12:47 a.m.
Hi all,
I'll resend this patchset because what sent yesterday was broken.
Sorry to be a nuisance.
--
This patchset enables us to configure inactive domains' CPU affinity setting.
This is the rebased version, but retains the version number.
*[PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
*[PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu driver
*[PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
*[PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin" command
Best regards,
Taku Izumi
12:54 a.m.
New subject: [libvirt] [PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
This patch introduces a new libvirt API (virDomainPinVcpuFlags).
Signd-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
include/libvirt/libvirt.h.in | 5 ++
src/driver.h | 7 +++
src/libvirt.c | 76 +++++++++++++++++++++++++++++++++++++++++++
src/libvirt_public.syms | 5 ++
4 files changed, 93 insertions(+)
Index: libvirt/include/libvirt/libvirt.h.in
===================================================================
--- libvirt.orig/include/libvirt/libvirt.h.in
+++ libvirt/include/libvirt/libvirt.h.in
@@ -1049,6 +1049,11 @@ int virDomainPinVcpu
unsigned int vcpu,
unsigned char *cpumap,
int maplen);
+int virDomainPinVcpuFlags (virDomainPtr domain,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags);
/**
* VIR_USE_CPU:
Index: libvirt/src/driver.h
===================================================================
--- libvirt.orig/src/driver.h
+++ libvirt/src/driver.h
@@ -230,6 +230,12 @@ typedef int
unsigned char *cpumap,
int maplen);
typedef int
+ (*virDrvDomainPinVcpuFlags) (virDomainPtr domain,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags);
+typedef int
(*virDrvDomainGetVcpus) (virDomainPtr domain,
virVcpuInfoPtr info,
int maxinfo,
@@ -671,6 +677,7 @@ struct _virDriver {
virDrvDomainSetVcpusFlags domainSetVcpusFlags;
virDrvDomainGetVcpusFlags domainGetVcpusFlags;
virDrvDomainPinVcpu domainPinVcpu;
+ virDrvDomainPinVcpuFlags domainPinVcpuFlags;
virDrvDomainGetVcpus domainGetVcpus;
virDrvDomainGetMaxVcpus domainGetMaxVcpus;
virDrvDomainGetSecurityLabel domainGetSecurityLabel;
Index: libvirt/src/libvirt.c
===================================================================
--- libvirt.orig/src/libvirt.c
+++ libvirt/src/libvirt.c
@@ -6709,6 +6709,82 @@ error:
}
/**
+ * virDomainPinVcpuFlags:
+ * @domain: pointer to domain object, or NULL for Domain0
+ * @vcpu: virtual CPU number
+ * @cpumap: pointer to a bit map of real CPUs (in 8-bit bytes) (IN)
+ * Each bit set to 1 means that corresponding CPU is usable.
+ * Bytes are stored in little-endian order: CPU0-7, 8-15...
+ * In each byte, lowest CPU number is least significant bit.
+ * @maplen: number of bytes in cpumap, from 1 up to size of CPU map in
+ * underlying virtualization system (Xen...).
+ * If maplen < size, missing bytes are set to zero.
+ * If maplen > size, failure code is returned.
+ * @flags: bitwise-OR of virDomainModificationImpac
+ *
+ * Dynamically change the real CPUs which can be allocated to a virtual CPU.
+ * This function requires privileged access to the hypervisor.
+ *
+ * @flags may include VIR_DOMAIN_AFFECT_LIVE or VIR_DOMAIN_AFFECT_CONFIG.
+ * Both flags may be set.
+ * If VIR_DOMAIN_AFFECT_LIVE is set, the change affects a running domain
+ * and may fail if domain is not alive.
+ * If VIR_DOMAIN_AFFECT_CONFIG is set, the change affects persistent state,
+ * and will fail for transient domains. If neither flag is specified (that is,
+ * @flags is VIR_DOMAIN_AFFECT_CURRENT), then an inactive domain modifies
+ * persistent setup, while an active domain is hypervisor-dependent on whether
+ * just live or both live and persistent state is changed.
+ * Not all hypervisors can support all flag combinations.
+ *
+ * Returns 0 in case of success, -1 in case of failure.
+ *
+ */
+int
+virDomainPinVcpuFlags(virDomainPtr domain, unsigned int vcpu,
+ unsigned char *cpumap, int maplen, unsigned int flags)
+{
+ virConnectPtr conn;
+
+ VIR_DOMAIN_DEBUG(domain, "vcpu=%u, cpumap=%p, maplen=%d, flags=%u",
+ vcpu, cpumap, maplen, flags);
+
+ virResetLastError();
+
+ if (!VIR_IS_CONNECTED_DOMAIN(domain)) {
+ virLibDomainError(VIR_ERR_INVALID_DOMAIN, __FUNCTION__);
+ virDispatchError(NULL);
+ return -1;
+ }
+
+ if (domain->conn->flags & VIR_CONNECT_RO) {
+ virLibDomainError(VIR_ERR_OPERATION_DENIED, __FUNCTION__);
+ goto error;
+ }
+
+ if ((vcpu > 32000) || (cpumap == NULL) || (maplen < 1)) {
+ virLibDomainError(VIR_ERR_INVALID_ARG, __FUNCTION__);
+ goto error;
+ }
+
+ conn = domain->conn;
+
+ if (conn->driver->domainPinVcpuFlags) {
+ int ret;
+ ret = conn->driver->domainPinVcpuFlags (domain, vcpu, cpumap, maplen,
flags);
+ if (ret < 0)
+ goto error;
+ return ret;
+ }
+
+ virLibConnError(VIR_ERR_NO_SUPPORT, __FUNCTION__);
+
+error:
+ virDispatchError(domain->conn);
+ return -1;
+
+}
+
+/**
* virDomainGetVcpus:
* @domain: pointer to domain object, or NULL for Domain0
* @info: pointer to an array of virVcpuInfo structures (OUT)
Index: libvirt/src/libvirt_public.syms
===================================================================
--- libvirt.orig/src/libvirt_public.syms
+++ libvirt/src/libvirt_public.syms
@@ -450,4 +450,9 @@ LIBVIRT_0.9.2 {
virInterfaceChangeRollback;
} LIBVIRT_0.9.0;
+LIBVIRT_0.9.3 {
+ global:
+ virDomainPinVcpuFlags;
+} LIBVIRT_0.9.2;
+
# .... define new API here using predicted next version number ....
8:44 a.m.
New subject: [libvirt] [PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
On Fri, Jun 10, 2011 at 02:54:43PM +0900, Taku Izumi wrote:
This patch introduces a new libvirt API (virDomainPinVcpuFlags).
