vcpu threads pin are implemented using sched_setaffinity(), but
not controlled by cgroup. This patch does the following things:
1) enable cpuset cgroup
2) reflect all the vcpu threads pin info to cgroup
Signed-off-by: Tang Chen <tangchen@xxxxxxxxxxxxxx>
Signed-off-by: Hu Tao <hutao@xxxxxxxxxxxxxx>
---
src/libvirt_private.syms | 2 ++
src/qemu/qemu_cgroup.c | 43 ++++++++++++++++++++++++
src/qemu/qemu_cgroup.h | 4 +++
src/qemu/qemu_driver.c | 82 +++++++++++++++++++++++++++++++++-------------
src/util/cgroup.c | 35 +++++++++++++++++++-
src/util/cgroup.h | 3 ++
6 files changed, 146 insertions(+), 23 deletions(-)
diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms
index c3b05d3..29eaef5 100644
--- a/src/libvirt_private.syms
+++ b/src/libvirt_private.syms
@@ -82,6 +82,7 @@ virCgroupGetCpuShares;
virCgroupGetCpuacctPercpuUsage;
virCgroupGetCpuacctStat;
virCgroupGetCpuacctUsage;
+virCgroupGetCpusetCpus;
virCgroupGetCpusetMems;
virCgroupGetFreezerState;
virCgroupGetMemSwapHardLimit;
@@ -100,6 +101,7 @@ virCgroupSetBlkioWeight;
virCgroupSetCpuCfsPeriod;
virCgroupSetCpuCfsQuota;
virCgroupSetCpuShares;
+virCgroupSetCpusetCpus;
virCgroupSetCpusetMems;
virCgroupSetFreezerState;
virCgroupSetMemSwapHardLimit;
diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c
index 8a5a536..37874d3 100644
--- a/src/qemu/qemu_cgroup.c
+++ b/src/qemu/qemu_cgroup.c
@@ -491,11 +491,45 @@ cleanup:
return -1;
}
+int qemuSetupCgroupVcpuPin(virCgroupPtr cgroup,
+ virDomainVcpuPinDefPtr *vcpupin,
+ int nvcpupin,
+ int vcpuid)
+{
+ int i, rc = 0;
+ char *new_cpus = NULL;
+
+ for (i = 0; i < nvcpupin; i++) {
+ if (vcpuid == vcpupin[i]->vcpuid) {
+ new_cpus = virDomainCpuSetFormat(vcpupin[i]->cpumask,
+ VIR_DOMAIN_CPUMASK_LEN);
+ if (!new_cpus) {
+ virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("failed to convert cpu mask"));
+ rc = -1;
+ goto cleanup;
+ }
+ rc = virCgroupSetCpusetCpus(cgroup, new_cpus);
+ if (rc != 0) {
+ virReportSystemError(-rc,
+ "%s",
+ _("Unable to set cpuset.cpus"));
+ goto cleanup;
+ }
+ }
+ }
+
+cleanup:
+ VIR_FREE(new_cpus);
+ return rc;
+}
+
int qemuSetupCgroupForVcpu(struct qemud_driver *driver, virDomainObjPtr vm)
{
virCgroupPtr cgroup = NULL;
virCgroupPtr cgroup_vcpu = NULL;
qemuDomainObjPrivatePtr priv = vm->privateData;
+ virDomainDefPtr def = vm->def;
int rc;
unsigned int i;
unsigned long long period = vm->def->cputune.period;
@@ -567,6 +601,15 @@ int qemuSetupCgroupForVcpu(struct qemud_driver *driver, virDomainObjPtr vm)
}
}
+ /* Set vcpupin in cgroup if vcpupin xml is provided */
+ if (def->cputune.nvcpupin &&
+ qemuCgroupControllerActive(driver, VIR_CGROUP_CONTROLLER_CPUSET) &&
+ qemuSetupCgroupVcpuPin(cgroup_vcpu,
+ def->cputune.vcpupin,
+ def->cputune.nvcpupin,
+ i) < 0)
+ goto cleanup;
+
virCgroupFree(&cgroup_vcpu);
}
diff --git a/src/qemu/qemu_cgroup.h b/src/qemu/qemu_cgroup.h
index 34a9312..fa93cdb 100644
--- a/src/qemu/qemu_cgroup.h
+++ b/src/qemu/qemu_cgroup.h
@@ -53,6 +53,10 @@ int qemuSetupCgroup(struct qemud_driver *driver,
