* [Qemu-devel] [PATCH 00/12] Migration next v11
@ 2012-08-06 17:51 Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 01/12] Add migration capabilities Orit Wasserman
` (11 more replies)
0 siblings, 12 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
Changes from v10:
remove query-migrate-supported-capabilities commands.
Update documentation.
Replace total_time with total-time
fix other comments by Eric
Changes from v9:
rename query-migrtion-supported-capabilities to query-migrate-supported-capabilities
rename 'info migration_supported_capabilities' to 'info migrate_supported_capabilities'
renamse migrate_set_cachesize to migrate_set_cache_size.
Add missing state to qmp_query_migrate_supported_capabilities.
remove unused MigrationParameters.
fix query-migrate examples.
fix other comments by Luiz and Eric.
Changes from v8:
remove xbzrle from xbzrle-cache fields name.
fix other review comments by Eric
Changes from v7:
rename migrate-set-parameter to migrate-set-capabilities
rename query-migrate-parameters to query-migrate-capabilities
renamse query-migration-capabilities to query-migration-supported-capabilities
fix query_migration_supported_capabilties to return a list
add query-migrate-cache-size commands
fix other review comment by Luiz and Eric
Changes from v6:
fix memory leak in cache_init.
using size_t for cache_get_cache_pos.
fix qapi-schema.json comments.
fix other comments by Eric Blake.
changes from v5:
Add query-migrate-parameters that display the migration parameters
(currently only capabilities I will add the other parameters separately).
Remove capabilities from query-migrate.
patches are based on git://repo.or.cz/qemu/quintela.git Migration-next-v5 branch
Please review
Juan Quintela (1):
Restart optimization on stage3 update version
Orit Wasserman (11):
Add migration capabilities
Add migrate-set-capabilities
Add XBZRLE documentation
Add cache handling functions
Add uleb encoding/decoding functions
Add xbzrle_encode_buffer and xbzrle_decode_buffer functions
Add XBZRLE to ram_save_block and ram_save_live
Add migrate_set_cache_size command
Add migration accounting for normal and duplicate pages
Change total_time to total-time in MigrationStats
Add XBZRLE statistics
Makefile.objs | 1 +
arch_init.c | 246 ++++++++++++++++++++++++++++++++++++++++++++-
cutils.c | 42 ++++++++
docs/xbzrle.txt | 128 +++++++++++++++++++++++
hmp-commands.hx | 38 +++++++
hmp.c | 103 +++++++++++++++++++
hmp.h | 4 +
include/qemu/page_cache.h | 79 +++++++++++++++
migration.c | 112 ++++++++++++++++++++
migration.h | 21 ++++
monitor.c | 14 +++
page_cache.c | 218 +++++++++++++++++++++++++++++++++++++++
qapi-schema.json | 119 +++++++++++++++++++++-
qemu-common.h | 21 ++++
qmp-commands.hx | 159 ++++++++++++++++++++++++++++-
savevm.c | 159 +++++++++++++++++++++++++++++
16 files changed, 1451 insertions(+), 13 deletions(-)
create mode 100644 docs/xbzrle.txt
create mode 100644 include/qemu/page_cache.h
create mode 100644 page_cache.c
--
1.7.7.6
^ permalink raw reply [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 01/12] Add migration capabilities
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 02/12] Add migrate-set-capabilities Orit Wasserman
` (10 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
The management can query the current migration capabilities using
query-migrate-capabilities QMP command.
The user can use 'info migrate_capabilities' HMP command.
Currently only XBZRLE capability is available.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
hmp-commands.hx | 2 ++
hmp.c | 33 +++++++++++++++++++++++++++++++++
hmp.h | 1 +
migration.c | 30 ++++++++++++++++++++++++++++++
migration.h | 2 ++
monitor.c | 7 +++++++
qapi-schema.json | 39 +++++++++++++++++++++++++++++++++++++++
qmp-commands.hx | 26 ++++++++++++++++++++++++++
8 files changed, 140 insertions(+), 0 deletions(-)
diff --git a/hmp-commands.hx b/hmp-commands.hx
index eea8b32..31d7672 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -1417,6 +1417,8 @@ show CPU statistics
show user network stack connection states
@item info migrate
show migration status
+@item info migrate_capabilities
+show current migration capabilities
@item info balloon
show balloon information
@item info qtree
diff --git a/hmp.c b/hmp.c
index 6b72a64..da82ffe 100644
--- a/hmp.c
+++ b/hmp.c
@@ -131,8 +131,21 @@ void hmp_info_mice(Monitor *mon)
void hmp_info_migrate(Monitor *mon)
{
MigrationInfo *info;
+ MigrationCapabilityStatusList *caps, *cap;
info = qmp_query_migrate(NULL);
+ caps = qmp_query_migrate_capabilities(NULL);
+
+ /* do not display parameters during setup */
+ if (info->has_status && caps) {
+ monitor_printf(mon, "capabilities: ");
+ for (cap = caps; cap; cap = cap->next) {
+ monitor_printf(mon, "%s: %s ",
+ MigrationCapability_lookup[cap->value->capability],
+ cap->value->state ? "on" : "off");
+ }
+ monitor_printf(mon, "\n");
+ }
if (info->has_status) {
monitor_printf(mon, "Migration status: %s\n", info->status);
@@ -159,6 +172,26 @@ void hmp_info_migrate(Monitor *mon)
}
qapi_free_MigrationInfo(info);
+ qapi_free_MigrationCapabilityStatusList(caps);
+}
+
+void hmp_info_migrate_capabilities(Monitor *mon)
+{
+ MigrationCapabilityStatusList *caps, *cap;
+
+ caps = qmp_query_migrate_capabilities(NULL);
+
+ if (caps) {
+ monitor_printf(mon, "capabilities: ");
+ for (cap = caps; cap; cap = cap->next) {
+ monitor_printf(mon, "%s: %s ",
+ MigrationCapability_lookup[cap->value->capability],
+ cap->value->state ? "on" : "off");
+ }
+ monitor_printf(mon, "\n");
+ }
+
+ qapi_free_MigrationCapabilityStatusList(caps);
}
void hmp_info_cpus(Monitor *mon)
diff --git a/hmp.h b/hmp.h
index 8d2b0d7..d111351 100644
--- a/hmp.h
+++ b/hmp.h
@@ -25,6 +25,7 @@ void hmp_info_uuid(Monitor *mon);
void hmp_info_chardev(Monitor *mon);
void hmp_info_mice(Monitor *mon);
void hmp_info_migrate(Monitor *mon);
+void hmp_info_migrate_capabilities(Monitor *mon);
void hmp_info_cpus(Monitor *mon);
void hmp_info_block(Monitor *mon);
void hmp_info_blockstats(Monitor *mon);
diff --git a/migration.c b/migration.c
index 8db1b43..fb620ac 100644
--- a/migration.c
+++ b/migration.c
@@ -113,6 +113,30 @@ uint64_t migrate_max_downtime(void)
return max_downtime;
}
+MigrationCapabilityStatusList *qmp_query_migrate_capabilities(Error **errp)
+{
+ MigrationCapabilityStatusList *head = NULL;
+ MigrationCapabilityStatusList *caps;
+ MigrationState *s = migrate_get_current();
+ int i;
+
+ for (i = 0; i < MIGRATION_CAPABILITY_MAX; i++) {
+ if (head == NULL) {
+ head = g_malloc0(sizeof(*caps));
+ caps = head;
+ } else {
+ caps->next = g_malloc0(sizeof(*caps));
+ caps = caps->next;
+ }
+ caps->value =
+ g_malloc(sizeof(*caps->value));
+ caps->value->capability = i;
+ caps->value->state = s->enabled_capabilities[i];
+ }
+
+ return head;
+}
+
MigrationInfo *qmp_query_migrate(Error **errp)
{
MigrationInfo *info = g_malloc0(sizeof(*info));
@@ -375,10 +399,16 @@ static MigrationState *migrate_init(const MigrationParams *params)
{
MigrationState *s = migrate_get_current();
int64_t bandwidth_limit = s->bandwidth_limit;
+ bool enabled_capabilities[MIGRATION_CAPABILITY_MAX];
+
+ memcpy(enabled_capabilities, s->enabled_capabilities,
+ sizeof(enabled_capabilities));
memset(s, 0, sizeof(*s));
s->bandwidth_limit = bandwidth_limit;
s->params = *params;
+ memcpy(s->enabled_capabilities, enabled_capabilities,
+ sizeof(enabled_capabilities));
s->bandwidth_limit = bandwidth_limit;
s->state = MIG_STATE_SETUP;
diff --git a/migration.h b/migration.h
index 57572a6..713aae0 100644
--- a/migration.h
+++ b/migration.h
@@ -19,6 +19,7 @@
#include "notify.h"
#include "error.h"
#include "vmstate.h"
+#include "qapi-types.h"
struct MigrationParams {
bool blk;
@@ -39,6 +40,7 @@ struct MigrationState
void *opaque;
MigrationParams params;
int64_t total_time;
+ bool enabled_capabilities[MIGRATION_CAPABILITY_MAX];
};
void process_incoming_migration(QEMUFile *f);
diff --git a/monitor.c b/monitor.c
index 09aa3cd..5121277 100644
--- a/monitor.c
+++ b/monitor.c
@@ -2662,6 +2662,13 @@ static mon_cmd_t info_cmds[] = {
.mhandler.info = hmp_info_migrate,
},
{
+ .name = "migrate_capabilities",
+ .args_type = "",
+ .params = "",
+ .help = "show current migration capabilities",
+ .mhandler.info = hmp_info_migrate_capabilities,
+ },
+ {
.name = "balloon",
.args_type = "",
.params = "",
diff --git a/qapi-schema.json b/qapi-schema.json
index a92adb1..77f39a1 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -306,6 +306,45 @@
{ 'command': 'query-migrate', 'returns': 'MigrationInfo' }
##
+# @MigrationCapability
+#
+# Migration capabilities enumeration
+#
+# @xbzrle: Migration supports xbzrle (Xor Based Zero Run Length Encoding).
+# This feature allows us to minimize migration traffic for certain work
+# loads, by sending compressed difference of the pages
+#
+# Since: 1.2
+##
+{ 'enum': 'MigrationCapability',
+ 'data': ['xbzrle'] }
+
+##
+# @MigrationCapabilityStatus
+#
+# Migration capability information
+#
+# @capability: capability enum
+#
+# @state: capability state bool
+#
+# Since: 1.2
+##
+{ 'type': 'MigrationCapabilityStatus',
+ 'data': { 'capability' : 'MigrationCapability', 'state' : 'bool' } }
+
+##
+# @query-migrate-capabilities
+#
+# Returns information about the current migration capabilities status
+#
+# Returns: @MigrationCapabilitiesStatus
+#
+# Since: 1.2
+##
+{ 'command': 'query-migrate-capabilities', 'returns': ['MigrationCapabilityStatus']}
+
+##
# @MouseInfo:
#
# Information about a mouse device.
diff --git a/qmp-commands.hx b/qmp-commands.hx
index e3cf3c5..e6fbbe5 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -2141,6 +2141,32 @@ EQMP
},
SQMP
+query-migrate-capabilities
+-------
+
+Query current migration capabilities
+
+- "capabilities": migration capabilities state
+ - "xbzrle" : XBZRLE state (json-bool)
+
+Arguments:
+
+Example:
+
+-> { "execute": "query-migrate-capabilities" }
+<- { "return": {
+ "capabilities" : [ { "capability" : "xbzrle", "state" : false } ]
+ }
+ }
+EQMP
+
+ {
+ .name = "query-migrate-capabilities",
+ .args_type = "",
+ .mhandler.cmd_new = qmp_marshal_input_query_migrate_capabilities,
+ },
+
+SQMP
query-balloon
-------------
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 02/12] Add migrate-set-capabilities
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 01/12] Add migration capabilities Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 03/12] Add XBZRLE documentation Orit Wasserman
` (9 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
The management can enable/disable a capability for the next migration by using
migrate-set-capabilities QMP command.
The user can use migrate_set_capability HMP command.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
hmp-commands.hx | 14 ++++++++++++++
hmp.c | 32 ++++++++++++++++++++++++++++++++
hmp.h | 1 +
migration.c | 16 ++++++++++++++++
qapi-schema.json | 12 ++++++++++++
qmp-commands.hx | 23 +++++++++++++++++++++++
6 files changed, 98 insertions(+), 0 deletions(-)
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 31d7672..8420794 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -861,6 +861,20 @@ Set maximum tolerated downtime (in seconds) for migration.
