8:39 a.m.
First patch is just a refactoring.
Nikolay Shirokovskiy (2):
qemu: make explicit that formatting migratable imposes secure
qemu: don't update cpu unconditionally for migratable flag
src/qemu/qemu_domain.c | 27 ++++++++++++++-------------
src/qemu/qemu_domain.h | 4 ----
src/qemu/qemu_driver.c | 6 +-----
src/qemu/qemu_migration.c | 4 +---
src/qemu/qemu_migration_cookie.c | 1 -
5 files changed, 16 insertions(+), 26 deletions(-)
--
1.8.3.1
8:39 a.m.
New subject: [libvirt] [PATCH 1/2] qemu: make explicit that formatting migratable imposes secure
qemu code always set VIR_DOMAIN_XML_SECURE flags when VIR_DOMAIN_XML_MIGRATABLE
is set. At the same time qemu code itself does not analyse VIR_DOMAIN_XML_SECURE
presense in any way. Thus we can just append secure flag conditionally
down the stack before call to virDomainDefFormatInternal which actually checks
secure flag. This drops a couple of adhoc macros also.
I would even put appending secure flag if migratable is present into
virDomainDefFormatConvertXMLFlags but not all hypervisors drivers agree on
this point.
---
src/qemu/qemu_domain.c | 27 ++++++++++++++-------------
src/qemu/qemu_domain.h | 4 ----
src/qemu/qemu_driver.c | 5 ++---
src/qemu/qemu_migration.c | 4 +---
src/qemu/qemu_migration_cookie.c | 1 -
5 files changed, 17 insertions(+), 24 deletions(-)
diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
index 50b536e..6b458c1 100644
--- a/src/qemu/qemu_domain.c
+++ b/src/qemu/qemu_domain.c
@@ -4682,6 +4682,9 @@ qemuDomainDefFormatBufInternal(virQEMUDriverPtr driver,
}
format:
+ if (flags & VIR_DOMAIN_XML_MIGRATABLE)
+ flags |= VIR_DOMAIN_XML_SECURE;
+
ret = virDomainDefFormatInternal(def, caps,
virDomainDefFormatConvertXMLFlags(flags),
buf);
@@ -4762,10 +4765,10 @@ qemuDomainDefFormatLive(virQEMUDriverPtr driver,
bool inactive,
bool compatible)
{
- unsigned int flags = QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
+ unsigned int flags = VIR_DOMAIN_XML_UPDATE_CPU;
if (inactive)
- flags |= VIR_DOMAIN_XML_INACTIVE;
+ flags |= VIR_DOMAIN_XML_INACTIVE | VIR_DOMAIN_XML_SECURE;
if (compatible)
flags |= VIR_DOMAIN_XML_MIGRATABLE;
@@ -5212,8 +5215,8 @@ qemuDomainSnapshotWriteMetadata(virDomainObjPtr vm,
virUUIDFormat(vm->def->uuid, uuidstr);
newxml = virDomainSnapshotDefFormat(
uuidstr, snapshot->def, caps, xmlopt,
- virDomainDefFormatConvertXMLFlags(QEMU_DOMAIN_FORMAT_LIVE_FLAGS),
- 1);
+ virDomainDefFormatConvertXMLFlags(VIR_DOMAIN_XML_SECURE |
+ VIR_DOMAIN_XML_UPDATE_CPU), 1);
if (newxml == NULL)
return -1;
@@ -6263,9 +6266,6 @@ qemuDomainMigratableDefCheckABIStability(virQEMUDriverPtr driver,
}
-#define COPY_FLAGS (VIR_DOMAIN_XML_SECURE | \
- VIR_DOMAIN_XML_MIGRATABLE)
-
bool
qemuDomainDefCheckABIStability(virQEMUDriverPtr driver,
virDomainDefPtr src,
@@ -6275,8 +6275,10 @@ qemuDomainDefCheckABIStability(virQEMUDriverPtr driver,
virDomainDefPtr migratableDefDst = NULL;
bool ret = false;
- if (!(migratableDefSrc = qemuDomainDefCopy(driver, src, COPY_FLAGS)) ||
- !(migratableDefDst = qemuDomainDefCopy(driver, dst, COPY_FLAGS)))
