| Message ID | 1485327241-15104-3-git-send-email-zhang.zhanghailiang@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
* zhanghailiang (zhang.zhanghailiang@huawei.com) wrote: > There are several stages during loadvm/savevm process. In different stage, > migration incoming processes different types of sections. > We want to control these stages more accuracy, it will benefit COLO > performance, we don't have to save type of QEMU_VM_SECTION_START > sections everytime while do checkpoint, besides, we want to separate > the process of saving/loading memory and devices state. > > So we add three new helper functions: qemu_loadvm_state_begin(), > qemu_load_device_state() and qemu_savevm_live_state() to achieve > different process during migration. > > Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com> > Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com> > Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> I don't think this one can go in without the patch that follows which uses these functions; we don't normally add functions without the patch that uses them. Dave > --- > include/sysemu/sysemu.h | 4 ++++ > migration/savevm.c | 41 +++++++++++++++++++++++++++++++++++++++++ > 2 files changed, 45 insertions(+) > > diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h > index ff8ffb5..d9214bf 100644 > --- a/include/sysemu/sysemu.h > +++ b/include/sysemu/sysemu.h > @@ -126,7 +126,11 @@ void qemu_savevm_send_postcopy_ram_discard(QEMUFile *f, const char *name, > uint64_t *start_list, > uint64_t *length_list); > > +void qemu_savevm_live_state(QEMUFile *f); > + > int qemu_loadvm_state(QEMUFile *f); > +int qemu_loadvm_state_begin(QEMUFile *f); > +int qemu_load_device_state(QEMUFile *f); > > extern int autostart; > > diff --git a/migration/savevm.c b/migration/savevm.c > index 92b3d6c..04dee4f 100644 > --- a/migration/savevm.c > +++ b/migration/savevm.c > @@ -1264,6 +1264,13 @@ done: > return ret; > } > > +void qemu_savevm_live_state(QEMUFile *f) > +{ > + /* save QEMU_VM_SECTION_END section */ > + qemu_savevm_state_complete_precopy(f, true); > + qemu_put_byte(f, QEMU_VM_EOF); > +} > + > static int qemu_save_device_state(QEMUFile *f) > { > SaveStateEntry *se; > @@ -2064,6 +2071,40 @@ int qemu_loadvm_state(QEMUFile *f) > return ret; > } > > +int qemu_loadvm_state_begin(QEMUFile *f) > +{ > + MigrationIncomingState *mis = migration_incoming_get_current(); > + Error *local_err = NULL; > + int ret; > + > + if (qemu_savevm_state_blocked(&local_err)) { > + error_report_err(local_err); > + return -EINVAL; > + } > + /* Load QEMU_VM_SECTION_START section */ > + ret = qemu_loadvm_state_main(f, mis); > + if (ret < 0) { > + error_report("Failed to loadvm begin work: %d", ret); > + } > + return ret; > +} > + > +int qemu_load_device_state(QEMUFile *f) > +{ > + MigrationIncomingState *mis = migration_incoming_get_current(); > + int ret; > + > + /* Load QEMU_VM_SECTION_FULL section */ > + ret = qemu_loadvm_state_main(f, mis); > + if (ret < 0) { > + error_report("Failed to load device state: %d", ret); > + return ret; > + } > + > + cpu_synchronize_all_post_init(); > + return 0; > +} > + > void hmp_savevm(Monitor *mon, const QDict *qdict) > { > BlockDriverState *bs, *bs1; > -- > 1.8.3.1 > > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
"Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote: > * zhanghailiang (zhang.zhanghailiang@huawei.com) wrote: >> There are several stages during loadvm/savevm process. In different stage, >> migration incoming processes different types of sections. >> We want to control these stages more accuracy, it will benefit COLO >> performance, we don't have to save type of QEMU_VM_SECTION_START >> sections everytime while do checkpoint, besides, we want to separate >> the process of saving/loading memory and devices state. >> >> So we add three new helper functions: qemu_loadvm_state_begin(), >> qemu_load_device_state() and qemu_savevm_live_state() to achieve >> different process during migration. >> >> Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com> >> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com> >> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> > > I don't think this one can go in without the patch that follows which > uses these functions; we don't normally add functions > without the patch that uses them. Agreed. If you want to add functions, you need new code that use them, or make old code use them. It is up to you. Thanks, Juan.