ETEXI
{
+ .name = "migrate_set_capability",
+ .args_type = "capability:s,state:b",
+ .params = "capability state",
+ .help = "Enable/Disable the usage of a capability for migration",
+ .mhandler.cmd = hmp_migrate_set_capability,
+ },
+
+STEXI
+@item migrate_set_capability @var{capability} @var{state}
+@findex migrate_set_capability
+Enable/Disable the usage of a capability @var{capability} for migration.
+ETEXI
+
+ {
.name = "client_migrate_info",
.args_type = "protocol:s,hostname:s,port:i?,tls-port:i?,cert-subject:s?",
.params = "protocol hostname port tls-port cert-subject",
diff --git a/hmp.c b/hmp.c
index da82ffe..6131120 100644
--- a/hmp.c
+++ b/hmp.c
@@ -768,6 +768,38 @@ void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict)
qmp_migrate_set_speed(value, NULL);
}
+void hmp_migrate_set_capability(Monitor *mon, const QDict *qdict)
+{
+ const char *cap = qdict_get_str(qdict, "capability");
+ bool state = qdict_get_bool(qdict, "state");
+ Error *err = NULL;
+ MigrationCapabilityStatusList *caps = g_malloc0(sizeof(*caps));
+ int i;
+
+ for (i = 0; i < MIGRATION_CAPABILITY_MAX; i++) {
+ if (strcmp(cap, MigrationCapability_lookup[i]) == 0) {
+ caps->value = g_malloc0(sizeof(*caps->value));
+ caps->value->capability = i;
+ caps->value->state = state;
+ caps->next = NULL;
+ qmp_migrate_set_capabilities(caps, &err);
+ break;
+ }
+ }
+
+ if (i == MIGRATION_CAPABILITY_MAX) {
+ error_set(&err, QERR_INVALID_PARAMETER, cap);
+ }
+
+ qapi_free_MigrationCapabilityStatusList(caps);
+
+ if (err) {
+ monitor_printf(mon, "migrate_set_parameter: %s\n",
+ error_get_pretty(err));
+ error_free(err);
+ }
+}
+
void hmp_set_password(Monitor *mon, const QDict *qdict)
{
const char *protocol = qdict_get_str(qdict, "protocol");
diff --git a/hmp.h b/hmp.h
index d111351..3390a92 100644
--- a/hmp.h
+++ b/hmp.h
@@ -52,6 +52,7 @@ void hmp_snapshot_blkdev(Monitor *mon, const QDict *qdict);
void hmp_migrate_cancel(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_downtime(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict);
+void hmp_migrate_set_capability(Monitor *mon, const QDict *qdict);
void hmp_set_password(Monitor *mon, const QDict *qdict);
void hmp_expire_password(Monitor *mon, const QDict *qdict);
void hmp_eject(Monitor *mon, const QDict *qdict);
diff --git a/migration.c b/migration.c
index fb620ac..ba81df7 100644
--- a/migration.c
+++ b/migration.c
@@ -190,6 +190,22 @@ MigrationInfo *qmp_query_migrate(Error **errp)
return info;
}
+void qmp_migrate_set_capabilities(MigrationCapabilityStatusList *params,
+ Error **errp)
+{
+ MigrationState *s = migrate_get_current();
+ MigrationCapabilityStatusList *cap;
+
+ if (s->state == MIG_STATE_ACTIVE) {
+ error_set(errp, QERR_MIGRATION_ACTIVE);
+ return;
+ }
+
+ for (cap = params; cap; cap = cap->next) {
+ s->enabled_capabilities[cap->value->capability] = cap->value->state;
+ }
+}
+
/* shared migration helpers */
static int migrate_fd_cleanup(MigrationState *s)
diff --git a/qapi-schema.json b/qapi-schema.json
index 77f39a1..4909058 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -334,6 +334,18 @@
'data': { 'capability' : 'MigrationCapability', 'state' : 'bool' } }
##
+# @migrate-set-capabilities
+#
+# Enable/Disable the following migration capabilities (like xbzrle)
+#
+# @capabilities: json array of capability modifications to make
+#
+# Since: 1.2
+##
+{ 'command': 'migrate-set-capabilities',
+ 'data': { 'capabilities': ['MigrationCapabilityStatus'] } }
+
+##
# @query-migrate-capabilities
#
# Returns information about the current migration capabilities status
diff --git a/qmp-commands.hx b/qmp-commands.hx
index e6fbbe5..1f488f4 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -2141,6 +2141,29 @@ EQMP
},
SQMP
+migrate-set-capabilities
+-------
+
+Enable/Disable migration capabilities
+
+- "xbzrle": xbzrle support
+
+Arguments:
+
+Example:
+
+-> { "execute": "migrate-set-capabilities" , "arguments":
+ { "capabilities": [ { "capability": "xbzrle", "state": true } ] } }
+
+EQMP
+
+ {
+ .name = "migrate-set-capabilities",
+ .args_type = "capabilities:O",
+ .params = "capability:s,state:b",
+ .mhandler.cmd_new = qmp_marshal_input_migrate_set_capabilities,
+ },
+SQMP
query-migrate-capabilities
-------
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 03/12] Add XBZRLE documentation
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 01/12] Add migration capabilities Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 02/12] Add migrate-set-capabilities Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 04/12] Add cache handling functions Orit Wasserman
` (8 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
docs/xbzrle.txt | 128 +++++++++++++++++++++++++++++++++++++++++++++++++++++++
1 files changed, 128 insertions(+), 0 deletions(-)
create mode 100644 docs/xbzrle.txt
diff --git a/docs/xbzrle.txt b/docs/xbzrle.txt
new file mode 100644
index 0000000..cc3a26a
--- /dev/null
+++ b/docs/xbzrle.txt
@@ -0,0 +1,128 @@
+XBZRLE (Xor Based Zero Run Length Encoding)
+===========================================
+
+Using XBZRLE (Xor Based Zero Run Length Encoding) allows for the reduction
+of VM downtime and the total live-migration time of Virtual machines.
+It is particularly useful for virtual machines running memory write intensive
+workloads that are typical of large enterprise applications such as SAP ERP
+Systems, and generally speaking for any application that uses a sparse memory
+update pattern.
+
+Instead of sending the changed guest memory page this solution will send a
+compressed version of the updates, thus reducing the amount of data sent during
+live migration.
+In order to be able to calculate the update, the previous memory pages need to
+be stored on the source. Those pages are stored in a dedicated cache
+(hash table) and are accessed by their address.
+The larger the cache size the better the chances are that the page has already
+been stored in the cache.
+A small cache size will result in high cache miss rate.
+Cache size can be changed before and during migration.
+
+Format
+=======
+
+The compression format performs a XOR between the previous and current content
+of the page, where zero represents an unchanged value.
+The page data delta is represented by zero and non zero runs.
+A zero run is represented by its length (in bytes).
+A non zero run is represented by its length (in bytes) and the new data.
+The run length is encoded using ULEB128 (http://en.wikipedia.org/wiki/LEB128)
+
+There can be more than one valid encoding, the sender may send a longer encoding
+for the benefit of reducing computation cost.
+
+page = zrun nzrun
+ | zrun nzrun page
+
+zrun = length
+
+nzrun = length byte...
+
+length = uleb128 encoded integer
+
+On the sender side XBZRLE is used as a compact delta encoding of page updates,
+retrieving the old page content from the cache (default size of 512 MB). The
+receiving side uses the existing page's content and XBZRLE to decode the new
+page's content.
+
+This work was originally based on research results published
+VEE 2011: Evaluation of Delta Compression Techniques for Efficient Live
+Migration of Large Virtual Machines by Benoit, Svard, Tordsson and Elmroth.
+Additionally the delta encoder XBRLE was improved further using the XBZRLE
+instead.
+
+XBZRLE has a sustained bandwidth of 2-2.5 GB/s for typical workloads making it
+ideal for in-line, real-time encoding such as is needed for live-migration.
+
+Example
+old buffer:
+1001 zeros
+05 06 07 08 09 0a 0b 0c 0d 0e 0f 10 11 12 13 68 00 00 6b 00 6d
+3074 zeros
+
+new buffer:
+1001 zeros
+01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 68 00 00 67 00 69
+3074 zeros
+
+encoded buffer:
+
+encoded length 24
+e9 07 0f 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f 03 01 67 01 01 69
+
+Usage
+======================
+1. Verify the destination QEMU version is able to decode the new format.
+ {qemu} info migrate_capabilities
+ {qemu} xbzrle: off , ...
+
+2. Activate xbzrle on both source and destination:
+ {qemu} migrate_set_capability xbzrle on
+
+3. Set the XBZRLE cache size - the cache size is in MBytes and should be a
+power of 2. The cache default value is 64MBytes. (on source only)
+ {qemu} migrate_set_cache_size 256m
+
+4. Start outgoing migration
+ {qemu} migrate -d tcp:destination.host:4444
+ {qemu} info migrate
+ capabilities: xbzrle: on
+ Migration status: active
+ transferred ram: A kbytes
+ remaining ram: B kbytes
+ total ram: C kbytes
+ total time: D milliseconds
+ duplicate: E pages
+ normal: F pages
+ normal bytes: G kbytes
+ cache size: H bytes
+ xbzrle transferred: I kbytes
+ xbzrle pages: J pages
+ xbzrle cache miss: K
+ xbzrle overflow : L
+
+xbzrle cache-miss: the number of cache misses to date - high cache-miss rate
+indicates that the cache size is set too low.
+xbzrle overflow: the number of overflows in the decoding which where the delta
+could not be compressed. This can happen if the changes in the pages are too
+large or there are many short changes; for example, changing every second byte
+(half a page).
+
+Testing: Testing indicated that live migration with XBZRLE was completed in 110
+seconds, whereas without it would not be able to complete.
+
+A simple synthetic memory r/w load generator:
+.. include <stdlib.h>
+.. include <stdio.h>
+.. int main()
+.. {
+.. char *buf = (char *) calloc(4096, 4096);
+.. while (1) {
+.. int i;
+.. for (i = 0; i < 4096 * 4; i++) {
+.. buf[i * 4096 / 4]++;
+.. }
+.. printf(".");
+.. }
+.. }
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 04/12] Add cache handling functions
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (2 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 03/12] Add XBZRLE documentation Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 05/12] Add uleb encoding/decoding functions Orit Wasserman
` (7 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
Add MRU page cache mechanism.
The page are accessed by their address.
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
Makefile.objs | 1 +
cutils.c | 9 ++
include/qemu/page_cache.h | 79 ++++++++++++++++
page_cache.c | 218 +++++++++++++++++++++++++++++++++++++++++++++
qemu-common.h | 13 +++
5 files changed, 320 insertions(+), 0 deletions(-)
create mode 100644 include/qemu/page_cache.h
create mode 100644 page_cache.c
diff --git a/Makefile.objs b/Makefile.objs
index 5ebbcfa..e0fb69b 100644
--- a/Makefile.objs
+++ b/Makefile.objs
@@ -77,6 +77,7 @@ common-obj-y += qemu-char.o #aio.o
common-obj-y += block-migration.o iohandler.o
common-obj-y += pflib.o
common-obj-y += bitmap.o bitops.o
+common-obj-y += page_cache.o
common-obj-$(CONFIG_POSIX) += migration-exec.o migration-unix.o migration-fd.o
common-obj-$(CONFIG_WIN32) += version.o
diff --git a/cutils.c b/cutils.c
index e2bc1b8..b0bdd4b 100644
--- a/cutils.c
+++ b/cutils.c
@@ -375,3 +375,12 @@ int qemu_parse_fd(const char *param)
}
return fd;
}
+
+/* round down to the nearest power of 2*/
+int64_t pow2floor(int64_t value)
+{
+ if (!is_power_of_2(value)) {
+ value = 0x8000000000000000ULL >> clz64(value);
+ }
+ return value;
+}
diff --git a/include/qemu/page_cache.h b/include/qemu/page_cache.h
new file mode 100644
index 0000000..3839ac7
--- /dev/null
+++ b/include/qemu/page_cache.h
@@ -0,0 +1,79 @@
+/*
+ * Page cache for QEMU
+ * The cache is base on a hash of the page address
+ *
+ * Copyright 2012 Red Hat, Inc. and/or its affiliates
+ *
+ * Authors:
+ * Orit Wasserman <owasserm@redhat.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.