int qemuSetupCgroupVcpuBW(virCgroupPtr cgroup,
unsigned long long period,
long long quota);
+int qemuSetupCgroupVcpuPin(virCgroupPtr cgroup,
+ virDomainVcpuPinDefPtr *vcpupin,
+ int nvcpupin,
+ int vcpuid);
int qemuSetupCgroupForVcpu(struct qemud_driver *driver, virDomainObjPtr vm);
int qemuSetupCgroupForEmulator(struct qemud_driver *driver,
virDomainObjPtr vm);
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c
index 8fe16b2..c52be95 100644
--- a/src/qemu/qemu_driver.c
+++ b/src/qemu/qemu_driver.c
@@ -3643,11 +3643,15 @@ qemudDomainPinVcpuFlags(virDomainPtr dom,
struct qemud_driver *driver = dom->conn->privateData;
virDomainObjPtr vm;
virDomainDefPtr persistentDef = NULL;
+ virCgroupPtr cgroup_dom = NULL;
+ virCgroupPtr cgroup_vcpu = NULL;
int maxcpu, hostcpus;
virNodeInfo nodeinfo;
int ret = -1;
qemuDomainObjPrivatePtr priv;
bool canResetting = true;
+ int newVcpuPinNum = 0;
+ virDomainVcpuPinDefPtr *newVcpuPin = NULL;
int pcpu;
virCheckFlags(VIR_DOMAIN_AFFECT_LIVE |
@@ -3696,43 +3700,73 @@ qemudDomainPinVcpuFlags(virDomainPtr dom,
if (flags & VIR_DOMAIN_AFFECT_LIVE) {
- if (priv->vcpupids != NULL) {
- if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
- cpumap, maplen, maxcpu) < 0)
- goto cleanup;
- } else {
+ if (priv->vcpupids == NULL) {
virReportError(VIR_ERR_OPERATION_INVALID,
"%s", _("cpu affinity is not supported"));
goto cleanup;
}
- if (canResetting) {
- if (virDomainVcpuPinDel(vm->def, vcpu) < 0) {
- virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
- _("failed to delete vcpupin xml of "
- "a running domain"));
+ if (vm->def->cputune.vcpupin) {
+ newVcpuPin = virDomainVcpuPinDefCopy(vm->def->cputune.vcpupin,
+ vm->def->cputune.nvcpupin);
+ if (!newVcpuPin)
+ goto cleanup;
+
+ newVcpuPinNum = vm->def->cputune.nvcpupin;
+ } else {
+ if (VIR_ALLOC(newVcpuPin) < 0) {
+ virReportOOMError();
+ goto cleanup;
+ }
+ newVcpuPinNum = 0;
+ }
+
+ if (virDomainVcpuPinAdd(newVcpuPin, &newVcpuPinNum, cpumap, maplen, vcpu) < 0) {
+ virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
+ _("failed to update vcpupin"));
+ virDomainVcpuPinDefFree(newVcpuPin, newVcpuPinNum);
+ goto cleanup;
+ }
+
+ /* Configure the corresponding cpuset cgroup before set affinity. */
+ if (qemuCgroupControllerActive(driver, VIR_CGROUP_CONTROLLER_CPUSET)) {
+ if (virCgroupForDomain(driver->cgroup, vm->def->name, &cgroup_dom, 0) == 0 &&
+ virCgroupForVcpu(cgroup_dom, vcpu, &cgroup_vcpu, 0) == 0 &&
+ qemuSetupCgroupVcpuPin(cgroup_vcpu, newVcpuPin, newVcpuPinNum, vcpu) < 0) {
+ virReportError(VIR_ERR_OPERATION_INVALID,
+ _("failed to set cpuset.cpus in cgroup"
+ " for vcpu %d"), vcpu);
goto cleanup;
}
} else {
- if (!vm->def->cputune.vcpupin) {
- if (VIR_ALLOC(vm->def->cputune.vcpupin) < 0) {
- virReportOOMError();
- goto cleanup;
- }
- vm->def->cputune.nvcpupin = 0;
+ if (virProcessInfoSetAffinity(priv->vcpupids[vcpu],
+ cpumap, maplen, maxcpu) < 0) {
+ virReportError(VIR_ERR_SYSTEM_ERROR,
+ _("failed to set cpu affinity for vcpu %d"),
+ vcpu);