Signd-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
include/libvirt/libvirt.h.in | 5 ++
src/driver.h | 7 +++
src/libvirt.c | 76 +++++++++++++++++++++++++++++++++++++++++++
src/libvirt_public.syms | 5 ++
4 files changed, 93 insertions(+)
Index: libvirt/include/libvirt/libvirt.h.in
===================================================================
--- libvirt.orig/include/libvirt/libvirt.h.in
+++ libvirt/include/libvirt/libvirt.h.in
@@ -1049,6 +1049,11 @@ int virDomainPinVcpu
unsigned int vcpu,
unsigned char *cpumap,
int maplen);
+int virDomainPinVcpuFlags (virDomainPtr domain,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags);
/**
* VIR_USE_CPU:
Index: libvirt/src/driver.h
===================================================================
--- libvirt.orig/src/driver.h
+++ libvirt/src/driver.h
@@ -230,6 +230,12 @@ typedef int
unsigned char *cpumap,
int maplen);
typedef int
+ (*virDrvDomainPinVcpuFlags) (virDomainPtr domain,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags);
+typedef int
(*virDrvDomainGetVcpus) (virDomainPtr domain,
virVcpuInfoPtr info,
int maxinfo,
@@ -671,6 +677,7 @@ struct _virDriver {
virDrvDomainSetVcpusFlags domainSetVcpusFlags;
virDrvDomainGetVcpusFlags domainGetVcpusFlags;
virDrvDomainPinVcpu domainPinVcpu;
+ virDrvDomainPinVcpuFlags domainPinVcpuFlags;
virDrvDomainGetVcpus domainGetVcpus;
virDrvDomainGetMaxVcpus domainGetMaxVcpus;
virDrvDomainGetSecurityLabel domainGetSecurityLabel;
Index: libvirt/src/libvirt.c
===================================================================
--- libvirt.orig/src/libvirt.c
+++ libvirt/src/libvirt.c
@@ -6709,6 +6709,82 @@ error:
}
/**
+ * virDomainPinVcpuFlags:
+ * @domain: pointer to domain object, or NULL for Domain0
+ * @vcpu: virtual CPU number
+ * @cpumap: pointer to a bit map of real CPUs (in 8-bit bytes) (IN)
+ * Each bit set to 1 means that corresponding CPU is usable.
+ * Bytes are stored in little-endian order: CPU0-7, 8-15...
+ * In each byte, lowest CPU number is least significant bit.
+ * @maplen: number of bytes in cpumap, from 1 up to size of CPU map in
+ * underlying virtualization system (Xen...).
+ * If maplen < size, missing bytes are set to zero.
+ * If maplen > size, failure code is returned.
+ * @flags: bitwise-OR of virDomainModificationImpac
+ *
+ * Dynamically change the real CPUs which can be allocated to a virtual CPU.
+ * This function requires privileged access to the hypervisor.
+ *
+ * @flags may include VIR_DOMAIN_AFFECT_LIVE or VIR_DOMAIN_AFFECT_CONFIG.
+ * Both flags may be set.
+ * If VIR_DOMAIN_AFFECT_LIVE is set, the change affects a running domain
+ * and may fail if domain is not alive.
+ * If VIR_DOMAIN_AFFECT_CONFIG is set, the change affects persistent state,
+ * and will fail for transient domains. If neither flag is specified (that is,
+ * @flags is VIR_DOMAIN_AFFECT_CURRENT), then an inactive domain modifies
+ * persistent setup, while an active domain is hypervisor-dependent on whether
+ * just live or both live and persistent state is changed.
+ * Not all hypervisors can support all flag combinations.
+ *
+ * Returns 0 in case of success, -1 in case of failure.
+ *
+ */
+int
+virDomainPinVcpuFlags(virDomainPtr domain, unsigned int vcpu,
+ unsigned char *cpumap, int maplen, unsigned int flags)
+{
+ virConnectPtr conn;
+
+ VIR_DOMAIN_DEBUG(domain, "vcpu=%u, cpumap=%p, maplen=%d, flags=%u",
+ vcpu, cpumap, maplen, flags);
+
+ virResetLastError();
+
+ if (!VIR_IS_CONNECTED_DOMAIN(domain)) {
+ virLibDomainError(VIR_ERR_INVALID_DOMAIN, __FUNCTION__);
+ virDispatchError(NULL);
+ return -1;
+ }
+
+ if (domain->conn->flags & VIR_CONNECT_RO) {
+ virLibDomainError(VIR_ERR_OPERATION_DENIED, __FUNCTION__);
+ goto error;
+ }
+
+ if ((vcpu > 32000) || (cpumap == NULL) || (maplen < 1)) {
+ virLibDomainError(VIR_ERR_INVALID_ARG, __FUNCTION__);
+ goto error;
+ }
+
+ conn = domain->conn;
+
+ if (conn->driver->domainPinVcpuFlags) {
+ int ret;
+ ret = conn->driver->domainPinVcpuFlags (domain, vcpu, cpumap, maplen,
flags);
+ if (ret < 0)
+ goto error;
+ return ret;
+ }
+
+ virLibConnError(VIR_ERR_NO_SUPPORT, __FUNCTION__);
+
+error:
+ virDispatchError(domain->conn);
+ return -1;
+
+}
+
+/**
* virDomainGetVcpus:
* @domain: pointer to domain object, or NULL for Domain0
* @info: pointer to an array of virVcpuInfo structures (OUT)
Index: libvirt/src/libvirt_public.syms
===================================================================
--- libvirt.orig/src/libvirt_public.syms
+++ libvirt/src/libvirt_public.syms
@@ -450,4 +450,9 @@ LIBVIRT_0.9.2 {
virInterfaceChangeRollback;
} LIBVIRT_0.9.0;
+LIBVIRT_0.9.3 {
+ global:
+ virDomainPinVcpuFlags;
+} LIBVIRT_0.9.2;
+
Looks fine to me, ACK,
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
11:01 a.m.
New subject: [libvirt] [PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
On 06/13/2011 07:44 AM, Daniel Veillard wrote:
>
> /**
> + * virDomainPinVcpuFlags:
> +
> +/**
> * virDomainGetVcpus:
We need a virDomainGetVcpusFlags counterpart, so that we can query
either the config or the live state.
--
Eric Blake eblake(a)redhat.com +1-801-349-2682
Libvirt virtualization library http://libvirt.org
11:13 a.m.
New subject: [libvirt] [PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
On 06/13/2011 10:01 AM, Eric Blake wrote:
On 06/13/2011 07:44 AM, Daniel Veillard wrote:
>>
>> /**
>> + * virDomainPinVcpuFlags:
>> +
>> +/**
>> * virDomainGetVcpus:
We need a virDomainGetVcpusFlags counterpart, so that we can query
either the config or the live state.
That didn't come across quite as clear as I intended, so I'll try again
with more details:
Right now, the 'vcpupin' command in virsh uses only the
virDomainPinVcpu{,Flags} API to set the vcpu pinning, while the
'vcpuinfo' command queries the existing vcpu details, including pinning
details. But 'vcpuinfo' currently fails on a non-running information,
and it uses the virDomainGetVcpus command; which means we need a
counterpart command (virDomainGetVcpusFlags) which will let us query the
persistent vcpu settings of an inactive domain.