+ if (!(migratableDefSrc = qemuDomainDefCopy(driver, src,
+ VIR_DOMAIN_XML_MIGRATABLE)) ||
+ !(migratableDefDst = qemuDomainDefCopy(driver, dst,
+ VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
ret = qemuDomainMigratableDefCheckABIStability(driver,
@@ -6300,9 +6302,10 @@ qemuDomainCheckABIStability(virQEMUDriverPtr driver,
char *xml = NULL;
bool ret = false;
- if (!(xml = qemuDomainFormatXML(driver, vm, COPY_FLAGS)) ||
+ if (!(xml = qemuDomainFormatXML(driver, vm, VIR_DOMAIN_XML_MIGRATABLE)) ||
!(migratableSrc = qemuDomainDefFromXML(driver, xml)) ||
- !(migratableDst = qemuDomainDefCopy(driver, dst, COPY_FLAGS)))
+ !(migratableDst = qemuDomainDefCopy(driver, dst,
+ VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
ret = qemuDomainMigratableDefCheckABIStability(driver,
@@ -6316,8 +6319,6 @@ qemuDomainCheckABIStability(virQEMUDriverPtr driver,
return ret;
}
-#undef COPY_FLAGS
-
bool
qemuDomainAgentAvailable(virDomainObjPtr vm,
diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h
index b291dc3..7c2f91a 100644
--- a/src/qemu/qemu_domain.h
+++ b/src/qemu/qemu_domain.h
@@ -39,10 +39,6 @@
# include "virobject.h"
# include "logging/log_manager.h"
-# define QEMU_DOMAIN_FORMAT_LIVE_FLAGS \
- (VIR_DOMAIN_XML_SECURE | \
- VIR_DOMAIN_XML_UPDATE_CPU)
-
# if ULONG_MAX == 4294967295
/* QEMU has a 64-bit limit, but we are limited by our historical choice of
* representing bandwidth in a long instead of a 64-bit int. */
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index 9a21995..d9dff93 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -6282,7 +6282,7 @@ qemuDomainSaveImageUpdateDef(virQEMUDriverPtr driver,
if (!(newdef_migr = qemuDomainDefCopy(driver,
newdef,
- QEMU_DOMAIN_FORMAT_LIVE_FLAGS |
+ VIR_DOMAIN_XML_UPDATE_CPU |
VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
@@ -6789,7 +6789,6 @@ qemuDomainSaveImageDefineXML(virConnectPtr conn, const char *path,
if (!(data->xml = qemuDomainDefFormatXML(driver, newdef,
VIR_DOMAIN_XML_INACTIVE |
- VIR_DOMAIN_XML_SECURE |
VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
@@ -6995,7 +6994,7 @@ static char
goto cleanup;
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
- flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
+ flags |= VIR_DOMAIN_XML_UPDATE_CPU;
ret = qemuDomainFormatXML(driver, vm, flags);
diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c
index 492815b..db0b4c5 100644
--- a/src/qemu/qemu_migration.c
+++ b/src/qemu/qemu_migration.c
@@ -2575,7 +2575,6 @@ qemuMigrationPrepareAny(virQEMUDriverPtr driver,
int hookret;
if (!(xml = qemuDomainDefFormatXML(driver, *def,
- VIR_DOMAIN_XML_SECURE |
VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
@@ -3639,7 +3638,6 @@ qemuMigrationRun(virQEMUDriverPtr driver,
} else {
virDomainDefPtr def = vm->newDef ? vm->newDef : vm->def;
if (!(persistDef = qemuDomainDefCopy(driver, def,
- VIR_DOMAIN_XML_SECURE |
VIR_DOMAIN_XML_MIGRATABLE)))
goto cleanup;
}
@@ -4108,7 +4106,7 @@ static int doPeer2PeerMigrate2(virQEMUDriverPtr driver,
* and pass it to Prepare2.