+ *
+ */
+
+#ifndef PAGE_CACHE_H
+#define PAGE_CACHE_H
+
+/* Page cache for storing guest pages */
+typedef struct PageCache PageCache;
+
+/**
+ * cache_init: Initialize the page cache
+ *
+ *
+ * Returns new allocated cache or NULL on error
+ *
+ * @cache pointer to the PageCache struct
+ * @num_pages: cache maximal number of cached pages
+ * @page_size: cache page size
+ */
+PageCache *cache_init(int64_t num_pages, unsigned int page_size);
+
+/**
+ * cache_fini: free all cache resources
+ * @cache pointer to the PageCache struct
+ */
+void cache_fini(PageCache *cache);
+
+/**
+ * cache_is_cached: Checks to see if the page is cached
+ *
+ * Returns %true if page is cached
+ *
+ * @cache pointer to the PageCache struct
+ * @addr: page addr
+ */
+bool cache_is_cached(const PageCache *cache, uint64_t addr);
+
+/**
+ * get_cached_data: Get the data cached for an addr
+ *
+ * Returns pointer to the data cached or NULL if not cached
+ *
+ * @cache pointer to the PageCache struct
+ * @addr: page addr
+ */
+uint8_t *get_cached_data(const PageCache *cache, uint64_t addr);
+
+/**
+ * cache_insert: insert the page into the cache. the previous value will be overwritten
+ *
+ * @cache pointer to the PageCache struct
+ * @addr: page address
+ * @pdata: pointer to the page
+ */
+void cache_insert(PageCache *cache, uint64_t addr, uint8_t *pdata);
+
+/**
+ * cache_resize: resize the page cache. In case of size reduction the extra
+ * pages will be freed
+ *
+ * Returns -1 on error new cache size on success
+ *
+ * @cache pointer to the PageCache struct
+ * @num_pages: new page cache size (in pages)
+ */
+int64_t cache_resize(PageCache *cache, int64_t num_pages);
+
+#endif
diff --git a/page_cache.c b/page_cache.c
new file mode 100644
index 0000000..0294f7e
--- /dev/null
+++ b/page_cache.c
@@ -0,0 +1,218 @@
+/*
+ * Page cache for QEMU
+ * The cache is base on a hash of the page address
+ *
+ * Copyright 2012 Red Hat, Inc. and/or its affiliates
+ *
+ * Authors:
+ * Orit Wasserman <owasserm@redhat.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 <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <strings.h>
+#include <string.h>
+#include <sys/time.h>
+#include <sys/types.h>
+#include <stdbool.h>
+#include <glib.h>
+#include <strings.h>
+
+#include "qemu-common.h"
+#include "qemu/page_cache.h"
+
+#ifdef DEBUG_CACHE
+#define DPRINTF(fmt, ...) \
+ do { fprintf(stdout, "cache: " fmt, ## __VA_ARGS__); } while (0)
+#else
+#define DPRINTF(fmt, ...) \
+ do { } while (0)
+#endif
+
+typedef struct CacheItem CacheItem;
+
+struct CacheItem {
+ uint64_t it_addr;
+ uint64_t it_age;
+ uint8_t *it_data;
+};
+
+struct PageCache {
+ CacheItem *page_cache;
+ unsigned int page_size;
+ int64_t max_num_items;
+ uint64_t max_item_age;
+ int64_t num_items;
+};
+
+PageCache *cache_init(int64_t num_pages, unsigned int page_size)
+{
+ int64_t i;
+
+ PageCache *cache;
+
+ if (num_pages <= 0) {
+ DPRINTF("invalid number of pages\n");
+ return NULL;
+ }
+
+ cache = g_malloc(sizeof(*cache));
+
+ /* round down to the nearest power of 2 */
+ if (!is_power_of_2(num_pages)) {
+ num_pages = pow2floor(num_pages);
+ DPRINTF("rounding down to %" PRId64 "\n", num_pages);
+ }
+ cache->page_size = page_size;
+ cache->num_items = 0;
+ cache->max_item_age = 0;
+ cache->max_num_items = num_pages;
+
+ DPRINTF("Setting cache buckets to %" PRId64 "\n", cache->max_num_items);
+
+ cache->page_cache = g_malloc((cache->max_num_items) *
+ sizeof(*cache->page_cache));
+
+ for (i = 0; i < cache->max_num_items; i++) {
+ cache->page_cache[i].it_data = NULL;
+ cache->page_cache[i].it_age = 0;
+ cache->page_cache[i].it_addr = -1;
+ }
+
+ return cache;
+}
+
+void cache_fini(PageCache *cache)
+{
+ int64_t i;
+
+ g_assert(cache);
+ g_assert(cache->page_cache);
+
+ for (i = 0; i < cache->max_num_items; i++) {
+ g_free(cache->page_cache[i].it_data);
+ }
+
+ g_free(cache->page_cache);
+ cache->page_cache = NULL;
+}
+
+static size_t cache_get_cache_pos(const PageCache *cache,
+ uint64_t address)
+{
+ size_t pos;
+
+ g_assert(cache->max_num_items);
+ pos = (address / cache->page_size) & (cache->max_num_items - 1);
+ return pos;
+}
+
+bool cache_is_cached(const PageCache *cache, uint64_t addr)
+{
+ size_t pos;
+
+ g_assert(cache);
+ g_assert(cache->page_cache);
+
+ pos = cache_get_cache_pos(cache, addr);
+
+ return (cache->page_cache[pos].it_addr == addr);
+}
+
+static CacheItem *cache_get_by_addr(const PageCache *cache, uint64_t addr)
+{
+ size_t pos;
+
+ g_assert(cache);
+ g_assert(cache->page_cache);
+
+ pos = cache_get_cache_pos(cache, addr);
+
+ return &cache->page_cache[pos];
+}
+
+uint8_t *get_cached_data(const PageCache *cache, uint64_t addr)
+{
+ return cache_get_by_addr(cache, addr)->it_data;
+}
+
+void cache_insert(PageCache *cache, uint64_t addr, uint8_t *pdata)
+{
+
+ CacheItem *it = NULL;
+
+ g_assert(cache);
+ g_assert(cache->page_cache);
+
+ /* actual update of entry */
+ it = cache_get_by_addr(cache, addr);
+
+ if (!it->it_data) {
+ cache->num_items++;
+ }
+
+ it->it_data = pdata;
+ it->it_age = ++cache->max_item_age;
+ it->it_addr = addr;
+}
+
+int64_t cache_resize(PageCache *cache, int64_t new_num_pages)
+{
+ PageCache *new_cache;
+ int64_t i;
+
+ CacheItem *old_it, *new_it;
+
+ g_assert(cache);
+
+ /* cache was not inited */
+ if (cache->page_cache == NULL) {
+ return -1;
+ }
+
+ /* same size */
+ if (pow2floor(new_num_pages) == cache->max_num_items) {
+ return cache->max_num_items;
+ }
+
+ new_cache = cache_init(new_num_pages, cache->page_size);
+ if (!(new_cache)) {
+ DPRINTF("Error creating new cache\n");
+ return -1;
+ }
+
+ /* move all data from old cache */
+ for (i = 0; i < cache->max_num_items; i++) {
+ old_it = &cache->page_cache[i];
+ if (old_it->it_addr != -1) {
+ /* check for collision, if there is, keep MRU page */
+ new_it = cache_get_by_addr(new_cache, old_it->it_addr);
+ if (new_it->it_data) {
+ /* keep the MRU page */
+ if (new_it->it_age >= old_it->it_age) {
+ g_free(old_it->it_data);
+ } else {
+ g_free(new_it->it_data);
+ new_it->it_data = old_it->it_data;
+ new_it->it_age = old_it->it_age;
+ new_it->it_addr = old_it->it_addr;
+ }
+ } else {
+ cache_insert(new_cache, old_it->it_addr, old_it->it_data);
+ }
+ }
+ }
+
+ cache->page_cache = new_cache->page_cache;
+ cache->max_num_items = new_cache->max_num_items;
+ cache->num_items = new_cache->num_items;
+
+ g_free(new_cache);
+
+ return cache->max_num_items;
+}
diff --git a/qemu-common.h b/qemu-common.h
index 09676f5..855f242 100644
--- a/qemu-common.h
+++ b/qemu-common.h
@@ -1,3 +1,4 @@
+
/* Common header file that is included by all of qemu. */
#ifndef QEMU_COMMON_H
#define QEMU_COMMON_H
@@ -411,6 +412,18 @@ static inline uint64_t muldiv64(uint64_t a, uint32_t b, uint32_t c)
/* Round number up to multiple */
#define QEMU_ALIGN_UP(n, m) QEMU_ALIGN_DOWN((n) + (m) - 1, (m))
+static inline bool is_power_of_2(uint64_t value)
+{
+ if (!value) {
+ return 0;
+ }
+
+ return !(value & (value - 1));
+}
+
+/* round down to the nearest power of 2*/
+int64_t pow2floor(int64_t value);
+
#include "module.h"
#endif
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 05/12] Add uleb encoding/decoding functions
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (3 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 04/12] Add cache handling functions Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 06/12] Add xbzrle_encode_buffer and xbzrle_decode_buffer functions Orit Wasserman
` (6 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
Implement Unsigned Little Endian Base 128.
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
cutils.c | 33 +++++++++++++++++++++++++++++++++
qemu-common.h | 8 ++++++++
2 files changed, 41 insertions(+), 0 deletions(-)
diff --git a/cutils.c b/cutils.c
index b0bdd4b..700f943 100644
--- a/cutils.c
+++ b/cutils.c
@@ -384,3 +384,36 @@ int64_t pow2floor(int64_t value)
}
return value;
}
+
+/*
+ * Implementation of ULEB128 (http://en.wikipedia.org/wiki/LEB128)
+ * Input is limited to 14-bit numbers
+ */
+int uleb128_encode_small(uint8_t *out, uint32_t n)
+{
+ g_assert(n <= 0x3fff);
+ if (n < 0x80) {
+ *out++ = n;
+ return 1;
+ } else {
+ *out++ = (n & 0x7f) | 0x80;
+ *out++ = n >> 7;
+ return 2;
+ }
+}
+
+int uleb128_decode_small(const uint8_t *in, uint32_t *n)
+{
+ if (!(*in & 0x80)) {
+ *n = *in++;
+ return 1;
+ } else {
+ *n = *in++ & 0x7f;
+ /* we exceed 14 bit number */
+ if (*in & 0x80) {
+ return -1;
+ }
+ *n |= *in++ << 7;
+ return 2;
+ }
+}
diff --git a/qemu-common.h b/qemu-common.h
index 855f242..93d1c63 100644
--- a/qemu-common.h
+++ b/qemu-common.h
@@ -426,4 +426,12 @@ int64_t pow2floor(int64_t value);
#include "module.h"
+/*
+ * Implementation of ULEB128 (http://en.wikipedia.org/wiki/LEB128)
+ * Input is limited to 14-bit numbers
+ */
+
+int uleb128_encode_small(uint8_t *out, uint32_t n);
+int uleb128_decode_small(const uint8_t *in, uint32_t *n);
+
#endif
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 06/12] Add xbzrle_encode_buffer and xbzrle_decode_buffer functions
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (4 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 05/12] Add uleb encoding/decoding functions Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 07/12] Add XBZRLE to ram_save_block and ram_save_live Orit Wasserman
` (5 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
For performance we are encoding long word at a time.
For nzrun we use long-word-at-a-time NULL-detection tricks from strcmp():
using ((lword - 0x0101010101010101) & (~lword) & 0x8080808080808080) test
to find out if any byte in the long word is zero.
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
---
migration.h | 4 ++
savevm.c | 159 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 163 insertions(+), 0 deletions(-)
diff --git a/migration.h b/migration.h
index 713aae0..743c366 100644
--- a/migration.h
+++ b/migration.h
@@ -100,4 +100,8 @@ void migrate_add_blocker(Error *reason);
*/
void migrate_del_blocker(Error *reason);
+int xbzrle_encode_buffer(uint8_t *old_buf, uint8_t *new_buf, int slen,
+ uint8_t *dst, int dlen);
+int xbzrle_decode_buffer(uint8_t *src, int slen, uint8_t *dst, int dlen);
+
#endif
diff --git a/savevm.c b/savevm.c
index 6e82b2d..0ea10c9 100644
--- a/savevm.c
+++ b/savevm.c
@@ -2392,3 +2392,162 @@ void vmstate_register_ram_global(MemoryRegion *mr)
{
vmstate_register_ram(mr, NULL);
}
+
+/*
+ page = zrun nzrun
+ | zrun nzrun page
+
+ zrun = length
+
+ nzrun = length byte...