Furthermore, it might be nice to add 'virsh vcpupin --query', which
prints the pinning information in the same manner of string as can be
used to set pinning, rather than the current 'virsh vcpuinfo' which is
rather complex for parsing.
So, I think we still need some more work on this front before 0.9.3 is
released.
--
Eric Blake eblake(a)redhat.com +1-801-349-2682
Libvirt virtualization library http://libvirt.org
7:32 p.m.
New subject: [libvirt] [PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
That didn't come across quite as clear as I intended, so I'll
try again
with more details:
Right now, the 'vcpupin' command in virsh uses only the
virDomainPinVcpu{,Flags} API to set the vcpu pinning, while the
'vcpuinfo' command queries the existing vcpu details, including pinning
details. But 'vcpuinfo' currently fails on a non-running information,
and it uses the virDomainGetVcpus command; which means we need a
counterpart command (virDomainGetVcpusFlags) which will let us query the
persistent vcpu settings of an inactive domain.
Furthermore, it might be nice to add 'virsh vcpupin --query', which
prints the pinning information in the same manner of string as can be
used to set pinning, rather than the current 'virsh vcpuinfo' which is
rather complex for parsing.
So, I think we still need some more work on this front before 0.9.3 is
released.
OK, I'll try.
Taku Izumi <izumi.taku(a)jp.fujitsu.com>
12:55 a.m.
New subject: [libvirt] [PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu driver
This patch implements the code to address the new API (virDomainPinVcpuFlags)
in the qemu driver.
Signed-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
src/qemu/qemu_driver.c | 99 +++++++++++++++++++++++++++++++++++++------------
1 file changed, 76 insertions(+), 23 deletions(-)
Index: libvirt/src/qemu/qemu_driver.c
===================================================================
--- libvirt.orig/src/qemu/qemu_driver.c
+++ libvirt/src/qemu/qemu_driver.c
@@ -2880,17 +2880,24 @@ qemudDomainSetVcpus(virDomainPtr dom, un
static int
-qemudDomainPinVcpu(virDomainPtr dom,
- unsigned int vcpu,
- unsigned char *cpumap,
- int maplen) {
+qemudDomainPinVcpuFlags(virDomainPtr dom,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags) {
+
struct qemud_driver *driver = dom->conn->privateData;
virDomainObjPtr vm;
+ virDomainDefPtr persistentDef = NULL;
int maxcpu, hostcpus;
virNodeInfo nodeinfo;
int ret = -1;
+ bool isActive;
qemuDomainObjPrivatePtr priv;
+ virCheckFlags(VIR_DOMAIN_AFFECT_LIVE |
+ VIR_DOMAIN_AFFECT_CONFIG, -1);
+
qemuDriverLock(driver);
vm = virDomainFindByUUID(&driver->domains, dom->uuid);
qemuDriverUnlock(driver);
@@ -2903,9 +2910,18 @@ qemudDomainPinVcpu(virDomainPtr dom,
goto cleanup;
}
- if (!virDomainObjIsActive(vm)) {
- qemuReportError(VIR_ERR_OPERATION_INVALID,
- "%s",_("cannot pin vcpus on an inactive
domain"));
+ isActive = virDomainObjIsActive(vm);
+ if (flags == VIR_DOMAIN_AFFECT_CURRENT) {
+ if (isActive)
+ flags = VIR_DOMAIN_AFFECT_LIVE;
+ else
+ flags = VIR_DOMAIN_AFFECT_CONFIG;
+ }
+
+ if (!isActive && (flags & VIR_DOMAIN_AFFECT_LIVE)) {
+ qemuReportError(VIR_ERR_OPERATION_INVALID, "%s",
+ _("a domain is inactive; can change only "
+ "persistent config"));
goto cleanup;
}
@@ -2918,27 +2934,54 @@ qemudDomainPinVcpu(virDomainPtr dom,
goto cleanup;
}
- if (nodeGetInfo(dom->conn, &nodeinfo) < 0)
- goto cleanup;
+ if (flags & VIR_DOMAIN_AFFECT_CONFIG) {
+ if (!vm->persistent) {
+ qemuReportError(VIR_ERR_OPERATION_INVALID, "%s",
+ _("cannot change persistent config of a transient
domain"));
+ goto cleanup;
+ }
+ if (!(persistentDef = virDomainObjGetPersistentDef(driver->caps, vm)))
+ goto cleanup;
+ }
- hostcpus = VIR_NODEINFO_MAXCPUS(nodeinfo);
- maxcpu = maplen * 8;
- if (maxcpu > hostcpus)
- maxcpu = hostcpus;
+ if (flags & VIR_DOMAIN_AFFECT_LIVE) {
- if (priv->vcpupids != NULL) {
- if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
- cpumap, maplen, maxcpu) < 0)
+ if (nodeGetInfo(dom->conn, &nodeinfo) < 0)
goto cleanup;
- } else {
- qemuReportError(VIR_ERR_NO_SUPPORT,
- "%s", _("cpu affinity is not supported"));
- goto cleanup;
+
+ hostcpus = VIR_NODEINFO_MAXCPUS(nodeinfo);
+ maxcpu = maplen * 8;
+ if (maxcpu > hostcpus)
+ maxcpu = hostcpus;
+
+ if (priv->vcpupids != NULL) {
+ if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
+ cpumap, maplen, maxcpu) < 0)
+ goto cleanup;
+ } else {
+ qemuReportError(VIR_ERR_NO_SUPPORT,
+ "%s", _("cpu affinity is not
supported"));
+ goto cleanup;
+ }
+
+ if (virDomainVcpupinAdd(vm->def, cpumap, maplen, vcpu) < 0) {
+ qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("failed to update or add vcpupin xml of "
+ "a running domain"));
+ goto cleanup;
+ }
+
}
- if (virDomainVcpupinAdd(vm->def, cpumap, maplen, vcpu) < 0) {
- qemuReportError(VIR_ERR_INTERNAL_ERROR,
- "%s", _("failed to update or add vcpupin
xml"));
+ if (flags & VIR_DOMAIN_AFFECT_CONFIG) {
+
+ if (virDomainVcpupinAdd(persistentDef, cpumap, maplen, vcpu) < 0) {
+ qemuReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("failed to update or add vcpupin xml of "
+ "a persistent domain"));
+ goto cleanup;
+ }
+ ret = virDomainSaveConfig(driver->configDir, persistentDef);
goto cleanup;
}
@@ -2951,6 +2994,15 @@ cleanup:
}
static int
+qemudDomainPinVcpu(virDomainPtr dom,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen) {
+ return qemudDomainPinVcpuFlags(dom, vcpu, cpumap, maplen,
+ VIR_DOMAIN_AFFECT_LIVE);
+}
+
+static int
qemudDomainGetVcpus(virDomainPtr dom,
virVcpuInfoPtr info,
int maxinfo,
@@ -8014,6 +8066,7 @@ static virDriver qemuDriver = {
.domainSetVcpusFlags = qemudDomainSetVcpusFlags, /* 0.8.5 */
.domainGetVcpusFlags = qemudDomainGetVcpusFlags, /* 0.8.5 */
.domainPinVcpu = qemudDomainPinVcpu, /* 0.4.4 */
+ .domainPinVcpuFlags = qemudDomainPinVcpuFlags, /* 0.9.3 */
.domainGetVcpus = qemudDomainGetVcpus, /* 0.4.4 */
.domainGetMaxVcpus = qemudDomainGetMaxVcpus, /* 0.4.4 */
.domainGetSecurityLabel = qemudDomainGetSecurityLabel, /* 0.6.1 */
8:50 a.m.