*/
if (!(dom_xml = qemuDomainFormatXML(driver, vm,
- QEMU_DOMAIN_FORMAT_LIVE_FLAGS |
+ VIR_DOMAIN_XML_UPDATE_CPU |
VIR_DOMAIN_XML_MIGRATABLE)))
return -1;
diff --git a/src/qemu/qemu_migration_cookie.c b/src/qemu/qemu_migration_cookie.c
index 4914c77..38213ae 100644
--- a/src/qemu/qemu_migration_cookie.c
+++ b/src/qemu/qemu_migration_cookie.c
@@ -738,7 +738,6 @@ qemuMigrationCookieXMLFormat(virQEMUDriverPtr driver,
if (qemuDomainDefFormatBuf(driver,
mig->persistent,
VIR_DOMAIN_XML_INACTIVE |
- VIR_DOMAIN_XML_SECURE |
VIR_DOMAIN_XML_MIGRATABLE,
buf) < 0)
return -1;
--
1.8.3.1
9:32 a.m.
New subject: [libvirt] [PATCH 1/2] qemu: make explicit that formatting migratable imposes secure
On Thu, Sep 21, 2017 at 16:39:38 +0300, Nikolay Shirokovskiy wrote:
qemu code always set VIR_DOMAIN_XML_SECURE flags when
VIR_DOMAIN_XML_MIGRATABLE
is set. At the same time qemu code itself does not analyse VIR_DOMAIN_XML_SECURE
presense in any way. Thus we can just append secure flag conditionally
down the stack before call to virDomainDefFormatInternal which actually checks
secure flag. This drops a couple of adhoc macros also.
I think the adhoc macros are actually useful. They make sure that all
places where a definition is copied or formatted in a certain way will
use the same flags. Even if it's just a single flag.
I would even put appending secure flag if migratable is present into
virDomainDefFormatConvertXMLFlags but not all hypervisors drivers agree on
this point.
...
@@ -6995,7 +6994,7 @@ static char
goto cleanup;
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
- flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
+ flags |= VIR_DOMAIN_XML_UPDATE_CPU;
Oh I se what was the local change now. The annoyingly slow mail delivery
from the mailing list caused me to see 2/2 without the rest of the
series.
Jirka
1:24 a.m.
New subject: [libvirt] [PATCH 1/2] qemu: make explicit that formatting migratable imposes secure
On 21.09.2017 17:32, Jiri Denemark wrote:
On Thu, Sep 21, 2017 at 16:39:38 +0300, Nikolay Shirokovskiy wrote:
> qemu code always set VIR_DOMAIN_XML_SECURE flags when VIR_DOMAIN_XML_MIGRATABLE
> is set. At the same time qemu code itself does not analyse VIR_DOMAIN_XML_SECURE
> presense in any way. Thus we can just append secure flag conditionally
> down the stack before call to virDomainDefFormatInternal which actually checks
> secure flag. This drops a couple of adhoc macros also.
I think the adhoc macros are actually useful. They make sure that all
places where a definition is copied or formatted in a certain way will
use the same flags. Even if it's just a single flag.
It is just flag QEMU_DOMAIN_FORMAT_LIVE_FLAGS was not very clear to me.
In one place it combined with VIR_DOMAIN_XML_INACTIVE which looks odd.
But the main purpose of this patch is not to remove macros but rather
to clarify the fact that migratable flag is always used with together
with secure flag.
Nikolay
> I would even put appending secure flag if migratable is present into
> virDomainDefFormatConvertXMLFlags but not all hypervisors drivers agree on
> this point.
...
> @@ -6995,7 +6994,7 @@ static char
> goto cleanup;
>
> if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
> - flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
> + flags |= VIR_DOMAIN_XML_UPDATE_CPU;
Oh I se what was the local change now. The annoyingly slow mail delivery
from the mailing list caused me to see 2/2 without the rest of the
series.
Jirka
8:39 a.m.
New subject: [libvirt] [PATCH 2/2] qemu: don't update cpu unconditionally for migratable flag
Imagine if we use 'virsh dumpxml --migratable' for offline domain
to get config to tweak before migration. Currenly cpu section will
be expanded, host-cpu mode turns into custom and migration fails
because of ABI check.
Looks like ABI check does not make much sence for offline migration
but we don't want host-cpu mode to turn into custom in the first place.
Using --migratable is reasonable in this case because this flags
makes changes for inactive configs too like removing automatically
added parts that old versions can not handle.
I suggest not to update cpu for inactive configs. This appoach is
coherent with the way migration works for inactive configs in case
it is not specified externally to migration: function qemuMigrationCookieXMLFormat
don't expand cpu too.
Adding flag for active configs is useless because qemuDomainFormatXML
will clear it in this case. Thus we can skip adding update flag
altogether.