+
+ length = uleb128 encoded integer
+ */
+int xbzrle_encode_buffer(uint8_t *old_buf, uint8_t *new_buf, int slen,
+ uint8_t *dst, int dlen)
+{
+ uint32_t zrun_len = 0, nzrun_len = 0;
+ int d = 0, i = 0;
+ long res, xor;
+ uint8_t *nzrun_start = NULL;
+
+ g_assert(!(((uintptr_t)old_buf | (uintptr_t)new_buf | slen) %
+ sizeof(long)));
+
+ while (i < slen) {
+ /* overflow */
+ if (d + 2 > dlen) {
+ return -1;
+ }
+
+ /* not aligned to sizeof(long) */
+ res = (slen - i) % sizeof(long);
+ while (res && old_buf[i] == new_buf[i]) {
+ zrun_len++;
+ i++;
+ res--;
+ }
+
+ /* word at a time for speed */
+ if (!res) {
+ while (i < slen &&
+ (*(long *)(old_buf + i)) == (*(long *)(new_buf + i))) {
+ i += sizeof(long);
+ zrun_len += sizeof(long);
+ }
+
+ /* go over the rest */
+ while (i < slen && old_buf[i] == new_buf[i]) {
+ zrun_len++;
+ i++;
+ }
+ }
+
+ /* buffer unchanged */
+ if (zrun_len == slen) {
+ return 0;
+ }
+
+ /* skip last zero run */
+ if (i == slen) {
+ return d;
+ }
+
+ d += uleb128_encode_small(dst + d, zrun_len);
+
+ zrun_len = 0;
+ nzrun_start = new_buf + i;
+
+ /* overflow */
+ if (d + 2 > dlen) {
+ return -1;
+ }
+ /* not aligned to sizeof(long) */
+ res = (slen - i) % sizeof(long);
+ while (res && old_buf[i] != new_buf[i]) {
+ i++;
+ nzrun_len++;
+ res--;
+ }
+
+ /* word at a time for speed, use of 32-bit long okay */
+ if (!res) {
+ /* truncation to 32-bit long okay */
+ long mask = 0x0101010101010101ULL;
+ while (i < slen) {
+ xor = *(long *)(old_buf + i) ^ *(long *)(new_buf + i);
+ if ((xor - mask) & ~xor & (mask << 7)) {
+ /* found the end of an nzrun within the current long */
+ while (old_buf[i] != new_buf[i]) {
+ nzrun_len++;
+ i++;
+ }
+ break;
+ } else {
+ i += sizeof(long);
+ nzrun_len += sizeof(long);
+ }
+ }
+ }
+
+ d += uleb128_encode_small(dst + d, nzrun_len);
+ /* overflow */
+ if (d + nzrun_len > dlen) {
+ return -1;
+ }
+ memcpy(dst + d, nzrun_start, nzrun_len);
+ d += nzrun_len;
+ nzrun_len = 0;
+ }
+
+ return d;
+}
+
+int xbzrle_decode_buffer(uint8_t *src, int slen, uint8_t *dst, int dlen)
+{
+ int i = 0, d = 0;
+ int ret;
+ uint32_t count = 0;
+
+ while (i < slen) {
+
+ /* zrun */
+ if ((slen - i) < 2) {
+ return -1;
+ }
+
+ ret = uleb128_decode_small(src + i, &count);
+ if (ret < 0 || (i && !count)) {
+ return -1;
+ }
+ i += ret;
+ d += count;
+
+ /* overflow */
+ if (d > dlen) {
+ return -1;
+ }
+
+ /* nzrun */
+ if ((slen - i) < 2) {
+ return -1;
+ }
+
+ ret = uleb128_decode_small(src + i, &count);
+ if (ret < 0 || !count) {
+ return -1;
+ }
+ i += ret;
+
+ /* overflow */
+ if (d + count > dlen || i + count > slen) {
+ return -1;
+ }
+
+ memcpy(dst + d, src + i, count);
+ d += count;
+ i += count;
+ }
+
+ return d;
+}
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 07/12] Add XBZRLE to ram_save_block and ram_save_live
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (5 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 06/12] Add xbzrle_encode_buffer and xbzrle_decode_buffer functions Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command Orit Wasserman
` (4 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
In the outgoing migration check to see if the page is cached and
changed, then send compressed page by using save_xbrle_page function.
In the incoming migration check to see if RAM_SAVE_FLAG_XBZRLE is set
and decompress the page (by using load_xbrle function).
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
arch_init.c | 158 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++-
migration.c | 24 +++++++++
migration.h | 4 ++
3 files changed, 184 insertions(+), 2 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 78cdf50..410ba4d 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -43,6 +43,7 @@
#include "hw/smbios.h"
#include "exec-memory.h"
#include "hw/pcspk.h"
+#include "qemu/page_cache.h"
#ifdef DEBUG_ARCH_INIT
#define DPRINTF(fmt, ...) \
@@ -102,6 +103,7 @@ const uint32_t arch_type = QEMU_ARCH;
#define RAM_SAVE_FLAG_PAGE 0x08
#define RAM_SAVE_FLAG_EOS 0x10
#define RAM_SAVE_FLAG_CONTINUE 0x20
+#define RAM_SAVE_FLAG_XBZRLE 0x40
#ifdef __ALTIVEC__
#include <altivec.h>
@@ -169,6 +171,24 @@ static int is_dup_page(uint8_t *page)
return 1;
}
+/* struct contains XBZRLE cache and a static page
+ used by the compression */
+static struct {
+ /* buffer used for XBZRLE encoding */
+ uint8_t *encoded_buf;
+ /* buffer for storing page content */
+ uint8_t *current_buf;
+ /* buffer used for XBZRLE decoding */
+ uint8_t *decoded_buf;
+ /* Cache for XBZRLE */
+ PageCache *cache;
+} XBZRLE = {
+ .encoded_buf = NULL,
+ .current_buf = NULL,
+ .decoded_buf = NULL,
+ .cache = NULL,
+};
+
static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
int cont, int flag)
{
@@ -181,6 +201,53 @@ static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
}
+#define ENCODING_FLAG_XBZRLE 0x1
+
+static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
+ ram_addr_t current_addr, RAMBlock *block,
+ ram_addr_t offset, int cont)
+{
+ int encoded_len = 0, bytes_sent = -1;
+ uint8_t *prev_cached_page;
+
+ if (!cache_is_cached(XBZRLE.cache, current_addr)) {
+ cache_insert(XBZRLE.cache, current_addr,
+ g_memdup(current_data, TARGET_PAGE_SIZE));
+ return -1;
+ }
+
+ prev_cached_page = get_cached_data(XBZRLE.cache, current_addr);
+
+ /* save current buffer into memory */
+ memcpy(XBZRLE.current_buf, current_data, TARGET_PAGE_SIZE);
+
+ /* XBZRLE encoding (if there is no overflow) */
+ encoded_len = xbzrle_encode_buffer(prev_cached_page, XBZRLE.current_buf,
+ TARGET_PAGE_SIZE, XBZRLE.encoded_buf,
+ TARGET_PAGE_SIZE);
+ if (encoded_len == 0) {
+ DPRINTF("Skipping unmodified page\n");
+ return 0;
+ } else if (encoded_len == -1) {
+ DPRINTF("Overflow\n");
+ /* update data in the cache */
+ memcpy(prev_cached_page, current_data, TARGET_PAGE_SIZE);
+ return -1;
+ }
+
+ /* we need to update the data in the cache, in order to get the same data */
+ memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
+
+ /* Send XBZRLE based compressed page */
+ save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_XBZRLE);
+ qemu_put_byte(f, ENCODING_FLAG_XBZRLE);
+ qemu_put_be16(f, encoded_len);
+ qemu_put_buffer(f, XBZRLE.encoded_buf, encoded_len);
+ bytes_sent = encoded_len + 1 + 2;
+
+ return bytes_sent;
+}
+
static RAMBlock *last_block;
static ram_addr_t last_offset;
@@ -198,6 +265,7 @@ static int ram_save_block(QEMUFile *f)
ram_addr_t offset = last_offset;
int bytes_sent = -1;
MemoryRegion *mr;
+ ram_addr_t current_addr;
if (!block)
block = QLIST_FIRST(&ram_list.blocks);
@@ -218,13 +286,24 @@ static int ram_save_block(QEMUFile *f)
save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_COMPRESS);
qemu_put_byte(f, *p);
bytes_sent = 1;
- } else {
+ } else if (migrate_use_xbzrle()) {
+ current_addr = block->offset + offset;
+ bytes_sent = save_xbzrle_page(f, p, current_addr, block,
+ offset, cont);
+ p = get_cached_data(XBZRLE.cache, current_addr);
+ }
+
+ /* either we didn't send yet (we may have had XBZRLE overflow) */
+ if (bytes_sent == -1) {
save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_PAGE);
qemu_put_buffer(f, p, TARGET_PAGE_SIZE);
bytes_sent = TARGET_PAGE_SIZE;
}
- break;
+ /* if page is unmodified, continue to the next */
+ if (bytes_sent != 0) {
+ break;
+ }
}
offset += TARGET_PAGE_SIZE;
@@ -302,6 +381,15 @@ static void sort_ram_list(void)
static void migration_end(void)
{
memory_global_dirty_log_stop();
+
+ if (migrate_use_xbzrle()) {
+ cache_fini(XBZRLE.cache);
+ g_free(XBZRLE.cache);
+ g_free(XBZRLE.encoded_buf);
+ g_free(XBZRLE.current_buf);
+ g_free(XBZRLE.decoded_buf);
+ XBZRLE.cache = NULL;
+ }
}
static void ram_migration_cancel(void *opaque)
@@ -321,6 +409,18 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
last_offset = 0;
sort_ram_list();
+ if (migrate_use_xbzrle()) {
+ XBZRLE.cache = cache_init(migrate_xbzrle_cache_size() /
+ TARGET_PAGE_SIZE,
+ TARGET_PAGE_SIZE);
+ if (!XBZRLE.cache) {
+ DPRINTF("Error creating cache\n");
+ return -1;
+ }
+ XBZRLE.encoded_buf = g_malloc0(TARGET_PAGE_SIZE);
+ XBZRLE.current_buf = g_malloc(TARGET_PAGE_SIZE);
+ }
+
/* Make sure all dirty bits are set */
QLIST_FOREACH(block, &ram_list.blocks, next) {
for (addr = 0; addr < block->length; addr += TARGET_PAGE_SIZE) {
@@ -436,6 +536,47 @@ static int ram_save_complete(QEMUFile *f, void *opaque)
return 0;
}
+static int load_xbzrle(QEMUFile *f, ram_addr_t addr, void *host)
+{
+ int ret, rc = 0;
+ unsigned int xh_len;
+ int xh_flags;
+
+ if (!XBZRLE.decoded_buf) {
+ XBZRLE.decoded_buf = g_malloc(TARGET_PAGE_SIZE);
+ }
+
+ /* extract RLE header */
+ xh_flags = qemu_get_byte(f);
+ xh_len = qemu_get_be16(f);
+
+ if (xh_flags != ENCODING_FLAG_XBZRLE) {
+ fprintf(stderr, "Failed to load XBZRLE page - wrong compression!\n");
+ return -1;
+ }
+
+ if (xh_len > TARGET_PAGE_SIZE) {
+ fprintf(stderr, "Failed to load XBZRLE page - len overflow!\n");
+ return -1;
+ }
+ /* load data and decode */
+ qemu_get_buffer(f, XBZRLE.decoded_buf, xh_len);
+
+ /* decode RLE */
+ ret = xbzrle_decode_buffer(XBZRLE.decoded_buf, xh_len, host,
+ TARGET_PAGE_SIZE);
+ if (ret == -1) {
+ fprintf(stderr, "Failed to load XBZRLE page - decode error!\n");
+ rc = -1;
+ } else if (ret > TARGET_PAGE_SIZE) {
+ fprintf(stderr, "Failed to load XBZRLE page - size %d exceeds %d!\n",
+ ret, TARGET_PAGE_SIZE);
+ abort();
+ }
+
+ return rc;
+}
+
static inline void *host_from_stream_offset(QEMUFile *f,
ram_addr_t offset,
int flags)
@@ -549,6 +690,19 @@ static int ram_load(QEMUFile *f, void *opaque, int version_id)
}
qemu_get_buffer(f, host, TARGET_PAGE_SIZE);
+ } else if (flags & RAM_SAVE_FLAG_XBZRLE) {
+ if (!migrate_use_xbzrle()) {
+ return -EINVAL;
+ }
+ void *host = host_from_stream_offset(f, addr, flags);
+ if (!host) {
+ return -EINVAL;
+ }
+
+ if (load_xbzrle(f, addr, host) < 0) {
+ ret = -EINVAL;
+ goto done;
+ }
}
error = qemu_file_get_error(f);
if (error) {
diff --git a/migration.c b/migration.c
index ba81df7..5a5333c 100644
--- a/migration.c
+++ b/migration.c
@@ -43,6 +43,9 @@ enum {
#define MAX_THROTTLE (32 << 20) /* Migration speed throttling */
+/* Migration XBZRLE default cache size */
+#define DEFAULT_MIGRATE_CACHE_SIZE (64 * 1024 * 1024)
+
static NotifierList migration_state_notifiers =
NOTIFIER_LIST_INITIALIZER(migration_state_notifiers);
@@ -55,6 +58,7 @@ static MigrationState *migrate_get_current(void)
static MigrationState current_migration = {
.state = MIG_STATE_SETUP,
.bandwidth_limit = MAX_THROTTLE,
+ .xbzrle_cache_size = DEFAULT_MIGRATE_CACHE_SIZE,
};
return ¤t_migration;
@@ -416,6 +420,7 @@ static MigrationState *migrate_init(const MigrationParams *params)
MigrationState *s = migrate_get_current();
int64_t bandwidth_limit = s->bandwidth_limit;
bool enabled_capabilities[MIGRATION_CAPABILITY_MAX];
+ int64_t xbzrle_cache_size = s->xbzrle_cache_size;
memcpy(enabled_capabilities, s->enabled_capabilities,
sizeof(enabled_capabilities));
@@ -425,6 +430,7 @@ static MigrationState *migrate_init(const MigrationParams *params)
s->params = *params;
memcpy(s->enabled_capabilities, enabled_capabilities,
sizeof(enabled_capabilities));
+ s->xbzrle_cache_size = xbzrle_cache_size;
s->bandwidth_limit = bandwidth_limit;
s->state = MIG_STATE_SETUP;
@@ -524,3 +530,21 @@ void qmp_migrate_set_downtime(double value, Error **errp)
value = MAX(0, MIN(UINT64_MAX, value));
max_downtime = (uint64_t)value;
}
+
+int migrate_use_xbzrle(void)
+{
+ MigrationState *s;
+
+ s = migrate_get_current();
+
+ return s->enabled_capabilities[MIGRATION_CAPABILITY_XBZRLE];
+}
+
+int64_t migrate_xbzrle_cache_size(void)
+{
+ MigrationState *s;
+
+ s = migrate_get_current();
+
+ return s->xbzrle_cache_size;
+}
diff --git a/migration.h b/migration.h
index 743c366..cdf6787 100644
--- a/migration.h
+++ b/migration.h
@@ -41,6 +41,7 @@ struct MigrationState
MigrationParams params;
int64_t total_time;
bool enabled_capabilities[MIGRATION_CAPABILITY_MAX];
+ int64_t xbzrle_cache_size;
};
void process_incoming_migration(QEMUFile *f);
@@ -104,4 +105,7 @@ int xbzrle_encode_buffer(uint8_t *old_buf, uint8_t *new_buf, int slen,
uint8_t *dst, int dlen);
int xbzrle_decode_buffer(uint8_t *src, int slen, uint8_t *dst, int dlen);
+int migrate_use_xbzrle(void);
+int64_t migrate_xbzrle_cache_size(void);
+
#endif
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (6 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 07/12] Add XBZRLE to ram_save_block and ram_save_live Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 18:13 ` Eric Blake
2012-08-06 17:51 ` [Qemu-devel] [PATCH 09/12] Add migration accounting for normal and duplicate pages Orit Wasserman
` (3 subsequent siblings)
11 siblings, 1 reply; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
Change XBZRLE cache size in bytes (the size should be a power of 2, it will be
rounded down to the nearest power of 2).