New subject: [libvirt] [PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu driver
On Fri, Jun 10, 2011 at 02:55:44PM +0900, Taku Izumi wrote:
This patch implements the code to address the new API (virDomainPinVcpuFlags)
in the qemu driver.
Signed-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
src/qemu/qemu_driver.c | 99 +++++++++++++++++++++++++++++++++++++------------
1 file changed, 76 insertions(+), 23 deletions(-)
Index: libvirt/src/qemu/qemu_driver.c
===================================================================
--- libvirt.orig/src/qemu/qemu_driver.c
+++ libvirt/src/qemu/qemu_driver.c
@@ -2880,17 +2880,24 @@ qemudDomainSetVcpus(virDomainPtr dom, un
static int
-qemudDomainPinVcpu(virDomainPtr dom,
- unsigned int vcpu,
- unsigned char *cpumap,
- int maplen) {
+qemudDomainPinVcpuFlags(virDomainPtr dom,
+ unsigned int vcpu,
+ unsigned char *cpumap,
+ int maplen,
+ unsigned int flags) {
+
struct qemud_driver *driver = dom->conn->privateData;
virDomainObjPtr vm;
+ virDomainDefPtr persistentDef = NULL;
int maxcpu, hostcpus;
virNodeInfo nodeinfo;
int ret = -1;
+ bool isActive;
qemuDomainObjPrivatePtr priv;
+ virCheckFlags(VIR_DOMAIN_AFFECT_LIVE |
+ VIR_DOMAIN_AFFECT_CONFIG, -1);
+
qemuDriverLock(driver);
vm = virDomainFindByUUID(&driver->domains, dom->uuid);
qemuDriverUnlock(driver);
@@ -2903,9 +2910,18 @@ qemudDomainPinVcpu(virDomainPtr dom,
goto cleanup;
}
- if (!virDomainObjIsActive(vm)) {
- qemuReportError(VIR_ERR_OPERATION_INVALID,
- "%s",_("cannot pin vcpus on an inactive
domain"));
+ isActive = virDomainObjIsActive(vm);
+ if (flags == VIR_DOMAIN_AFFECT_CURRENT) {
+ if (isActive)
+ flags = VIR_DOMAIN_AFFECT_LIVE;
+ else
+ flags = VIR_DOMAIN_AFFECT_CONFIG;
+ }
+
+ if (!isActive && (flags & VIR_DOMAIN_AFFECT_LIVE)) {
+ qemuReportError(VIR_ERR_OPERATION_INVALID, "%s",
+ _("a domain is inactive; can change only "
+ "persistent config"));
goto cleanup;
}
@@ -2918,27 +2934,54 @@ qemudDomainPinVcpu(virDomainPtr dom,
goto cleanup;
}
- if (nodeGetInfo(dom->conn, &nodeinfo) < 0)
- goto cleanup;
+ if (flags & VIR_DOMAIN_AFFECT_CONFIG) {
+ if (!vm->persistent) {
+ qemuReportError(VIR_ERR_OPERATION_INVALID, "%s",
+ _("cannot change persistent config of a transient
domain"));
+ goto cleanup;
+ }
+ if (!(persistentDef = virDomainObjGetPersistentDef(driver->caps, vm)))
+ goto cleanup;
+ }
- hostcpus = VIR_NODEINFO_MAXCPUS(nodeinfo);
- maxcpu = maplen * 8;
- if (maxcpu > hostcpus)
- maxcpu = hostcpus;
+ if (flags & VIR_DOMAIN_AFFECT_LIVE) {
- if (priv->vcpupids != NULL) {
- if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
- cpumap, maplen, maxcpu) < 0)
+ if (nodeGetInfo(dom->conn, &nodeinfo) < 0)
goto cleanup;
- } else {
- qemuReportError(VIR_ERR_NO_SUPPORT,
- "%s", _("cpu affinity is not supported"));
- goto cleanup;
+
+ hostcpus = VIR_NODEINFO_MAXCPUS(nodeinfo);
+ maxcpu = maplen * 8;
+ if (maxcpu > hostcpus)
+ maxcpu = hostcpus;
I know it's the old code behaviour but one could argue that when
modifying the XML config file, maybe what the user requested should be
saved to the permanent definition, and adjusted when the guest is
(re)booted accoding to current running host. But it's the same debate
as "cpu pinning informations is tuning" and whether tuning should or
should not be saved in the persistent long term config.
+ if (priv->vcpupids != NULL) {
+ if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
+ cpumap, maplen, maxcpu) < 0)
+ goto cleanup;
+ } else {
+ qemuReportError(VIR_ERR_NO_SUPPORT,
+ "%s", _("cpu affinity is not
supported"));
+ goto cleanup;
But snce it's the existing behaviour, ACK :-)
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
12:57 a.m.
New subject: [libvirt] [PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
This patch implements the remote protocol to address the new API
(virDomainPinVcpuFlags).
Signd-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
daemon/remote.c | 42 ++++++++++++++++++++++++++++++++++++++++++
src/remote/remote_driver.c | 1 +
src/remote/remote_protocol.x | 10 +++++++++-
src/remote_protocol-structs | 9 +++++++++
4 files changed, 61 insertions(+), 1 deletion(-)
Index: libvirt/src/remote/remote_protocol.x
===================================================================
--- libvirt.orig/src/remote/remote_protocol.x
+++ libvirt/src/remote/remote_protocol.x
@@ -837,6 +837,13 @@ struct remote_domain_pin_vcpu_args {
opaque cpumap<REMOTE_CPUMAP_MAX>;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ int vcpu;
+ opaque cpumap<REMOTE_CPUMAP_MAX>;
+ unsigned int flags;
+};
+
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
@@ -2297,7 +2304,8 @@ enum remote_procedure {
REMOTE_PROC_INTERFACE_CHANGE_COMMIT = 221, /* autogen autogen */
REMOTE_PROC_INTERFACE_CHANGE_ROLLBACK = 222, /* autogen autogen */
REMOTE_PROC_DOMAIN_GET_SCHEDULER_PARAMETERS_FLAGS = 223, /* skipgen autogen */
- REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224 /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224, /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_PIN_VCPU_FLAGS = 225 /* skipgen autogen */
/*
* Notice how the entries are grouped in sets of 10 ?