+ goto cleanup;
}
- if (virDomainVcpuPinAdd(vm->def->cputune.vcpupin,
- &vm->def->cputune.nvcpupin,
- cpumap,
- maplen,
- vcpu) < 0) {
+ }
+
+ if (canResetting) {
+ if (virDomainVcpuPinDel(vm->def, vcpu) < 0) {
virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
- _("failed to update or add vcpupin xml of "
+ _("failed to delete vcpupin xml of "
"a running domain"));
goto cleanup;
}
+ } else {
+ if (vm->def->cputune.vcpupin)
+ virDomainVcpuPinDefFree(vm->def->cputune.vcpupin, vm->def->cputune.nvcpupin);
+
+ vm->def->cputune.vcpupin = newVcpuPin;
+ vm->def->cputune.nvcpupin = newVcpuPinNum;
+ newVcpuPin = NULL;
}
+ if (newVcpuPin)
+ virDomainVcpuPinDefFree(newVcpuPin, newVcpuPinNum);
+
if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0)
goto cleanup;
}
@@ -3773,6 +3807,10 @@ qemudDomainPinVcpuFlags(virDomainPtr dom,
ret = 0;
cleanup:
+ if (cgroup_vcpu)
+ virCgroupFree(&cgroup_vcpu);
+ if (cgroup_dom)
+ virCgroupFree(&cgroup_dom);
if (vm)
virDomainObjUnlock(vm);
return ret;
diff --git a/src/util/cgroup.c b/src/util/cgroup.c
index 374516a..840a952 100644
--- a/src/util/cgroup.c
+++ b/src/util/cgroup.c
@@ -543,7 +543,8 @@ static int virCgroupMakeGroup(virCgroupPtr parent, virCgroupPtr group,
/* We need to control cpu bandwidth for each vcpu now */
if ((flags & VIR_CGROUP_VCPU) &&
(i != VIR_CGROUP_CONTROLLER_CPU &&
- i != VIR_CGROUP_CONTROLLER_CPUACCT)) {
+ i != VIR_CGROUP_CONTROLLER_CPUACCT &&
+ i != VIR_CGROUP_CONTROLLER_CPUSET)) {
/* treat it as unmounted and we can use virCgroupAddTask */
VIR_FREE(group->controllers[i].mountPoint);
continue;
@@ -1402,6 +1403,38 @@ int virCgroupGetCpusetMems(virCgroupPtr group, char **mems)
}
/**
+ * virCgroupSetCpusetCpus:
+ *
+ * @group: The cgroup to set cpuset.cpus for
+ * @cpus: the cpus to set
+ *
+ * Retuens: 0 on success
+ */
+int virCgroupSetCpusetCpus(virCgroupPtr group, const char *cpus)
+{
+ return virCgroupSetValueStr(group,
+ VIR_CGROUP_CONTROLLER_CPUSET,
+ "cpuset.cpus",
+ cpus);
+}
+
+/**
+ * virCgroupGetCpusetCpus:
+ *
+ * @group: The cgroup to get cpuset.cpus for
+ * @cpus: the cpus to get
+ *
+ * Retuens: 0 on success
+ */
+int virCgroupGetCpusetCpus(virCgroupPtr group, char **cpus)
+{
+ return virCgroupGetValueStr(group,
+ VIR_CGROUP_CONTROLLER_CPUSET,
+ "cpuset.cpus",
+ cpus);
+}
+
+/**
* virCgroupDenyAllDevices:
*
* @group: The cgroup to deny all permissions, for all devices
diff --git a/src/util/cgroup.h b/src/util/cgroup.h
index 727e536..68ac232 100644
--- a/src/util/cgroup.h
+++ b/src/util/cgroup.h
@@ -151,6 +151,9 @@ int virCgroupGetFreezerState(virCgroupPtr group, char **state);
int virCgroupSetCpusetMems(virCgroupPtr group, const char *mems);
int virCgroupGetCpusetMems(virCgroupPtr group, char **mems);
+int virCgroupSetCpusetCpus(virCgroupPtr group, const char *cpus);
+int virCgroupGetCpusetCpus(virCgroupPtr group, char **cpus);
+
int virCgroupRemove(virCgroupPtr group);
void virCgroupFree(virCgroupPtr *group);
--
1.7.10.2
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list