If XBZRLE cache size is too small there will be many cache miss.
New query-migrate-cache-size QMP command and 'info migrate_cache_size' HMP
command to query cache value.
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
arch_init.c | 10 ++++++++++
hmp-commands.hx | 22 ++++++++++++++++++++++
hmp.c | 19 +++++++++++++++++++
hmp.h | 2 ++
migration.c | 19 +++++++++++++++++++
migration.h | 2 ++
monitor.c | 7 +++++++
qapi-schema.json | 27 +++++++++++++++++++++++++++
qmp-commands.hx | 44 ++++++++++++++++++++++++++++++++++++++++++++
9 files changed, 152 insertions(+), 0 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 410ba4d..d709ccb 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -189,6 +189,16 @@ static struct {
.cache = NULL,
};
+
+int64_t xbzrle_cache_resize(int64_t new_size)
+{
+ if (XBZRLE.cache != NULL) {
+ return cache_resize(XBZRLE.cache, new_size / TARGET_PAGE_SIZE) *
+ TARGET_PAGE_SIZE;
+ }
+ return pow2floor(new_size);
+}
+
static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
int cont, int flag)
{
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 8420794..6bb5b4a 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -829,6 +829,26 @@ STEXI
@item migrate_cancel
@findex migrate_cancel
Cancel the current VM migration.
+
+ETEXI
+
+ {
+ .name = "migrate_set_cache_size",
+ .args_type = "value:o",
+ .params = "value",
+ .help = "set cache size (in bytes) for XBZRLE migrations,"
+ "the cache size will be rounded down to the nearest "
+ "power of 2.\n"
+ "The cache size affects the number of cache misses."
+ "In case of a high cache miss ratio you need to increase"
+ " the cache size",
+ .mhandler.cmd = hmp_migrate_set_cache_size,
+ },
+
+STEXI
+@item migrate_set_cache_size @var{value}
+@findex migrate_set_cache_size
+Set cache size to @var{value} (in bytes) for xbzrle migrations.
ETEXI
{
@@ -1433,6 +1453,8 @@ show user network stack connection states
show migration status
@item info migrate_capabilities
show current migration capabilities
+@item info migrate_cache_size
+show current migration XBZRLE cache size
@item info balloon
show balloon information
@item info qtree
diff --git a/hmp.c b/hmp.c
index 6131120..42cbfb9 100644
--- a/hmp.c
+++ b/hmp.c
@@ -194,6 +194,12 @@ void hmp_info_migrate_capabilities(Monitor *mon)
qapi_free_MigrationCapabilityStatusList(caps);
}
+void hmp_info_migrate_cache_size(Monitor *mon)
+{
+ monitor_printf(mon, "xbzrel cache size: %" PRId64 " kbytes\n",
+ qmp_query_migrate_cache_size(NULL) >> 10);
+}
+
void hmp_info_cpus(Monitor *mon)
{
CpuInfoList *cpu_list, *cpu;
@@ -762,6 +768,19 @@ void hmp_migrate_set_downtime(Monitor *mon, const QDict *qdict)
qmp_migrate_set_downtime(value, NULL);
}
+void hmp_migrate_set_cache_size(Monitor *mon, const QDict *qdict)
+{
+ int64_t value = qdict_get_int(qdict, "value");
+ Error *err = NULL;
+
+ qmp_migrate_set_cache_size(value, &err);
+ if (err) {
+ monitor_printf(mon, "%s\n", error_get_pretty(err));
+ error_free(err);
+ return;
+ }
+}
+
void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict)
{
int64_t value = qdict_get_int(qdict, "value");
diff --git a/hmp.h b/hmp.h
index 3390a92..6d6e53b 100644
--- a/hmp.h
+++ b/hmp.h
@@ -26,6 +26,7 @@ void hmp_info_chardev(Monitor *mon);
void hmp_info_mice(Monitor *mon);
void hmp_info_migrate(Monitor *mon);
void hmp_info_migrate_capabilities(Monitor *mon);
+void hmp_info_migrate_cache_size(Monitor *mon);
void hmp_info_cpus(Monitor *mon);
void hmp_info_block(Monitor *mon);
void hmp_info_blockstats(Monitor *mon);
@@ -53,6 +54,7 @@ void hmp_migrate_cancel(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_downtime(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_capability(Monitor *mon, const QDict *qdict);
+void hmp_migrate_set_cache_size(Monitor *mon, const QDict *qdict);
void hmp_set_password(Monitor *mon, const QDict *qdict);
void hmp_expire_password(Monitor *mon, const QDict *qdict);
void hmp_eject(Monitor *mon, const QDict *qdict);
diff --git a/migration.c b/migration.c
index 5a5333c..a6c0072 100644
--- a/migration.c
+++ b/migration.c
@@ -511,6 +511,25 @@ void qmp_migrate_cancel(Error **errp)
migrate_fd_cancel(migrate_get_current());
}
+void qmp_migrate_set_cache_size(int64_t value, Error **errp)
+{
+ MigrationState *s = migrate_get_current();
+
+ /* Check for truncation */
+ if (value != (size_t)value) {
+ error_set(errp, QERR_INVALID_PARAMETER_VALUE, "cache size",
+ "exceeding address space");
+ return;
+ }
+
+ s->xbzrle_cache_size = xbzrle_cache_resize(value);
+}
+
+int64_t qmp_query_migrate_cache_size(Error **errp)
+{
+ return migrate_xbzrle_cache_size();
+}
+
void qmp_migrate_set_speed(int64_t value, Error **errp)
{
MigrationState *s;
diff --git a/migration.h b/migration.h
index cdf6787..337e225 100644
--- a/migration.h
+++ b/migration.h
@@ -108,4 +108,6 @@ int xbzrle_decode_buffer(uint8_t *src, int slen, uint8_t *dst, int dlen);
int migrate_use_xbzrle(void);
int64_t migrate_xbzrle_cache_size(void);
+int64_t xbzrle_cache_resize(int64_t new_size);
+
#endif
diff --git a/monitor.c b/monitor.c
index 5121277..14ab1c0 100644
--- a/monitor.c
+++ b/monitor.c
@@ -2669,6 +2669,13 @@ static mon_cmd_t info_cmds[] = {
.mhandler.info = hmp_info_migrate_capabilities,
},
{
+ .name = "migrate_cache_size",
+ .args_type = "",
+ .params = "",
+ .help = "show current migration xbzrle cache size",
+ .mhandler.info = hmp_info_migrate_cache_size,
+ },
+ {
.name = "balloon",
.args_type = "",
.params = "",
diff --git a/qapi-schema.json b/qapi-schema.json
index 4909058..c899032 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -1392,6 +1392,33 @@
{ 'command': 'migrate_set_speed', 'data': {'value': 'int'} }
##
+# @migrate-set-cache-size
+#
+# Set XBZRLE cache size
+#
+# @value: cache size in bytes
+#
+# The size will be rounded down to the nearest power of 2.
+# The cache size can be modified before and during ongoing migration
+#
+# Returns: nothing on success
+#
+# Since: 1.2
+##
+{ 'command': 'migrate-set-cache-size', 'data': {'value': 'int'} }
+
+##
+# @query-migrate-cache-size
+#
+# query XBZRLE cache size
+#
+# Returns: XBZRLE cache size in bytes
+#
+# Since: 1.2
+##
+{ 'command': 'query-migrate-cache-size', 'returns': 'int' }
+
+##
# @ObjectPropertyInfo:
#
# @name: the name of the property
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 1f488f4..2672a73 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -520,6 +520,50 @@ Example:
<- { "return": {} }
EQMP
+{
+ .name = "migrate-set-cache-size",
+ .args_type = "value:o",
+ .mhandler.cmd_new = qmp_marshal_input_migrate_set_cache_size,
+ },
+
+SQMP
+migrate-set-cache-size
+---------------------
+
+Set cache size to be used by XBZRLE migration, the cache size will be rounded
+down to the nearest power of 2
+
+Arguments:
+
+- "value": cache size in bytes (json-int)
+
+Example:
+
+-> { "execute": "migrate-set-cache-size", "arguments": { "value": 536870912 } }
+<- { "return": {} }
+
+EQMP
+ {
+ .name = "query-migrate-cache-size",
+ .args_type = "",
+ .mhandler.cmd_new = qmp_marshal_input_query_migrate_cache_size,
+ },
+
+SQMP
+query-migrate-cache-size
+---------------------
+
+Show cache size to be used by XBZRLE migration
+
+returns a json-object with the following information:
+- "size" : json-int
+
+Example:
+
+-> { "execute": "query-migrate-cache-size" }
+<- { "return": { "67108864" } }
+
+EQMP
{
.name = "migrate_set_speed",
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 09/12] Add migration accounting for normal and duplicate pages
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (7 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats Orit Wasserman
` (2 subsequent siblings)
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
arch_init.c | 38 ++++++++++++++++++++++++++++++++++++++
hmp.c | 6 ++++++
migration.c | 6 ++++++
migration.h | 5 +++++
qapi-schema.json | 11 +++++++++--
qmp-commands.hx | 26 +++++++++++++++++++++++---
6 files changed, 87 insertions(+), 5 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index d709ccb..7f12317 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -199,6 +199,40 @@ int64_t xbzrle_cache_resize(int64_t new_size)
return pow2floor(new_size);
}
+/* accounting for migration statistics */
+typedef struct AccountingInfo {
+ uint64_t dup_pages;
+ uint64_t norm_pages;
+ uint64_t iterations;
+} AccountingInfo;
+
+static AccountingInfo acct_info;
+
+static void acct_clear(void)
+{
+ memset(&acct_info, 0, sizeof(acct_info));
+}
+
+uint64_t dup_mig_bytes_transferred(void)
+{
+ return acct_info.dup_pages * TARGET_PAGE_SIZE;
+}
+
+uint64_t dup_mig_pages_transferred(void)
+{
+ return acct_info.dup_pages;
+}
+
+uint64_t norm_mig_bytes_transferred(void)
+{
+ return acct_info.norm_pages * TARGET_PAGE_SIZE;
+}
+
+uint64_t norm_mig_pages_transferred(void)
+{
+ return acct_info.norm_pages;
+}
+
static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
int cont, int flag)
{
@@ -293,6 +327,7 @@ static int ram_save_block(QEMUFile *f)
p = memory_region_get_ram_ptr(mr) + offset;
if (is_dup_page(p)) {
+ acct_info.dup_pages++;
save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_COMPRESS);
qemu_put_byte(f, *p);
bytes_sent = 1;
@@ -308,6 +343,7 @@ static int ram_save_block(QEMUFile *f)
save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_PAGE);
qemu_put_buffer(f, p, TARGET_PAGE_SIZE);
bytes_sent = TARGET_PAGE_SIZE;
+ acct_info.norm_pages++;
}
/* if page is unmodified, continue to the next */
@@ -429,6 +465,7 @@ static int ram_save_setup(QEMUFile *f, void *opaque)
}
XBZRLE.encoded_buf = g_malloc0(TARGET_PAGE_SIZE);
XBZRLE.current_buf = g_malloc(TARGET_PAGE_SIZE);
+ acct_clear();
}
/* Make sure all dirty bits are set */
@@ -477,6 +514,7 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
break;
}
bytes_transferred += bytes_sent;
+ acct_info.iterations++;
/* we want to check in the 1st loop, just in case it was the 1st time
and we had to sync the dirty bitmap.