Index: libvirt/src/remote_protocol-structs
===================================================================
--- libvirt.orig/src/remote_protocol-structs
+++ libvirt/src/remote_protocol-structs
@@ -529,6 +529,15 @@ struct remote_domain_pin_vcpu_args {
char * cpumap_val;
} cpumap;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ int vcpu;
+ struct {
+ u_int cpumap_len;
+ char * cpumap_val;
+ } cpumap;
+ u_int flags;
+};
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
Index: libvirt/src/remote/remote_driver.c
===================================================================
--- libvirt.orig/src/remote/remote_driver.c
+++ libvirt/src/remote/remote_driver.c
@@ -6259,6 +6259,7 @@ static virDriver remote_driver = {
.domainSetVcpusFlags = remoteDomainSetVcpusFlags, /* 0.8.5 */
.domainGetVcpusFlags = remoteDomainGetVcpusFlags, /* 0.8.5 */
.domainPinVcpu = remoteDomainPinVcpu, /* 0.3.0 */
+ .domainPinVcpuFlags = remoteDomainPinVcpuFlags, /* 0.9.3 */
.domainGetVcpus = remoteDomainGetVcpus, /* 0.3.0 */
.domainGetMaxVcpus = remoteDomainGetMaxVcpus, /* 0.3.0 */
.domainGetSecurityLabel = remoteDomainGetSecurityLabel, /* 0.6.1 */
Index: libvirt/daemon/remote.c
===================================================================
--- libvirt.orig/daemon/remote.c
+++ libvirt/daemon/remote.c
@@ -1277,6 +1277,48 @@ cleanup:
}
static int
+remoteDispatchDomainPinVcpuFlags(struct qemud_server *server ATTRIBUTE_UNUSED,
+ struct qemud_client *client ATTRIBUTE_UNUSED,
+ virConnectPtr conn,
+ remote_message_header *hdr ATTRIBUTE_UNUSED,
+ remote_error *rerr,
+ remote_domain_pin_vcpu_flags_args *args,
+ void *ret ATTRIBUTE_UNUSED)
+{
+ virDomainPtr dom = NULL;
+ int rv = -1;
+
+ if (!conn) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not
open"));
+ goto cleanup;
+ }
+
+ if (!(dom = get_nonnull_domain(conn, args->dom)))
+ goto cleanup;
+
+ if (args->cpumap.cpumap_len > REMOTE_CPUMAP_MAX) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("cpumap_len >
REMOTE_CPUMAP_MAX"));
+ goto cleanup;
+ }
+
+ if (virDomainPinVcpuFlags(dom,
+ args->vcpu,
+ (unsigned char *) args->cpumap.cpumap_val,
+ args->cpumap.cpumap_len,
+ args->flags) < 0)
+ goto cleanup;
+
+ rv = 0;
+
+cleanup:
+ if (rv < 0)
+ remoteDispatchError(rerr);
+ if (dom)
+ virDomainFree(dom);
+ return rv;
+}
+
+static int
remoteDispatchDomainGetMemoryParameters(struct qemud_server *server
ATTRIBUTE_UNUSED,
struct qemud_client *client
3:25 a.m.
New subject: [libvirt] [PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
Index: libvirt/src/remote/remote_driver.c
===================================================================
--- libvirt.orig/src/remote/remote_driver.c
+++ libvirt/src/remote/remote_driver.c
@@ -6259,6 +6259,7 @@ static virDriver remote_driver = {
.domainSetVcpusFlags = remoteDomainSetVcpusFlags, /* 0.8.5 */
.domainGetVcpusFlags = remoteDomainGetVcpusFlags, /* 0.8.5 */
.domainPinVcpu = remoteDomainPinVcpu, /* 0.3.0 */
+ .domainPinVcpuFlags = remoteDomainPinVcpuFlags, /* 0.9.3 */
gcc gives a warning here. the generated remoteDomainPinVcpuFlags is
defined as:
static int
remoteDomainPinVcpuFlags(virDomainPtr dom, int vcpu, const char *cpumap,
int cpumaplen, unsigned int flags)
but .domainPinVcpuFlags is of type:
typedef int
(*virDrvDomainPinVcpuFlags) (virDomainPtr domain,
unsigned int vcpu,
unsigned char *cpumap,
int maplen,
unsigned int flags);
--
Thanks,
Hu Tao
9:16 p.m.
New subject: [libvirt] [PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
I updated according to Hu Tao's comment. Thanks.
---
This patch implements the remote protocol to address the new API
(virDomainPinVcpuFlags).
Signd-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
daemon/remote.c | 42 ++++++++++++++++++++++++++++++++++++++++++
daemon/remote_generator.pl | 3 +++
src/remote/remote_driver.c | 1 +
src/remote/remote_protocol.x | 10 +++++++++-
src/remote_protocol-structs | 9 +++++++++
5 files changed, 64 insertions(+), 1 deletion(-)
Index: libvirt/src/remote/remote_protocol.x
===================================================================
--- libvirt.orig/src/remote/remote_protocol.x
+++ libvirt/src/remote/remote_protocol.x
@@ -837,6 +837,13 @@ struct remote_domain_pin_vcpu_args {
opaque cpumap<REMOTE_CPUMAP_MAX>;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ unsigned int vcpu;
+ opaque cpumap<REMOTE_CPUMAP_MAX>;
+ unsigned int flags;
+};
+
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
@@ -2297,7 +2304,8 @@ enum remote_procedure {
REMOTE_PROC_INTERFACE_CHANGE_COMMIT = 221, /* autogen autogen */
REMOTE_PROC_INTERFACE_CHANGE_ROLLBACK = 222, /* autogen autogen */
REMOTE_PROC_DOMAIN_GET_SCHEDULER_PARAMETERS_FLAGS = 223, /* skipgen autogen */
- REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224 /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224, /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_PIN_VCPU_FLAGS = 225 /* skipgen autogen */
/*
* Notice how the entries are grouped in sets of 10 ?