---
I suppose next usage of libvirt in case one need to migrate with tweaked configs:
inactive domain:
virsh dumpxml DOM --migratable > dom.xml
virsh migrate DOM --offline --persistent --xml dom.xml
active domain:
virsh dumpxml DOM --migratable > active.xml
virsh dumpxml DOM --migratable --inactive > inactive.xml
virsh migrate DOM --persistent --xml active.xml --persistent-xml inactive.xml
src/qemu/qemu_driver.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index d9dff93..3223554 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -6993,9 +6993,6 @@ static char
if (qemuDomainUpdateCurrentMemorySize(driver, vm) < 0)
goto cleanup;
- if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
- flags |= VIR_DOMAIN_XML_UPDATE_CPU;
-
ret = qemuDomainFormatXML(driver, vm, flags);
cleanup:
--
1.8.3.1
9:24 a.m.
New subject: [libvirt] [PATCH 2/2] qemu: don't update cpu unconditionally for migratable flag
On Thu, Sep 21, 2017 at 16:39:39 +0300, Nikolay Shirokovskiy wrote:
Imagine if we use 'virsh dumpxml --migratable' for offline
domain
to get config to tweak before migration. Currenly cpu section will
be expanded, host-cpu mode turns into custom and migration fails
because of ABI check.
Looks like ABI check does not make much sence for offline migration
but we don't want host-cpu mode to turn into custom in the first place.
Using --migratable is reasonable in this case because this flags
makes changes for inactive configs too like removing automatically
added parts that old versions can not handle.
I suggest not to update cpu for inactive configs. This appoach is
coherent with the way migration works for inactive configs in case
it is not specified externally to migration: function qemuMigrationCookieXMLFormat
don't expand cpu too.
Adding flag for active configs is useless because qemuDomainFormatXML
will clear it in this case.
Not anymore, see commit v3.7.0-151-g06f75ff2cb.
...
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index d9dff93..3223554 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -6993,9 +6993,6 @@ static char
if (qemuDomainUpdateCurrentMemorySize(driver, vm) < 0)
goto cleanup;
- if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
- flags |= VIR_DOMAIN_XML_UPDATE_CPU;
-
ret = qemuDomainFormatXML(driver, vm, flags);
cleanup:
This area of code changed recently and it looks like the following now:
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
/* The CPU is already updated in the domain's live definition, we need to
* ignore the VIR_DOMAIN_XML_UPDATE_CPU flag.
*/
if (virDomainObjIsActive(vm) &&
!(flags & VIR_DOMAIN_DEF_FORMAT_INACTIVE))
flags &= ~VIR_DOMAIN_XML_UPDATE_CPU;
But even before the changes, the two lines you are removing were
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
since commit v0.10.2-133-g28f8dfdccc, i.e., for the last 5 years. I
guess you have some local changes applied to the libvirt tree.
Jirka
9:32 a.m.
New subject: [libvirt] [PATCH 2/2] qemu: don't update cpu unconditionally for migratable flag
On 21.09.2017 17:24, Jiri Denemark wrote:
On Thu, Sep 21, 2017 at 16:39:39 +0300, Nikolay Shirokovskiy wrote:
> Imagine if we use 'virsh dumpxml --migratable' for offline domain
> to get config to tweak before migration. Currenly cpu section will
> be expanded, host-cpu mode turns into custom and migration fails
> because of ABI check.
>
> Looks like ABI check does not make much sence for offline migration
> but we don't want host-cpu mode to turn into custom in the first place.
> Using --migratable is reasonable in this case because this flags
> makes changes for inactive configs too like removing automatically
> added parts that old versions can not handle.
>
> I suggest not to update cpu for inactive configs. This appoach is
> coherent with the way migration works for inactive configs in case
> it is not specified externally to migration: function qemuMigrationCookieXMLFormat
> don't expand cpu too.
>
> Adding flag for active configs is useless because qemuDomainFormatXML
> will clear it in this case.
Not anymore, see commit v3.7.0-151-g06f75ff2cb.
Is g06f75ff2cb commit in public libvirt tree? I can't see it even on
https://libvirt.org/git/?p=libvirt.git;a=summary.
...
> diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
> index d9dff93..3223554 100644
> --- a/src/qemu/qemu_driver.c
> +++ b/src/qemu/qemu_driver.c
> @@ -6993,9 +6993,6 @@ static char
> if (qemuDomainUpdateCurrentMemorySize(driver, vm) < 0)
> goto cleanup;
>
> - if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
> - flags |= VIR_DOMAIN_XML_UPDATE_CPU;
> -
> ret = qemuDomainFormatXML(driver, vm, flags);
>
> cleanup:
This area of code changed recently and it looks like the following now:
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
/* The CPU is already updated in the domain's live definition, we need to
* ignore the VIR_DOMAIN_XML_UPDATE_CPU flag.
*/
if (virDomainObjIsActive(vm) &&
!(flags & VIR_DOMAIN_DEF_FORMAT_INACTIVE))
flags &= ~VIR_DOMAIN_XML_UPDATE_CPU;
But even before the changes, the two lines you are removing were
if ((flags & VIR_DOMAIN_XML_MIGRATABLE))
flags |= QEMU_DOMAIN_FORMAT_LIVE_FLAGS;
since commit v0.10.2-133-g28f8dfdccc, i.e., for the last 5 years. I
guess you have some local changes applied to the libvirt tree.
))) The previous commit in series touch this place.
Nikolay
10:50 a.m.
New subject: [libvirt] [PATCH 2/2] qemu: don't update cpu unconditionally for migratable flag
On Thu, Sep 21, 2017 at 17:32:33 +0300, Nikolay Shirokovskiy wrote:
On 21.09.2017 17:24, Jiri Denemark wrote:
> On Thu, Sep 21, 2017 at 16:39:39 +0300, Nikolay Shirokovskiy wrote:
>> Imagine if we use 'virsh dumpxml --migratable' for offline domain
>> to get config to tweak before migration. Currenly cpu section will
>> be expanded, host-cpu mode turns into custom and migration fails
>> because of ABI check.
>>
>> Looks like ABI check does not make much sence for offline migration
>> but we don't want host-cpu mode to turn into custom in the first place.
>> Using --migratable is reasonable in this case because this flags
>> makes changes for inactive configs too like removing automatically
>> added parts that old versions can not handle.
>>
>> I suggest not to update cpu for inactive configs. This appoach is
>> coherent with the way migration works for inactive configs in case
>> it is not specified externally to migration: function
qemuMigrationCookieXMLFormat
>> don't expand cpu too.
>>
>> Adding flag for active configs is useless because qemuDomainFormatXML
>> will clear it in this case.
>
> Not anymore, see commit v3.7.0-151-g06f75ff2cb.
Is g06f75ff2cb commit in public libvirt tree? I can't see it even on
https://libvirt.org/git/?p=libvirt.git;a=summary.
Note that 'g' isn't really a hex digit as it's usual in commit hashes.
That's a tricky part of the git describe string.
https://libvirt.org/git/?p=libvirt.git;a=commit;h=06f75ff2cb
commit 06f75ff2cb292e2658b4f2f6949c700550006272
Author: Jiri Denemark <jdenemar(a)redhat.com>
AuthorDate: Fri Jun 30 16:55:20 2017 +0200
Commit: Jiri Denemark <jdenemar(a)redhat.com>
CommitDate: Thu Sep 21 15:27:39 2017 +0200
qemu: Don't update CPU when formatting live def
Since commit v2.2.0-199-g7ce711a30e libvirt stores an updated guest CPU
in domain's live definition and there's no need to update it every time
we want to format the definition. The commit itself tried to address
this in qemuDomainFormatXML, but forgot to fix qemuDomainDefFormatLive.
Not to mention that masking a previously set flag is only acceptable if
the flag was set by a public API user. Internally, libvirt should have
never set the flag in the first place.
https://bugzilla.redhat.com/show_bug.cgi?id=1485022
Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com>
1:18 a.m.
New subject: [libvirt] [PATCH 2/2] qemu: don't update cpu unconditionally for migratable flag
On 21.09.2017 18:50, Peter Krempa wrote:
On Thu, Sep 21, 2017 at 17:32:33 +0300, Nikolay Shirokovskiy wrote:
>
>
> On 21.09.2017 17:24, Jiri Denemark wrote:
>> On Thu, Sep 21, 2017 at 16:39:39 +0300, Nikolay Shirokovskiy wrote:
>>> Imagine if we use 'virsh dumpxml --migratable' for offline domain
>>> to get config to tweak before migration. Currenly cpu section will
>>> be expanded, host-cpu mode turns into custom and migration fails
>>> because of ABI check.