qemu_get_clock_ns() is a bit expensive, so we only check each some
diff --git a/hmp.c b/hmp.c
index 42cbfb9..1fd75ab 100644
--- a/hmp.c
+++ b/hmp.c
@@ -160,6 +160,12 @@ void hmp_info_migrate(Monitor *mon)
info->ram->total >> 10);
monitor_printf(mon, "total time: %" PRIu64 " milliseconds\n",
info->ram->total_time);
+ monitor_printf(mon, "duplicate: %" PRIu64 " pages\n",
+ info->ram->duplicate);
+ monitor_printf(mon, "normal: %" PRIu64 " pages\n",
+ info->ram->normal);
+ monitor_printf(mon, "normal bytes: %" PRIu64 " kbytes\n",
+ info->ram->normal_bytes >> 10);
}
if (info->has_disk) {
diff --git a/migration.c b/migration.c
index a6c0072..ab5e09d 100644
--- a/migration.c
+++ b/migration.c
@@ -161,6 +161,9 @@ MigrationInfo *qmp_query_migrate(Error **errp)
info->ram->total = ram_bytes_total();
info->ram->total_time = qemu_get_clock_ms(rt_clock)
- s->total_time;
+ info->ram->duplicate = dup_mig_pages_transferred();
+ info->ram->normal = norm_mig_pages_transferred();
+ info->ram->normal_bytes = norm_mig_bytes_transferred();
if (blk_mig_active()) {
info->has_disk = true;
@@ -180,6 +183,9 @@ MigrationInfo *qmp_query_migrate(Error **errp)
info->ram->remaining = 0;
info->ram->total = ram_bytes_total();
info->ram->total_time = s->total_time;
+ info->ram->duplicate = dup_mig_pages_transferred();
+ info->ram->normal = norm_mig_pages_transferred();
+ info->ram->normal_bytes = norm_mig_bytes_transferred();
break;
case MIG_STATE_ERROR:
info->has_status = true;
diff --git a/migration.h b/migration.h
index 337e225..e4a7cd7 100644
--- a/migration.h
+++ b/migration.h
@@ -87,6 +87,11 @@ uint64_t ram_bytes_total(void);
extern SaveVMHandlers savevm_ram_handlers;
+uint64_t dup_mig_bytes_transferred(void);
+uint64_t dup_mig_pages_transferred(void);
+uint64_t norm_mig_bytes_transferred(void);
+uint64_t norm_mig_pages_transferred(void);
+
/**
* @migrate_add_blocker - prevent migration from proceeding
*
diff --git a/qapi-schema.json b/qapi-schema.json
index c899032..8cc40e2 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -264,11 +264,18 @@
# migration has ended, it returns the total migration
# time. (since 1.2)
#
-# Since: 0.14.0.
+# @duplicate: number of duplicate pages (since 1.2)
+#
+# @normal : number of normal pages (since 1.2)
+#
+# @normal-bytes : number of normal bytes sent (since 1.2)
+#
+# Since: 0.14.0
##
{ 'type': 'MigrationStats',
'data': {'transferred': 'int', 'remaining': 'int', 'total': 'int' ,
- 'total_time': 'int' } }
+ 'total_time': 'int', 'duplicate': 'int', 'normal': 'int',
+ 'normal-bytes': 'int' } }
##
# @MigrationInfo
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 2672a73..0087079 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -2120,6 +2120,9 @@ The main json-object contains the following:
- "transferred": amount transferred (json-int)
- "remaining": amount remaining (json-int)
- "total": total (json-int)
+ - "duplicate": number of duplicated pages (json-int)
+ - "normal" : number of normal pages transferred (json-int)
+ - "normal-bytes" : number of normal bytes transferred (json-int)
- "disk": only present if "status" is "active" and it is a block migration,
it is a json-object with the following disk information (in bytes):
- "transferred": amount transferred (json-int)
@@ -2136,7 +2139,18 @@ Examples:
2. Migration is done and has succeeded
-> { "execute": "query-migrate" }
-<- { "return": { "status": "completed" } }
+<- { "return": {
+ "status": "completed",
+ "ram":{
+ "transferred":123,
+ "remaining":123,
+ "total":246,
+ "duplicate":123,
+ "normal":123,
+ "normal-bytes":123456
+ }
+ }
+ }
3. Migration is done and has failed
@@ -2152,7 +2166,10 @@ Examples:
"ram":{
"transferred":123,
"remaining":123,
- "total":246
+ "total":246,
+ "duplicate":123,
+ "normal":123,
+ "normal-bytes":123456
}
}
}
@@ -2166,7 +2183,10 @@ Examples:
"ram":{
"total":1057024,
"remaining":1053304,
- "transferred":3720
+ "transferred":3720,
+ "duplicate":123,
+ "normal":123,
+ "normal-bytes":123456
},
"disk":{
"total":20971520,
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (8 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 09/12] Add migration accounting for normal and duplicate pages Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 18:15 ` Eric Blake
2012-08-06 17:51 ` [Qemu-devel] [PATCH 11/12] Add XBZRLE statistics Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 12/12] Restart optimization on stage3 update version Orit Wasserman
11 siblings, 1 reply; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, Orit Wasserman, chegu_vinod, avi, pbonzini, eblake
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
qapi-schema.json | 4 ++--
qmp-commands.hx | 6 ++++++
2 files changed, 8 insertions(+), 2 deletions(-)
diff --git a/qapi-schema.json b/qapi-schema.json
index 8cc40e2..3238403 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -260,7 +260,7 @@
#
# @total: total amount of bytes involved in the migration process
#
-# @total_time: tota0l amount of ms since migration started. If
+# @total-time: total amount of ms since migration started. If
# migration has ended, it returns the total migration
# time. (since 1.2)
#
@@ -274,7 +274,7 @@
##
{ 'type': 'MigrationStats',
'data': {'transferred': 'int', 'remaining': 'int', 'total': 'int' ,
- 'total_time': 'int', 'duplicate': 'int', 'normal': 'int',
+ 'total-time': 'int', 'duplicate': 'int', 'normal': 'int',
'normal-bytes': 'int' } }
##
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 0087079..1a1b2cf 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -2120,6 +2120,9 @@ The main json-object contains the following:
- "transferred": amount transferred (json-int)
- "remaining": amount remaining (json-int)
- "total": total (json-int)
+ - "total-time": total amount of ms since migration started. If
+ migration has ended, it returns the total migration time
+ (json-int)
- "duplicate": number of duplicated pages (json-int)
- "normal" : number of normal pages transferred (json-int)
- "normal-bytes" : number of normal bytes transferred (json-int)
@@ -2145,6 +2148,7 @@ Examples:
"transferred":123,
"remaining":123,
"total":246,
+ "total-time":12345,
"duplicate":123,
"normal":123,
"normal-bytes":123456
@@ -2167,6 +2171,7 @@ Examples:
"transferred":123,
"remaining":123,
"total":246,
+ "total-time":12345,
"duplicate":123,
"normal":123,
"normal-bytes":123456
@@ -2184,6 +2189,7 @@ Examples:
"total":1057024,
"remaining":1053304,
"transferred":3720,
+ "total-time":12345,
"duplicate":123,
"normal":123,
"normal-bytes":123456
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 11/12] Add XBZRLE statistics
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (9 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 12/12] Restart optimization on stage3 update version Orit Wasserman
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
arch_init.c | 28 ++++++++++++++++++++++++++++
hmp.c | 13 +++++++++++++
migration.c | 17 +++++++++++++++++
migration.h | 4 ++++
qapi-schema.json | 28 +++++++++++++++++++++++++++-
qmp-commands.hx | 34 +++++++++++++++++++++++++++++++++-
6 files changed, 122 insertions(+), 2 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 7f12317..5f864a6 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -204,6 +204,10 @@ typedef struct AccountingInfo {
uint64_t dup_pages;
uint64_t norm_pages;
uint64_t iterations;
+ uint64_t xbzrle_bytes;
+ uint64_t xbzrle_pages;
+ uint64_t xbzrle_cache_miss;
+ uint64_t xbzrle_overflows;
} AccountingInfo;
static AccountingInfo acct_info;
@@ -233,6 +237,26 @@ uint64_t norm_mig_pages_transferred(void)
return acct_info.norm_pages;
}
+uint64_t xbzrle_mig_bytes_transferred(void)
+{
+ return acct_info.xbzrle_bytes;
+}
+
+uint64_t xbzrle_mig_pages_transferred(void)
+{
+ return acct_info.xbzrle_pages;
+}
+
+uint64_t xbzrle_mig_pages_cache_miss(void)
+{
+ return acct_info.xbzrle_cache_miss;
+}
+
+uint64_t xbzrle_mig_pages_overflow(void)
+{
+ return acct_info.xbzrle_overflows;
+}
+
static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
int cont, int flag)
{
@@ -257,6 +281,7 @@ static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
if (!cache_is_cached(XBZRLE.cache, current_addr)) {
cache_insert(XBZRLE.cache, current_addr,
g_memdup(current_data, TARGET_PAGE_SIZE));
+ acct_info.xbzrle_cache_miss++;
return -1;
}
@@ -274,6 +299,7 @@ static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
return 0;
} else if (encoded_len == -1) {
DPRINTF("Overflow\n");
+ acct_info.xbzrle_overflows++;
/* update data in the cache */
memcpy(prev_cached_page, current_data, TARGET_PAGE_SIZE);
return -1;
@@ -288,6 +314,8 @@ static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
qemu_put_be16(f, encoded_len);
qemu_put_buffer(f, XBZRLE.encoded_buf, encoded_len);
bytes_sent = encoded_len + 1 + 2;
+ acct_info.xbzrle_pages++;
+ acct_info.xbzrle_bytes += bytes_sent;
return bytes_sent;
}
diff --git a/hmp.c b/hmp.c
index 1fd75ab..1854280 100644
--- a/hmp.c
+++ b/hmp.c
@@ -177,6 +177,19 @@ void hmp_info_migrate(Monitor *mon)
info->disk->total >> 10);
}
+ if (info->has_xbzrle_cache) {
+ monitor_printf(mon, "cache size: %" PRIu64 " bytes\n",
+ info->xbzrle_cache->cache_size);
+ monitor_printf(mon, "xbzrle transferred: %" PRIu64 " kbytes\n",
+ info->xbzrle_cache->bytes >> 10);
+ monitor_printf(mon, "xbzrle pages: %" PRIu64 " pages\n",
+ info->xbzrle_cache->pages);
+ monitor_printf(mon, "xbzrle cache miss: %" PRIu64 "\n",
+ info->xbzrle_cache->cache_miss);
+ monitor_printf(mon, "xbzrle overflow : %" PRIu64 "\n",
+ info->xbzrle_cache->overflow);
+ }
+
qapi_free_MigrationInfo(info);
qapi_free_MigrationCapabilityStatusList(caps);
}
diff --git a/migration.c b/migration.c
index ab5e09d..653a3c1 100644
--- a/migration.c
+++ b/migration.c
@@ -141,6 +141,19 @@ MigrationCapabilityStatusList *qmp_query_migrate_capabilities(Error **errp)
return head;
}
+static void get_xbzrle_cache_stats(MigrationInfo *info)
+{
+ if (migrate_use_xbzrle()) {
+ info->has_xbzrle_cache = true;
+ info->xbzrle_cache = g_malloc0(sizeof(*info->xbzrle_cache));
+ info->xbzrle_cache->cache_size = migrate_xbzrle_cache_size();
+ info->xbzrle_cache->bytes = xbzrle_mig_bytes_transferred();
+ info->xbzrle_cache->pages = xbzrle_mig_pages_transferred();
+ info->xbzrle_cache->cache_miss = xbzrle_mig_pages_cache_miss();
+ info->xbzrle_cache->overflow = xbzrle_mig_pages_overflow();
+ }
+}
+
MigrationInfo *qmp_query_migrate(Error **errp)
{
MigrationInfo *info = g_malloc0(sizeof(*info));
@@ -172,8 +185,12 @@ MigrationInfo *qmp_query_migrate(Error **errp)
info->disk->remaining = blk_mig_bytes_remaining();
info->disk->total = blk_mig_bytes_total();
}
+
+ get_xbzrle_cache_stats(info);
break;
case MIG_STATE_COMPLETED:
+ get_xbzrle_cache_stats(info);
+
info->has_status = true;
info->status = g_strdup("completed");
diff --git a/migration.h b/migration.h
index e4a7cd7..a9852fc 100644
--- a/migration.h
+++ b/migration.h
@@ -91,6 +91,10 @@ uint64_t dup_mig_bytes_transferred(void);
uint64_t dup_mig_pages_transferred(void);
uint64_t norm_mig_bytes_transferred(void);
uint64_t norm_mig_pages_transferred(void);
+uint64_t xbzrle_mig_bytes_transferred(void);
+uint64_t xbzrle_mig_pages_transferred(void);
+uint64_t xbzrle_mig_pages_overflow(void);
+uint64_t xbzrle_mig_pages_cache_miss(void);
/**
* @migrate_add_blocker - prevent migration from proceeding
diff --git a/qapi-schema.json b/qapi-schema.json
index 3238403..b7d5366 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -278,6 +278,27 @@