Index: libvirt/src/remote_protocol-structs
===================================================================
--- libvirt.orig/src/remote_protocol-structs
+++ libvirt/src/remote_protocol-structs
@@ -529,6 +529,15 @@ struct remote_domain_pin_vcpu_args {
char * cpumap_val;
} cpumap;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ u_int vcpu;
+ struct {
+ u_int cpumap_len;
+ char * cpumap_val;
+ } cpumap;
+ u_int flags;
+};
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
Index: libvirt/src/remote/remote_driver.c
===================================================================
--- libvirt.orig/src/remote/remote_driver.c
+++ libvirt/src/remote/remote_driver.c
@@ -6259,6 +6259,7 @@ static virDriver remote_driver = {
.domainSetVcpusFlags = remoteDomainSetVcpusFlags, /* 0.8.5 */
.domainGetVcpusFlags = remoteDomainGetVcpusFlags, /* 0.8.5 */
.domainPinVcpu = remoteDomainPinVcpu, /* 0.3.0 */
+ .domainPinVcpuFlags = remoteDomainPinVcpuFlags, /* 0.9.3 */
.domainGetVcpus = remoteDomainGetVcpus, /* 0.3.0 */
.domainGetMaxVcpus = remoteDomainGetMaxVcpus, /* 0.3.0 */
.domainGetSecurityLabel = remoteDomainGetSecurityLabel, /* 0.6.1 */
Index: libvirt/daemon/remote.c
===================================================================
--- libvirt.orig/daemon/remote.c
+++ libvirt/daemon/remote.c
@@ -1277,6 +1277,48 @@ cleanup:
}
static int
+remoteDispatchDomainPinVcpuFlags(struct qemud_server *server ATTRIBUTE_UNUSED,
+ struct qemud_client *client ATTRIBUTE_UNUSED,
+ virConnectPtr conn,
+ remote_message_header *hdr ATTRIBUTE_UNUSED,
+ remote_error *rerr,
+ remote_domain_pin_vcpu_flags_args *args,
+ void *ret ATTRIBUTE_UNUSED)
+{
+ virDomainPtr dom = NULL;
+ int rv = -1;
+
+ if (!conn) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not
open"));
+ goto cleanup;
+ }
+
+ if (!(dom = get_nonnull_domain(conn, args->dom)))
+ goto cleanup;
+
+ if (args->cpumap.cpumap_len > REMOTE_CPUMAP_MAX) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("cpumap_len >
REMOTE_CPUMAP_MAX"));
+ goto cleanup;
+ }
+
+ if (virDomainPinVcpuFlags(dom,
+ args->vcpu,
+ (unsigned char *) args->cpumap.cpumap_val,
+ args->cpumap.cpumap_len,
+ args->flags) < 0)
+ goto cleanup;
+
+ rv = 0;
+
+cleanup:
+ if (rv < 0)
+ remoteDispatchError(rerr);
+ if (dom)
+ virDomainFree(dom);
+ return rv;
+}
+
+static int
remoteDispatchDomainGetMemoryParameters(struct qemud_server *server
ATTRIBUTE_UNUSED,
struct qemud_client *client
Index: libvirt/daemon/remote_generator.pl
===================================================================
--- libvirt.orig/daemon/remote_generator.pl
+++ libvirt/daemon/remote_generator.pl
@@ -977,6 +977,9 @@ elsif ($opt_k) {
} elsif ($call->{ProcName} eq "DomainPinVcpu") {
push(@args_list, "unsigned char *$arg_name");
push(@args_list, "int ${arg_name}len");
+ } elsif ($call->{ProcName} eq "DomainPinVcpuFlags") {
+ push(@args_list, "unsigned char *$arg_name");
+ push(@args_list, "int ${arg_name}len");
} else {
push(@args_list, "const char *$arg_name");
push(@args_list, "int ${arg_name}len");
8:51 a.m.
New subject: [libvirt] [PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
On Mon, Jun 13, 2011 at 11:16:41AM +0900, Taku Izumi wrote:
I updated according to Hu Tao's comment. Thanks.
---
This patch implements the remote protocol to address the new API
(virDomainPinVcpuFlags).
Signd-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
daemon/remote.c | 42 ++++++++++++++++++++++++++++++++++++++++++
daemon/remote_generator.pl | 3 +++
src/remote/remote_driver.c | 1 +
src/remote/remote_protocol.x | 10 +++++++++-
src/remote_protocol-structs | 9 +++++++++
5 files changed, 64 insertions(+), 1 deletion(-)
Index: libvirt/src/remote/remote_protocol.x
===================================================================
--- libvirt.orig/src/remote/remote_protocol.x
+++ libvirt/src/remote/remote_protocol.x
@@ -837,6 +837,13 @@ struct remote_domain_pin_vcpu_args {
opaque cpumap<REMOTE_CPUMAP_MAX>;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ unsigned int vcpu;
+ opaque cpumap<REMOTE_CPUMAP_MAX>;
+ unsigned int flags;
+};
+
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
@@ -2297,7 +2304,8 @@ enum remote_procedure {
REMOTE_PROC_INTERFACE_CHANGE_COMMIT = 221, /* autogen autogen */
REMOTE_PROC_INTERFACE_CHANGE_ROLLBACK = 222, /* autogen autogen */
REMOTE_PROC_DOMAIN_GET_SCHEDULER_PARAMETERS_FLAGS = 223, /* skipgen autogen */
- REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224 /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_EVENT_CONTROL_ERROR = 224, /* skipgen skipgen */
+ REMOTE_PROC_DOMAIN_PIN_VCPU_FLAGS = 225 /* skipgen autogen */
/*
* Notice how the entries are grouped in sets of 10 ?
Index: libvirt/src/remote_protocol-structs
===================================================================
--- libvirt.orig/src/remote_protocol-structs
+++ libvirt/src/remote_protocol-structs
@@ -529,6 +529,15 @@ struct remote_domain_pin_vcpu_args {
char * cpumap_val;
} cpumap;
};
+struct remote_domain_pin_vcpu_flags_args {
+ remote_nonnull_domain dom;
+ u_int vcpu;
+ struct {
+ u_int cpumap_len;
+ char * cpumap_val;
+ } cpumap;
+ u_int flags;
+};
struct remote_domain_get_vcpus_args {
remote_nonnull_domain dom;
int maxinfo;
Index: libvirt/src/remote/remote_driver.c
===================================================================
--- libvirt.orig/src/remote/remote_driver.c
+++ libvirt/src/remote/remote_driver.c
@@ -6259,6 +6259,7 @@ static virDriver remote_driver = {
.domainSetVcpusFlags = remoteDomainSetVcpusFlags, /* 0.8.5 */
.domainGetVcpusFlags = remoteDomainGetVcpusFlags, /* 0.8.5 */
.domainPinVcpu = remoteDomainPinVcpu, /* 0.3.0 */
+ .domainPinVcpuFlags = remoteDomainPinVcpuFlags, /* 0.9.3 */
.domainGetVcpus = remoteDomainGetVcpus, /* 0.3.0 */
.domainGetMaxVcpus = remoteDomainGetMaxVcpus, /* 0.3.0 */
.domainGetSecurityLabel = remoteDomainGetSecurityLabel, /* 0.6.1 */
Index: libvirt/daemon/remote.c
===================================================================
--- libvirt.orig/daemon/remote.c
+++ libvirt/daemon/remote.c
@@ -1277,6 +1277,48 @@ cleanup:
}
static int
+remoteDispatchDomainPinVcpuFlags(struct qemud_server *server ATTRIBUTE_UNUSED,
+ struct qemud_client *client ATTRIBUTE_UNUSED,
+ virConnectPtr conn,
+ remote_message_header *hdr ATTRIBUTE_UNUSED,
+ remote_error *rerr,
+ remote_domain_pin_vcpu_flags_args *args,
+ void *ret ATTRIBUTE_UNUSED)
+{
+ virDomainPtr dom = NULL;
+ int rv = -1;
+
+ if (!conn) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("connection not
open"));
+ goto cleanup;
+ }
+
+ if (!(dom = get_nonnull_domain(conn, args->dom)))
+ goto cleanup;
+
+ if (args->cpumap.cpumap_len > REMOTE_CPUMAP_MAX) {
+ virNetError(VIR_ERR_INTERNAL_ERROR, "%s", _("cpumap_len >
REMOTE_CPUMAP_MAX"));
+ goto cleanup;
+ }
+
+ if (virDomainPinVcpuFlags(dom,
+ args->vcpu,
+ (unsigned char *) args->cpumap.cpumap_val,
+ args->cpumap.cpumap_len,
+ args->flags) < 0)
+ goto cleanup;
+
+ rv = 0;
+
+cleanup:
+ if (rv < 0)
+ remoteDispatchError(rerr);
+ if (dom)
+ virDomainFree(dom);
+ return rv;
+}
+
+static int
remoteDispatchDomainGetMemoryParameters(struct qemud_server *server
ATTRIBUTE_UNUSED,
struct qemud_client *client
Index: libvirt/daemon/remote_generator.pl
===================================================================
--- libvirt.orig/daemon/remote_generator.pl
+++ libvirt/daemon/remote_generator.pl
@@ -977,6 +977,9 @@ elsif ($opt_k) {
} elsif ($call->{ProcName} eq "DomainPinVcpu") {
push(@args_list, "unsigned char *$arg_name");
push(@args_list, "int ${arg_name}len");
+ } elsif ($call->{ProcName} eq "DomainPinVcpuFlags") {
+ push(@args_list, "unsigned char *$arg_name");
+ push(@args_list, "int ${arg_name}len");
} else {
push(@args_list, "const char *$arg_name");
push(@args_list, "int ${arg_name}len");
ACK,
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
12:57 a.m.
