()
self.vm_a.launch()
check_bitmaps(self.vm_a, 0)
# check that bitmaps are migrated and persistence works
check_bitmaps(self.vm_b, nb_bitmaps)
self.vm_b.shutdown()
# recreate vm_b, so there is no incoming option, which prevents
# loading bitmaps from disk
self.vm_b = iotests.VM(path_suffix='b').add_drive(disk_b)
self.vm_b.launch()
check_bitmaps(self.vm_b, nb_bitmaps)
# Check content of migrated bitmaps. Still, don't waste time checking
# every bitmap
for i in range(0, nb_bitmaps, 5):
result = self.vm_b.qmp('x-debug-block-dirty-bitmap-sha256', result = self.vm_b.qmp('x-debug-block-dirty-bitmap-sha256',
node='drive0', name='bitmap') node='drive0', name='bitmap{}'.format(i))
sha = self.discards1_sha256 if i % 2 else self.all_discards_sha256
self.assert_qmp(result, 'return/sha256', sha)
def test_early_shutdown_destination(self):
self.start_postcopy()
self.vm_b_events += self.vm_b.get_qmp_events()
self.vm_b.shutdown()
# recreate vm_b, so there is no incoming option, which prevents
# loading bitmaps from disk
self.vm_b = iotests.VM(path_suffix='b').add_drive(disk_b)
self.vm_b.launch()
check_bitmaps(self.vm_b, 0)
# Bitmaps will be lost if we just shutdown the vm, as they are marked
# to skip storing to disk when prepared for migration. And that's
# correct, as actual data may be modified in target vm, so we play
# safe.
# Still, this mark would be taken away if we do 'cont', and bitmaps
# become persistent again. (see iotest 169 for such behavior case)
result = self.vm_a.qmp('query-status')
assert not result['return']['running']
self.vm_a_events += self.vm_a.get_qmp_events()
self.vm_a.shutdown()
self.vm_a.launch()
check_bitmaps(self.vm_a, 0)
def test_early_kill_source(self):
self.start_postcopy()
self.vm_a_events = self.vm_a.get_qmp_events()
self.vm_a.kill()
self.vm_a.launch()
match = {'data': {'status': 'completed'}}
e_complete = self.vm_b.event_wait('MIGRATION', match=match)
self.vm_b_events.append(e_complete)
check_bitmaps(self.vm_a, 0)
check_bitmaps(self.vm_b, 0)
self.assert_qmp(result, 'return/sha256', sha256);
if __name__ == '__main__': if __name__ == '__main__':
iotests.main(supported_fmts=['qcow2'], supported_cache_modes=['none'], iotests.main(supported_fmts=['qcow2'])
supported_protocols=['file'])

4
tests/qemu-iotests/199.out
View File

@ -1,5 +1,5 @@
. ...
---------------------------------------------------------------------- ----------------------------------------------------------------------
Ran 1 tests Ran 3 tests
OK OK

12
tests/qemu-iotests/group
View File

@ -112,7 +112,7 @@
088 rw quick 088 rw quick
089 rw auto quick 089 rw auto quick
090 rw auto quick 090 rw auto quick
091 rw migration 091 rw migration quick
092 rw quick 092 rw quick
093 throttle 093 throttle
094 rw quick 094 rw quick
@ -186,7 +186,7 @@
162 quick 162 quick
163 rw 163 rw
165 rw quick 165 rw quick
169 rw quick migration 169 rw migration
170 rw auto quick 170 rw auto quick
171 rw quick 171 rw quick
172 auto 172 auto
@ -197,9 +197,9 @@
177 rw auto quick 177 rw auto quick
178 img 178 img
179 rw auto quick 179 rw auto quick
181 rw auto migration 181 rw auto migration quick
182 rw quick 182 rw quick
183 rw migration 183 rw migration quick
184 rw auto quick 184 rw auto quick
185 rw 185 rw
186 rw auto 186 rw auto
@ -216,9 +216,9 @@
198 rw 198 rw
199 rw migration 199 rw migration
200 rw 200 rw
201 rw migration 201 rw migration quick
202 rw quick 202 rw quick
203 rw auto migration 203 rw auto migration quick
204 rw quick 204 rw quick
205 rw quick 205 rw quick
206 rw 206 rw