'normal-bytes': 'int' } }
##
+# @XBZRLECacheStats
+#
+# Detailed XBZRLE migration cache statistics
+#
+# @cache-size: XBZRLE cache size
+#
+# @bytes: amount of bytes already transferred to the target VM
+#
+# @pages: amount of pages transferred to the target VM
+#
+# @cache-miss: number of cache miss
+#
+# @overflow: number of overflows
+#
+# Since: 1.2
+##
+{ 'type': 'XBZRLECacheStats',
+ 'data': {'cache-size': 'int', 'bytes': 'int', 'pages': 'int',
+ 'cache-miss': 'int', 'overflow': 'int' } }
+
+##
# @MigrationInfo
#
# Information about current migration process.
@@ -295,11 +316,16 @@
# status, only returned if status is 'active' and it is a block
# migration
#
+# @xbzrle-cache: #optional @XBZRLECacheStats containing detailed XBZRLE
+# migration statistics, only returned if XBZRLE feature is on and
+# status is 'active' or 'completed' (since 1.2)
+#
# Since: 0.14.0
##
{ 'type': 'MigrationInfo',
'data': {'*status': 'str', '*ram': 'MigrationStats',
- '*disk': 'MigrationStats'} }
+ '*disk': 'MigrationStats',
+ '*xbzrle-cache': 'XBZRLECacheStats'} }
##
# @query-migrate
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 1a1b2cf..73df8f9 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -2131,7 +2131,13 @@ The main json-object contains the following:
- "transferred": amount transferred (json-int)
- "remaining": amount remaining (json-int)
- "total": total (json-int)
-
+- "xbzrle-cache": only present if XBZRLE is active.
+ It is a json-object with the following XBZRLE information:
+ - "cache-size": XBZRLE cache size
+ - "bytes": total XBZRLE bytes transferred
+ - "pages": number of XBZRLE compressed pages
+ - "cache-miss": number of cache misses
+ - "overflow": number of XBZRLE overflows
Examples:
1. Before the first migration
@@ -2202,6 +2208,32 @@ Examples:
}
}
+6. Migration is being performed and XBZRLE is active:
+
+-> { "execute": "query-migrate" }
+<- {
+ "return":{
+ "status":"active",
+ "capabilities" : [ { "capability": "xbzrle", "state" : true } ],
+ "ram":{
+ "total":1057024,
+ "remaining":1053304,
+ "transferred":3720,
+ "total-time":12345,
+ "duplicate":10,
+ "normal":3333,
+ "normal-bytes":3412992
+ },
+ "xbzrle-cache":{
+ "cache-size":67108864,
+ "bytes":20971520,
+ "pages":2444343,
+ "cache-miss":2244,
+ "overflow":34434
+ }
+ }
+ }
+
EQMP
{
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 12/12] Restart optimization on stage3 update version
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
` (10 preceding siblings ...)
2012-08-06 17:51 ` [Qemu-devel] [PATCH 11/12] Add XBZRLE statistics Orit Wasserman
@ 2012-08-06 17:51 ` Orit Wasserman
11 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 17:51 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, stefanha, mdroth, lcapitulino,
blauwirbel, chegu_vinod, avi, pbonzini, eblake
From: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
---
arch_init.c | 24 +++++++++++++++---------
1 files changed, 15 insertions(+), 9 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 5f864a6..5181953 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -273,14 +273,16 @@ static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
ram_addr_t current_addr, RAMBlock *block,
- ram_addr_t offset, int cont)
+ ram_addr_t offset, int cont, bool last_stage)
{
int encoded_len = 0, bytes_sent = -1;
uint8_t *prev_cached_page;
if (!cache_is_cached(XBZRLE.cache, current_addr)) {
- cache_insert(XBZRLE.cache, current_addr,
- g_memdup(current_data, TARGET_PAGE_SIZE));
+ if (!last_stage) {
+ cache_insert(XBZRLE.cache, current_addr,
+ g_memdup(current_data, TARGET_PAGE_SIZE));
+ }
acct_info.xbzrle_cache_miss++;
return -1;
}
@@ -306,7 +308,9 @@ static int save_xbzrle_page(QEMUFile *f, uint8_t *current_data,
}
/* we need to update the data in the cache, in order to get the same data */
- memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
+ if (!last_stage) {
+ memcpy(prev_cached_page, XBZRLE.current_buf, TARGET_PAGE_SIZE);
+ }
/* Send XBZRLE based compressed page */
save_block_hdr(f, block, offset, cont, RAM_SAVE_FLAG_XBZRLE);
@@ -331,7 +335,7 @@ static ram_addr_t last_offset;
* n: the amount of bytes written in other case
*/
-static int ram_save_block(QEMUFile *f)
+static int ram_save_block(QEMUFile *f, bool last_stage)
{
RAMBlock *block = last_block;
ram_addr_t offset = last_offset;
@@ -362,8 +366,10 @@ static int ram_save_block(QEMUFile *f)
} else if (migrate_use_xbzrle()) {
current_addr = block->offset + offset;
bytes_sent = save_xbzrle_page(f, p, current_addr, block,
- offset, cont);
- p = get_cached_data(XBZRLE.cache, current_addr);
+ offset, cont, last_stage);
+ if (!last_stage) {
+ p = get_cached_data(XBZRLE.cache, current_addr);
+ }
}
/* either we didn't send yet (we may have had XBZRLE overflow) */
@@ -536,7 +542,7 @@ static int ram_save_iterate(QEMUFile *f, void *opaque)
while ((ret = qemu_file_rate_limit(f)) == 0) {
int bytes_sent;
- bytes_sent = ram_save_block(f);
+ bytes_sent = ram_save_block(f, false);
/* no more blocks to sent */
if (bytes_sent < 0) {
break;
@@ -598,7 +604,7 @@ static int ram_save_complete(QEMUFile *f, void *opaque)
while (true) {
int bytes_sent;
- bytes_sent = ram_save_block(f);
+ bytes_sent = ram_save_block(f, true);
/* no more blocks to sent */
if (bytes_sent < 0) {
break;
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
* Re: [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command
2012-08-06 17:51 ` [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command Orit Wasserman
@ 2012-08-06 18:13 ` Eric Blake
2012-08-06 18:24 ` Orit Wasserman
0 siblings, 1 reply; 17+ messages in thread
From: Eric Blake @ 2012-08-06 18:13 UTC (permalink / raw)
To: Orit Wasserman
Cc: peter.maydell, aliguori, quintela, stefanha, qemu-devel, mdroth,
blauwirbel, Petter Svard, Benoit Hudzia, avi, Aidan Shribman,
pbonzini, lcapitulino, chegu_vinod
[-- Attachment #1: Type: text/plain, Size: 1439 bytes --]
On 08/06/2012 11:51 AM, Orit Wasserman wrote:
> Change XBZRLE cache size in bytes (the size should be a power of 2, it will be
> rounded down to the nearest power of 2).
> If XBZRLE cache size is too small there will be many cache miss.
>
> New query-migrate-cache-size QMP command and 'info migrate_cache_size' HMP
> command to query cache value.
>
> Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
> Signed-off-by: Petter Svard <petters@cs.umu.se>
> Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
> Signed-off-by: Orit Wasserman <owasserm@redhat.com>
> +#
> +# Since: 1.2
> +##
> +{ 'command': 'query-migrate-cache-size', 'returns': 'int' }
This is the first 'returns':'int' in the schema, but we do have a
'returns':'str' for comparison of returning a native type instead of a
further struct.
> +Example:
> +
> +-> { "execute": "query-migrate-cache-size" }
> +<- { "return": { "67108864" } }
Invalid JSON. Inside an {}, everything has to be in name:value pairs,
but you gave no name to your string. But given the above, you aren't
returning a string, but an int. I think this should be:
{ "return": 67108864 }
That's based solely off my reading of the schema, and not of the code
itself, so please actually execute your code, to see what really came
over the wire.
--
Eric Blake eblake@redhat.com +1-919-301-3266
Libvirt virtualization library http://libvirt.org
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 620 bytes --]
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats
2012-08-06 17:51 ` [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats Orit Wasserman
@ 2012-08-06 18:15 ` Eric Blake
0 siblings, 0 replies; 17+ messages in thread
From: Eric Blake @ 2012-08-06 18:15 UTC (permalink / raw)
To: Orit Wasserman
Cc: peter.maydell, aliguori, quintela, stefanha, qemu-devel, mdroth,
blauwirbel, avi, pbonzini, lcapitulino, chegu_vinod
[-- Attachment #1: Type: text/plain, Size: 525 bytes --]
On 08/06/2012 11:51 AM, Orit Wasserman wrote:
> Signed-off-by: Orit Wasserman <owasserm@redhat.com>
Sparse on the commit message. You need to document why the change is
safe (namely, call out which commit introduced 'total_time' in the first
place, and the fact that the commit is not part of 1.1, so we are fixing
an unreleased naming glitch). Also, I would suggest rebasing this to
occur before 9/11.
--
Eric Blake eblake@redhat.com +1-919-301-3266
Libvirt virtualization library http://libvirt.org
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 620 bytes --]
^ permalink raw reply [flat|nested] 17+ messages in thread
* Re: [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command
2012-08-06 18:13 ` Eric Blake
@ 2012-08-06 18:24 ` Orit Wasserman
0 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 18:24 UTC (permalink / raw)
To: Eric Blake
Cc: peter.maydell, aliguori, quintela, stefanha, qemu-devel, mdroth,
blauwirbel, Petter Svard, Benoit Hudzia, avi, Aidan Shribman,
pbonzini, lcapitulino, chegu_vinod
On 08/06/2012 09:13 PM, Eric Blake wrote:
> On 08/06/2012 11:51 AM, Orit Wasserman wrote:
>> Change XBZRLE cache size in bytes (the size should be a power of 2, it will be
>> rounded down to the nearest power of 2).
>> If XBZRLE cache size is too small there will be many cache miss.
>>
>> New query-migrate-cache-size QMP command and 'info migrate_cache_size' HMP
>> command to query cache value.