New subject: [libvirt] [PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin" command
This patch adds the new option (--live, --config and --current) to
"virsh vcpupin" command. The behavior of above aption is the same as
that of "virsh setmem", "virsh setvcpus", and whatnot.
When the --config option is specified, the command affects a persistent domain,
while --live option is specified, it affects a running (live) domain.
The --current option cannot be used with --config or --live at the same time,
and when --current is specified, it affects a "current" domain.
Signed-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
tools/virsh.c | 33 +++++++++++++++++++++++++++++++--
tools/virsh.pod | 8 +++++++-
2 files changed, 38 insertions(+), 3 deletions(-)
Index: libvirt/tools/virsh.c
===================================================================
--- libvirt.orig/tools/virsh.c
+++ libvirt/tools/virsh.c
@@ -2929,6 +2929,9 @@ static const vshCmdOptDef opts_vcpupin[]
{"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or
uuid")},
{"vcpu", VSH_OT_INT, VSH_OFLAG_REQ, N_("vcpu number")},
{"cpulist", VSH_OT_DATA, VSH_OFLAG_REQ, N_("host cpu number(s) (comma
separated)")},
+ {"config", VSH_OT_BOOL, 0, N_("affect next boot")},
+ {"live", VSH_OT_BOOL, 0, N_("affect running domain")},
+ {"current", VSH_OT_BOOL, 0, N_("affect current domain")},
{NULL, 0, 0, NULL}
};
@@ -2945,6 +2948,26 @@ cmdVcpupin(vshControl *ctl, const vshCmd
int cpumaplen;
int i;
enum { expect_num, expect_num_or_comma } state;
+ int config = vshCommandOptBool(cmd, "config");
+ int live = vshCommandOptBool(cmd, "live");
+ int current = vshCommandOptBool(cmd, "current");
+ int flags = 0;
+
+ if (current) {
+ if (live || config) {
+ vshError(ctl, "%s", _("--current must be specified
exclusively"));
+ return false;
+ }
+ flags = VIR_DOMAIN_AFFECT_CURRENT;
+ } else {
+ if (config)
+ flags |= VIR_DOMAIN_AFFECT_CONFIG;
+ if (live)
+ flags |= VIR_DOMAIN_AFFECT_LIVE;
+ /* neither option is specified */
+ if (!live && !config)
+ flags = -1;
+ }
if (!vshConnectionUsability(ctl, ctl->conn))
return false;
@@ -3041,8 +3064,14 @@ cmdVcpupin(vshControl *ctl, const vshCmd
cpulist++;
} while (cpulist);
- if (virDomainPinVcpu(dom, vcpu, cpumap, cpumaplen) != 0) {
- ret = false;
+ if (flags == -1) {
+ if (virDomainPinVcpu(dom, vcpu, cpumap, cpumaplen) != 0) {
+ ret = false;
+ }
+ } else {
+ if (virDomainPinVcpuFlags(dom, vcpu, cpumap, cpumaplen, flags) != 0) {
+ ret = false;
+ }
}
VIR_FREE(cpumap);
Index: libvirt/tools/virsh.pod
===================================================================
--- libvirt.orig/tools/virsh.pod
+++ libvirt/tools/virsh.pod
@@ -790,10 +790,16 @@ values; these two flags cannot both be s
Returns basic information about the domain virtual CPUs, like the number of
vCPUs, the running time, the affinity to physical processors.
-=item B<vcpupin> I<domain-id> I<vcpu> I<cpulist>
+=item B<vcpupin> I<domain-id> I<vcpu> I<cpulist> optional
I<--live> I<--config>
+I<--current>
Pin domain VCPUs to host physical CPUs. The I<vcpu> number must be provided
and I<cpulist> is a comma separated list of physical CPU numbers.
+If I<--live> is specified, affect a running guest.
+If I<--config> is specified, affect the next boot of a persistent guest.
+If I<--current> is specified, affect the current guest state.
+Both I<--live> and I<--config> flags may be given, but I<--current> is
exclusive.
+If no flag is specified, behavior is different depending on hypervisor.
=item B<vncdisplay> I<domain-id>
8:54 a.m.
New subject: [libvirt] [PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin" command
On Fri, Jun 10, 2011 at 02:57:50PM +0900, Taku Izumi wrote:
This patch adds the new option (--live, --config and --current) to
"virsh vcpupin" command. The behavior of above aption is the same as
that of "virsh setmem", "virsh setvcpus", and whatnot.
When the --config option is specified, the command affects a persistent domain,
while --live option is specified, it affects a running (live) domain.
The --current option cannot be used with --config or --live at the same time,
and when --current is specified, it affects a "current" domain.