On 2017/1/31 18:05, Dr. David Alan Gilbert wrote: > * zhanghailiang (zhang.zhanghailiang@huawei.com) wrote: >> There are several stages during loadvm/savevm process. In different stage, >> migration incoming processes different types of sections. >> We want to control these stages more accuracy, it will benefit COLO >> performance, we don't have to save type of QEMU_VM_SECTION_START >> sections everytime while do checkpoint, besides, we want to separate >> the process of saving/loading memory and devices state. >> >> So we add three new helper functions: qemu_loadvm_state_begin(), >> qemu_load_device_state() and qemu_savevm_live_state() to achieve >> different process during migration. >> >> Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com> >> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com> >> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> > > I don't think this one can go in without the patch that follows which > uses these functions; we don't normally add functions > without the patch that uses them. > Yes, this series is prerequisite for the later COLO optimization, that will be a bigger series, that's why i picked these two patches, but since it is irrational, I'll post these series with the later one where we use this functions. Thanks. :) > Dave > >> --- >> include/sysemu/sysemu.h | 4 ++++ >> migration/savevm.c | 41 +++++++++++++++++++++++++++++++++++++++++ >> 2 files changed, 45 insertions(+) >> >> diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h >> index ff8ffb5..d9214bf 100644 >> --- a/include/sysemu/sysemu.h >> +++ b/include/sysemu/sysemu.h >> @@ -126,7 +126,11 @@ void qemu_savevm_send_postcopy_ram_discard(QEMUFile *f, const char *name, >> uint64_t *start_list, >> uint64_t *length_list); >> >> +void qemu_savevm_live_state(QEMUFile *f); >> + >> int qemu_loadvm_state(QEMUFile *f); >> +int qemu_loadvm_state_begin(QEMUFile *f); >> +int qemu_load_device_state(QEMUFile *f); >> >> extern int autostart; >> >> diff --git a/migration/savevm.c b/migration/savevm.c >> index 92b3d6c..04dee4f 100644 >> --- a/migration/savevm.c >> +++ b/migration/savevm.c >> @@ -1264,6 +1264,13 @@ done: >> return ret; >> } >> >> +void qemu_savevm_live_state(QEMUFile *f) >> +{ >> + /* save QEMU_VM_SECTION_END section */ >> + qemu_savevm_state_complete_precopy(f, true); >> + qemu_put_byte(f, QEMU_VM_EOF); >> +} >> + >> static int qemu_save_device_state(QEMUFile *f) >> { >> SaveStateEntry *se; >> @@ -2064,6 +2071,40 @@ int qemu_loadvm_state(QEMUFile *f) >> return ret; >> } >> >> +int qemu_loadvm_state_begin(QEMUFile *f) >> +{ >> + MigrationIncomingState *mis = migration_incoming_get_current(); >> + Error *local_err = NULL; >> + int ret; >> + >> + if (qemu_savevm_state_blocked(&local_err)) { >> + error_report_err(local_err); >> + return -EINVAL; >> + } >> + /* Load QEMU_VM_SECTION_START section */ >> + ret = qemu_loadvm_state_main(f, mis); >> + if (ret < 0) { >> + error_report("Failed to loadvm begin work: %d", ret); >> + } >> + return ret; >> +} >> + >> +int qemu_load_device_state(QEMUFile *f) >> +{ >> + MigrationIncomingState *mis = migration_incoming_get_current(); >> + int ret; >> + >> + /* Load QEMU_VM_SECTION_FULL section */ >> + ret = qemu_loadvm_state_main(f, mis); >> + if (ret < 0) { >> + error_report("Failed to load device state: %d", ret); >> + return ret; >> + } >> + >> + cpu_synchronize_all_post_init(); >> + return 0; >> +} >> + >> void hmp_savevm(Monitor *mon, const QDict *qdict) >> { >> BlockDriverState *bs, *bs1; >> -- >> 1.8.3.1 >> >> > -- > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK > > . >
On 2017/1/31 18:19, Juan Quintela wrote: > "Dr. David Alan Gilbert" <dgilbert@redhat.com> wrote: >> * zhanghailiang (zhang.zhanghailiang@huawei.com) wrote: >>> There are several stages during loadvm/savevm process. In different stage, >>> migration incoming processes different types of sections. >>> We want to control these stages more accuracy, it will benefit COLO >>> performance, we don't have to save type of QEMU_VM_SECTION_START >>> sections everytime while do checkpoint, besides, we want to separate >>> the process of saving/loading memory and devices state. >>> >>> So we add three new helper functions: qemu_loadvm_state_begin(), >>> qemu_load_device_state() and qemu_savevm_live_state() to achieve >>> different process during migration. >>> >>> Signed-off-by: zhanghailiang <zhang.zhanghailiang@huawei.com> >>> Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com> >>> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> >> >> I don't think this one can go in without the patch that follows which >> uses these functions; we don't normally add functions >> without the patch that uses them. > > Agreed. If you want to add functions, you need new code that use them, > or make old code use them. It is up to you. > Got it, thanks, I'll merge this series with the one where we use them. > Thanks, Juan. > > . >
diff --git a/include/sysemu/sysemu.h b/include/sysemu/sysemu.h index ff8ffb5..d9214bf 100644 --- a/include/sysemu/sysemu.h +++ b/include/sysemu/sysemu.h @@ -126,7 +126,11 @@ void qemu_savevm_send_postcopy_ram_discard(QEMUFile *f, const char *name, uint64_t *start_list, uint64_t *length_list); +void qemu_savevm_live_state(QEMUFile *f); + int qemu_loadvm_state(QEMUFile *f); +int qemu_loadvm_state_begin(QEMUFile *f); +int qemu_load_device_state(QEMUFile *f); extern int autostart; diff --git a/migration/savevm.c b/migration/savevm.c index 92b3d6c..04dee4f 100644 --- a/migration/savevm.c +++ b/migration/savevm.c @@ -1264,6 +1264,13 @@ done: return ret; } +void qemu_savevm_live_state(QEMUFile *f) +{ + /* save QEMU_VM_SECTION_END section */ + qemu_savevm_state_complete_precopy(f, true); + qemu_put_byte(f, QEMU_VM_EOF); +} + static int qemu_save_device_state(QEMUFile *f) { SaveStateEntry *se; @@ -2064,6 +2071,40 @@ int qemu_loadvm_state(QEMUFile *f) return ret; } +int qemu_loadvm_state_begin(QEMUFile *f) +{ + MigrationIncomingState *mis = migration_incoming_get_current(); + Error *local_err = NULL; + int ret; + + if (qemu_savevm_state_blocked(&local_err)) { + error_report_err(local_err); + return -EINVAL; + } + /* Load QEMU_VM_SECTION_START section */ + ret = qemu_loadvm_state_main(f, mis); + if (ret < 0) { + error_report("Failed to loadvm begin work: %d", ret); + } + return ret; +} + +int qemu_load_device_state(QEMUFile *f) +{ + MigrationIncomingState *mis = migration_incoming_get_current(); + int ret; + + /* Load QEMU_VM_SECTION_FULL section */ + ret = qemu_loadvm_state_main(f, mis); + if (ret < 0) { + error_report("Failed to load device state: %d", ret); + return ret; + } + + cpu_synchronize_all_post_init(); + return 0; +} + void hmp_savevm(Monitor *mon, const QDict *qdict) { BlockDriverState *bs, *bs1;