>>
>> Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
>> Signed-off-by: Petter Svard <petters@cs.umu.se>
>> Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
>> Signed-off-by: Orit Wasserman <owasserm@redhat.com>
>
>> +#
>> +# Since: 1.2
>> +##
>> +{ 'command': 'query-migrate-cache-size', 'returns': 'int' }
>
> This is the first 'returns':'int' in the schema, but we do have a
> 'returns':'str' for comparison of returning a native type instead of a
> further struct.
>
>> +Example:
>> +
>> +-> { "execute": "query-migrate-cache-size" }
>> +<- { "return": { "67108864" } }
>
> Invalid JSON. Inside an {}, everything has to be in name:value pairs,
> but you gave no name to your string. But given the above, you aren't
> returning a string, but an int. I think this should be:
>
> { "return": 67108864 }
>
> That's based solely off my reading of the schema, and not of the code
> itself, so please actually execute your code, to see what really came
> over the wire.
>
you are right I will fix the example.
Orit
^ permalink raw reply [flat|nested] 17+ messages in thread
* [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command
2012-08-06 18:42 [Qemu-devel] [PATCH 00/12] Migration next v12 Orit Wasserman
@ 2012-08-06 18:42 ` Orit Wasserman
0 siblings, 0 replies; 17+ messages in thread
From: Orit Wasserman @ 2012-08-06 18:42 UTC (permalink / raw)
To: qemu-devel
Cc: peter.maydell, aliguori, quintela, Petter Svard, stefanha, mdroth,
Benoit Hudzia, lcapitulino, blauwirbel, Orit Wasserman,
chegu_vinod, avi, Aidan Shribman, pbonzini, eblake
Change XBZRLE cache size in bytes (the size should be a power of 2, it will be
rounded down to the nearest power of 2).
If XBZRLE cache size is too small there will be many cache miss.
New query-migrate-cache-size QMP command and 'info migrate_cache_size' HMP
command to query cache value.
Signed-off-by: Benoit Hudzia <benoit.hudzia@sap.com>
Signed-off-by: Petter Svard <petters@cs.umu.se>
Signed-off-by: Aidan Shribman <aidan.shribman@sap.com>
Signed-off-by: Orit Wasserman <owasserm@redhat.com>
---
arch_init.c | 10 ++++++++++
hmp-commands.hx | 22 ++++++++++++++++++++++
hmp.c | 19 +++++++++++++++++++
hmp.h | 2 ++
migration.c | 19 +++++++++++++++++++
migration.h | 2 ++
monitor.c | 7 +++++++
qapi-schema.json | 27 +++++++++++++++++++++++++++
qmp-commands.hx | 44 ++++++++++++++++++++++++++++++++++++++++++++
9 files changed, 152 insertions(+), 0 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 410ba4d..d709ccb 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -189,6 +189,16 @@ static struct {
.cache = NULL,
};
+
+int64_t xbzrle_cache_resize(int64_t new_size)
+{
+ if (XBZRLE.cache != NULL) {
+ return cache_resize(XBZRLE.cache, new_size / TARGET_PAGE_SIZE) *
+ TARGET_PAGE_SIZE;
+ }
+ return pow2floor(new_size);
+}
+
static void save_block_hdr(QEMUFile *f, RAMBlock *block, ram_addr_t offset,
int cont, int flag)
{
diff --git a/hmp-commands.hx b/hmp-commands.hx
index 8420794..6bb5b4a 100644
--- a/hmp-commands.hx
+++ b/hmp-commands.hx
@@ -829,6 +829,26 @@ STEXI
@item migrate_cancel
@findex migrate_cancel
Cancel the current VM migration.
+
+ETEXI
+
+ {
+ .name = "migrate_set_cache_size",
+ .args_type = "value:o",
+ .params = "value",
+ .help = "set cache size (in bytes) for XBZRLE migrations,"
+ "the cache size will be rounded down to the nearest "
+ "power of 2.\n"
+ "The cache size affects the number of cache misses."
+ "In case of a high cache miss ratio you need to increase"
+ " the cache size",
+ .mhandler.cmd = hmp_migrate_set_cache_size,
+ },
+
+STEXI
+@item migrate_set_cache_size @var{value}
+@findex migrate_set_cache_size
+Set cache size to @var{value} (in bytes) for xbzrle migrations.
ETEXI
{
@@ -1433,6 +1453,8 @@ show user network stack connection states
show migration status
@item info migrate_capabilities
show current migration capabilities
+@item info migrate_cache_size
+show current migration XBZRLE cache size
@item info balloon
show balloon information
@item info qtree
diff --git a/hmp.c b/hmp.c
index 6131120..42cbfb9 100644
--- a/hmp.c
+++ b/hmp.c
@@ -194,6 +194,12 @@ void hmp_info_migrate_capabilities(Monitor *mon)
qapi_free_MigrationCapabilityStatusList(caps);
}
+void hmp_info_migrate_cache_size(Monitor *mon)
+{
+ monitor_printf(mon, "xbzrel cache size: %" PRId64 " kbytes\n",
+ qmp_query_migrate_cache_size(NULL) >> 10);
+}
+
void hmp_info_cpus(Monitor *mon)
{
CpuInfoList *cpu_list, *cpu;
@@ -762,6 +768,19 @@ void hmp_migrate_set_downtime(Monitor *mon, const QDict *qdict)
qmp_migrate_set_downtime(value, NULL);
}
+void hmp_migrate_set_cache_size(Monitor *mon, const QDict *qdict)
+{
+ int64_t value = qdict_get_int(qdict, "value");
+ Error *err = NULL;
+
+ qmp_migrate_set_cache_size(value, &err);
+ if (err) {
+ monitor_printf(mon, "%s\n", error_get_pretty(err));
+ error_free(err);
+ return;
+ }
+}
+
void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict)
{
int64_t value = qdict_get_int(qdict, "value");
diff --git a/hmp.h b/hmp.h
index 3390a92..6d6e53b 100644
--- a/hmp.h
+++ b/hmp.h
@@ -26,6 +26,7 @@ void hmp_info_chardev(Monitor *mon);
void hmp_info_mice(Monitor *mon);
void hmp_info_migrate(Monitor *mon);
void hmp_info_migrate_capabilities(Monitor *mon);
+void hmp_info_migrate_cache_size(Monitor *mon);
void hmp_info_cpus(Monitor *mon);
void hmp_info_block(Monitor *mon);
void hmp_info_blockstats(Monitor *mon);
@@ -53,6 +54,7 @@ void hmp_migrate_cancel(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_downtime(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_speed(Monitor *mon, const QDict *qdict);
void hmp_migrate_set_capability(Monitor *mon, const QDict *qdict);
+void hmp_migrate_set_cache_size(Monitor *mon, const QDict *qdict);
void hmp_set_password(Monitor *mon, const QDict *qdict);
void hmp_expire_password(Monitor *mon, const QDict *qdict);
void hmp_eject(Monitor *mon, const QDict *qdict);
diff --git a/migration.c b/migration.c
index 5a5333c..a6c0072 100644
--- a/migration.c
+++ b/migration.c
@@ -511,6 +511,25 @@ void qmp_migrate_cancel(Error **errp)
migrate_fd_cancel(migrate_get_current());
}
+void qmp_migrate_set_cache_size(int64_t value, Error **errp)
+{
+ MigrationState *s = migrate_get_current();
+
+ /* Check for truncation */
+ if (value != (size_t)value) {
+ error_set(errp, QERR_INVALID_PARAMETER_VALUE, "cache size",
+ "exceeding address space");
+ return;
+ }
+
+ s->xbzrle_cache_size = xbzrle_cache_resize(value);
+}
+
+int64_t qmp_query_migrate_cache_size(Error **errp)
+{
+ return migrate_xbzrle_cache_size();
+}
+
void qmp_migrate_set_speed(int64_t value, Error **errp)
{
MigrationState *s;
diff --git a/migration.h b/migration.h
index cdf6787..337e225 100644
--- a/migration.h
+++ b/migration.h
@@ -108,4 +108,6 @@ int xbzrle_decode_buffer(uint8_t *src, int slen, uint8_t *dst, int dlen);
int migrate_use_xbzrle(void);
int64_t migrate_xbzrle_cache_size(void);
+int64_t xbzrle_cache_resize(int64_t new_size);
+
#endif
diff --git a/monitor.c b/monitor.c
index 5121277..14ab1c0 100644
--- a/monitor.c
+++ b/monitor.c
@@ -2669,6 +2669,13 @@ static mon_cmd_t info_cmds[] = {
.mhandler.info = hmp_info_migrate_capabilities,
},
{
+ .name = "migrate_cache_size",
+ .args_type = "",
+ .params = "",
+ .help = "show current migration xbzrle cache size",
+ .mhandler.info = hmp_info_migrate_cache_size,
+ },
+ {
.name = "balloon",
.args_type = "",
.params = "",
diff --git a/qapi-schema.json b/qapi-schema.json
index 4909058..c899032 100644
--- a/qapi-schema.json
+++ b/qapi-schema.json
@@ -1392,6 +1392,33 @@
{ 'command': 'migrate_set_speed', 'data': {'value': 'int'} }
##
+# @migrate-set-cache-size
+#
+# Set XBZRLE cache size
+#
+# @value: cache size in bytes
+#
+# The size will be rounded down to the nearest power of 2.
+# The cache size can be modified before and during ongoing migration
+#
+# Returns: nothing on success
+#
+# Since: 1.2
+##
+{ 'command': 'migrate-set-cache-size', 'data': {'value': 'int'} }
+
+##
+# @query-migrate-cache-size
+#
+# query XBZRLE cache size
+#
+# Returns: XBZRLE cache size in bytes
+#
+# Since: 1.2
+##
+{ 'command': 'query-migrate-cache-size', 'returns': 'int' }
+
+##
# @ObjectPropertyInfo:
#
# @name: the name of the property
diff --git a/qmp-commands.hx b/qmp-commands.hx
index 1f488f4..50576bf 100644
--- a/qmp-commands.hx
+++ b/qmp-commands.hx
@@ -520,6 +520,50 @@ Example:
<- { "return": {} }
EQMP
+{
+ .name = "migrate-set-cache-size",
+ .args_type = "value:o",
+ .mhandler.cmd_new = qmp_marshal_input_migrate_set_cache_size,
+ },
+
+SQMP
+migrate-set-cache-size
+---------------------
+
+Set cache size to be used by XBZRLE migration, the cache size will be rounded
+down to the nearest power of 2
+
+Arguments:
+
+- "value": cache size in bytes (json-int)
+
+Example:
+
+-> { "execute": "migrate-set-cache-size", "arguments": { "value": 536870912 } }
+<- { "return": {} }
+
+EQMP
+ {
+ .name = "query-migrate-cache-size",
+ .args_type = "",
+ .mhandler.cmd_new = qmp_marshal_input_query_migrate_cache_size,
+ },
+
+SQMP
+query-migrate-cache-size
+---------------------
+
+Show cache size to be used by XBZRLE migration
+
+returns a json-object with the following information:
+- "size" : json-int
+
+Example:
+
+-> { "execute": "query-migrate-cache-size" }
+<- { "return": 67108864 }
+
+EQMP
{
.name = "migrate_set_speed",
--
1.7.7.6
^ permalink raw reply related [flat|nested] 17+ messages in thread
end of thread, other threads:[~2012-08-06 18:43 UTC | newest]
Thread overview: 17+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-08-06 17:51 [Qemu-devel] [PATCH 00/12] Migration next v11 Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 01/12] Add migration capabilities Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 02/12] Add migrate-set-capabilities Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 03/12] Add XBZRLE documentation Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 04/12] Add cache handling functions Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 05/12] Add uleb encoding/decoding functions Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 06/12] Add xbzrle_encode_buffer and xbzrle_decode_buffer functions Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 07/12] Add XBZRLE to ram_save_block and ram_save_live Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command Orit Wasserman
2012-08-06 18:13 ` Eric Blake
2012-08-06 18:24 ` Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 09/12] Add migration accounting for normal and duplicate pages Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 10/12] Change total_time to total-time in MigrationStats Orit Wasserman
2012-08-06 18:15 ` Eric Blake
2012-08-06 17:51 ` [Qemu-devel] [PATCH 11/12] Add XBZRLE statistics Orit Wasserman
2012-08-06 17:51 ` [Qemu-devel] [PATCH 12/12] Restart optimization on stage3 update version Orit Wasserman
-- strict thread matches above, loose matches on Subject: below --
2012-08-06 18:42 [Qemu-devel] [PATCH 00/12] Migration next v12 Orit Wasserman
2012-08-06 18:42 ` [Qemu-devel] [PATCH 08/12] Add migrate_set_cache_size command Orit Wasserman
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.