>>>
>>> Looks like ABI check does not make much sence for offline migration
>>> but we don't want host-cpu mode to turn into custom in the first place.
>>> Using --migratable is reasonable in this case because this flags
>>> makes changes for inactive configs too like removing automatically
>>> added parts that old versions can not handle.
>>>
>>> I suggest not to update cpu for inactive configs. This appoach is
>>> coherent with the way migration works for inactive configs in case
>>> it is not specified externally to migration: function
qemuMigrationCookieXMLFormat
>>> don't expand cpu too.
>>>
>>> Adding flag for active configs is useless because qemuDomainFormatXML
>>> will clear it in this case.
>>
>> Not anymore, see commit v3.7.0-151-g06f75ff2cb.
>
> Is g06f75ff2cb commit in public libvirt tree? I can't see it even on
https://libvirt.org/git/?p=libvirt.git;a=summary.
Note that 'g' isn't really a hex digit as it's usual in commit hashes.
That's a tricky part of the git describe string.
https://libvirt.org/git/?p=libvirt.git;a=commit;h=06f75ff2cb
commit 06f75ff2cb292e2658b4f2f6949c700550006272
Author: Jiri Denemark <jdenemar(a)redhat.com>
AuthorDate: Fri Jun 30 16:55:20 2017 +0200
Commit: Jiri Denemark <jdenemar(a)redhat.com>
CommitDate: Thu Sep 21 15:27:39 2017 +0200
qemu: Don't update CPU when formatting live def
Since commit v2.2.0-199-g7ce711a30e libvirt stores an updated guest CPU
in domain's live definition and there's no need to update it every time
we want to format the definition. The commit itself tried to address
this in qemuDomainFormatXML, but forgot to fix qemuDomainDefFormatLive.
Not to mention that masking a previously set flag is only acceptable if
the flag was set by a public API user. Internally, libvirt should have
never set the flag in the first place.
https://bugzilla.redhat.com/show_bug.cgi?id=1485022
Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com>
I see, thanx!
1:46 a.m.
On Thu, Sep 21, 2017 at 16:39:37 +0300, Nikolay Shirokovskiy wrote:
First patch is just a refactoring.
Nikolay Shirokovskiy (2):
qemu: make explicit that formatting migratable imposes secure
qemu: don't update cpu unconditionally for migratable flag
src/qemu/qemu_domain.c | 27 ++++++++++++++-------------
src/qemu/qemu_domain.h | 4 ----
src/qemu/qemu_driver.c | 6 +-----
src/qemu/qemu_migration.c | 4 +---
src/qemu/qemu_migration_cookie.c | 1 -
5 files changed, 16 insertions(+), 26 deletions(-)
It looks like the issue you were trying to fix by this series was
addressed in [1] which was pushed an hour ago.
Jirka
[1] https://www.redhat.com/archives/libvir-list/2017-September/msg00517.html
2:47 a.m.
On 22.09.2017 09:46, Jiri Denemark wrote:
On Thu, Sep 21, 2017 at 16:39:37 +0300, Nikolay Shirokovskiy wrote:
> First patch is just a refactoring.
>
> Nikolay Shirokovskiy (2):
> qemu: make explicit that formatting migratable imposes secure
> qemu: don't update cpu unconditionally for migratable flag
>
> src/qemu/qemu_domain.c | 27 ++++++++++++++-------------
> src/qemu/qemu_domain.h | 4 ----
> src/qemu/qemu_driver.c | 6 +-----
> src/qemu/qemu_migration.c | 4 +---
> src/qemu/qemu_migration_cookie.c | 1 -
> 5 files changed, 16 insertions(+), 26 deletions(-)
It looks like the issue you were trying to fix by this series was
addressed in [1] which was pushed an hour ago.
Jirka
[1] https://www.redhat.com/archives/libvir-list/2017-September/msg00517.html
True, [1] fixes issues I address in my series. But still please check out
the first patch of that series.
Nikolay
2654
days inactive
2655
days old
10 comments
3 participants
participants (3)
-
Jiri Denemark -
Nikolay Shirokovskiy -
Peter Krempa