Signed-off-by: Taku Izumi <izumi.taku(a)jp.fujitsu.com>
---
tools/virsh.c | 33 +++++++++++++++++++++++++++++++--
tools/virsh.pod | 8 +++++++-
2 files changed, 38 insertions(+), 3 deletions(-)
Index: libvirt/tools/virsh.c
===================================================================
--- libvirt.orig/tools/virsh.c
+++ libvirt/tools/virsh.c
@@ -2929,6 +2929,9 @@ static const vshCmdOptDef opts_vcpupin[]
{"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or
uuid")},
{"vcpu", VSH_OT_INT, VSH_OFLAG_REQ, N_("vcpu number")},
{"cpulist", VSH_OT_DATA, VSH_OFLAG_REQ, N_("host cpu number(s) (comma
separated)")},
+ {"config", VSH_OT_BOOL, 0, N_("affect next boot")},
+ {"live", VSH_OT_BOOL, 0, N_("affect running domain")},
+ {"current", VSH_OT_BOOL, 0, N_("affect current domain")},
{NULL, 0, 0, NULL}
};
@@ -2945,6 +2948,26 @@ cmdVcpupin(vshControl *ctl, const vshCmd
int cpumaplen;
int i;
enum { expect_num, expect_num_or_comma } state;
+ int config = vshCommandOptBool(cmd, "config");
+ int live = vshCommandOptBool(cmd, "live");
+ int current = vshCommandOptBool(cmd, "current");
+ int flags = 0;
+
+ if (current) {
+ if (live || config) {
+ vshError(ctl, "%s", _("--current must be specified
exclusively"));
+ return false;
+ }
+ flags = VIR_DOMAIN_AFFECT_CURRENT;
+ } else {
+ if (config)
+ flags |= VIR_DOMAIN_AFFECT_CONFIG;
+ if (live)
+ flags |= VIR_DOMAIN_AFFECT_LIVE;
+ /* neither option is specified */
+ if (!live && !config)
+ flags = -1;
+ }
if (!vshConnectionUsability(ctl, ctl->conn))
return false;
@@ -3041,8 +3064,14 @@ cmdVcpupin(vshControl *ctl, const vshCmd
cpulist++;
} while (cpulist);
- if (virDomainPinVcpu(dom, vcpu, cpumap, cpumaplen) != 0) {
- ret = false;
+ if (flags == -1) {
+ if (virDomainPinVcpu(dom, vcpu, cpumap, cpumaplen) != 0) {
+ ret = false;
+ }
+ } else {
+ if (virDomainPinVcpuFlags(dom, vcpu, cpumap, cpumaplen, flags) != 0) {
+ ret = false;
+ }
}
VIR_FREE(cpumap);
Index: libvirt/tools/virsh.pod
===================================================================
--- libvirt.orig/tools/virsh.pod
+++ libvirt/tools/virsh.pod
@@ -790,10 +790,16 @@ values; these two flags cannot both be s
Returns basic information about the domain virtual CPUs, like the number of
vCPUs, the running time, the affinity to physical processors.
-=item B<vcpupin> I<domain-id> I<vcpu> I<cpulist>
+=item B<vcpupin> I<domain-id> I<vcpu> I<cpulist> optional
I<--live> I<--config>
+I<--current>
Pin domain VCPUs to host physical CPUs. The I<vcpu> number must be provided
and I<cpulist> is a comma separated list of physical CPU numbers.
+If I<--live> is specified, affect a running guest.
+If I<--config> is specified, affect the next boot of a persistent guest.
+If I<--current> is specified, affect the current guest state.
+Both I<--live> and I<--config> flags may be given, but I<--current> is
exclusive.
+If no flag is specified, behavior is different depending on hypervisor.
=item B<vncdisplay> I<domain-id>
ACK,
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
10:53 a.m.
On Fri, Jun 10, 2011 at 02:47:08PM +0900, Taku Izumi wrote:
Hi all,
I'll resend this patchset because what sent yesterday was broken.
Sorry to be a nuisance.
--
This patchset enables us to configure inactive domains' CPU affinity setting.
This is the rebased version, but retains the version number.
*[PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
*[PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu driver
*[PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
*[PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin" command
Okay, I have commited the patch set, including modified 3/4
thanks a lot !
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
7:25 p.m.
On Mon, 13 Jun 2011 23:53:59 +0800
Daniel Veillard <veillard(a)redhat.com> wrote:
On Fri, Jun 10, 2011 at 02:47:08PM +0900, Taku Izumi wrote:
> Hi all,
>
> I'll resend this patchset because what sent yesterday was broken.
> Sorry to be a nuisance.
>
> --
>
> This patchset enables us to configure inactive domains' CPU affinity setting.
> This is the rebased version, but retains the version number.
>
> *[PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
> *[PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu
driver
> *[PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
> *[PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin" command
Okay, I have commited the patch set, including modified 3/4
thanks a lot !
Thank you!
Hopefully, could you commit the following ?
http://www.redhat.com/archives/libvir-list/2011-June/msg00289.html
I think it's a simple but serious bug.
--
Taku Izumi <izumi.taku(a)jp.fujitsu.com>
10:19 p.m.
On 06/14/2011 08:25 AM, Taku Izumi wrote:
On Mon, 13 Jun 2011 23:53:59 +0800
Daniel Veillard<veillard(a)redhat.com> wrote:
> On Fri, Jun 10, 2011 at 02:47:08PM +0900, Taku Izumi wrote:
>> Hi all,
>>
>> I'll resend this patchset because what sent yesterday was broken.
>> Sorry to be a nuisance.
>>
>> --
>>
>> This patchset enables us to configure inactive domains' CPU affinity
setting.
>> This is the rebased version, but retains the version number.
>>
>> *[PATCH v4 1/4] vcpupin: introduce a new libvirt API (virDomainPinVcpuFlags)
>> *[PATCH v4 2/4] vcpupin: implement the code to address the new API in the qemu
driver
>> *[PATCH v4 3/4] vcpupin: implement the remote protocol to address the new API
>> *[PATCH v4 4/4] vcpupin: add the new option to "virsh vcpupin"
command
>
> Okay, I have commited the patch set, including modified 3/4
>
> thanks a lot !
Thank you!
Hopefully, could you commit the following ?
http://www.redhat.com/archives/libvir-list/2011-June/msg00289.html
I think it's a simple but serious bug.
I applied the patch, thanks
Regards
Osier
2:16 a.m.
On Tue, Jun 14, 2011 at 09:25:45AM +0900, Taku Izumi wrote:
On Mon, 13 Jun 2011 23:53:59 +0800
Daniel Veillard <veillard(a)redhat.com> wrote:
[...]
> Okay, I have commited the patch set, including modified 3/4
>
> thanks a lot !
Thank you!
Hopefully, could you commit the following ?
http://www.redhat.com/archives/libvir-list/2011-June/msg00289.html
I think it's a simple but serious bug.
Osier just did that :-)
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library http://libvirt.org/
4911
days inactive
4915
days old
17 comments
5 participants
participants (5)
-
Daniel Veillard -
Eric Blake -
Hu Tao -
Osier Yang -
Taku Izumi