[libvirt] [PATCH] qemu: Error out if setting vcpu count would lead to invalid config
When the domain definition describes a machine with NUMA, setting the maximum vCPU count via the API might lead to an invalid config. Add a check that will forbid this until we add more advanced cpu config capabilities. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1327499 --- src/libvirt_private.syms | 1 + src/qemu/qemu_driver.c | 7 +++++++ 2 files changed, 8 insertions(+) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index ad5c382..769bdb7 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -673,6 +673,7 @@ virNodeDeviceObjUnlock; virDomainNumaCheckABIStability; virDomainNumaEquals; virDomainNumaFree; +virDomainNumaGetCPUCountTotal; virDomainNumaGetMaxCPUID; virDomainNumaGetMemorySize; virDomainNumaGetNodeCount; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 542d13c..44ec42d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4898,6 +4898,13 @@ qemuDomainSetVcpusFlags(virDomainPtr dom, unsigned int nvcpus, if (persistentDef) { if (flags & VIR_DOMAIN_VCPU_MAXIMUM) { + if (virDomainNumaGetCPUCountTotal(persistentDef->numa) > nvcpus) { + virReportError(VIR_ERR_INVALID_ARG, "%s", + _("Number of CPUs in <numa> exceeds the desired " + "maximum vcpu count")); + goto endjob; + } + if (virDomainDefSetVcpusMax(persistentDef, nvcpus) < 0) goto endjob; } else { -- 2.8.1
On Wed, Apr 27, 2016 at 04:48:43PM +0200, Peter Krempa wrote:
When the domain definition describes a machine with NUMA, setting the maximum vCPU count via the API might lead to an invalid config.
Add a check that will forbid this until we add more advanced cpu config capabilities.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1327499 --- src/libvirt_private.syms | 1 + src/qemu/qemu_driver.c | 7 +++++++ 2 files changed, 8 insertions(+)
diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index ad5c382..769bdb7 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -673,6 +673,7 @@ virNodeDeviceObjUnlock; virDomainNumaCheckABIStability; virDomainNumaEquals; virDomainNumaFree; +virDomainNumaGetCPUCountTotal; virDomainNumaGetMaxCPUID; virDomainNumaGetMemorySize; virDomainNumaGetNodeCount; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 542d13c..44ec42d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4898,6 +4898,13 @@ qemuDomainSetVcpusFlags(virDomainPtr dom, unsigned int nvcpus,
if (persistentDef) { if (flags & VIR_DOMAIN_VCPU_MAXIMUM) { + if (virDomainNumaGetCPUCountTotal(persistentDef->numa) > nvcpus) { + virReportError(VIR_ERR_INVALID_ARG, "%s", + _("Number of CPUs in <numa> exceeds the desired " + "maximum vcpu count")); + goto endjob; + } +
We should probably make sure that maxcpus == numacpus. But that's for the future. ACK to this one.
if (virDomainDefSetVcpusMax(persistentDef, nvcpus) < 0) goto endjob; } else { -- 2.8.1
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
On Wed, Apr 27, 2016 at 17:36:06 +0200, Martin Kletzander wrote:
On Wed, Apr 27, 2016 at 04:48:43PM +0200, Peter Krempa wrote:
When the domain definition describes a machine with NUMA, setting the maximum vCPU count via the API might lead to an invalid config.
Add a check that will forbid this until we add more advanced cpu config capabilities.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1327499 --- src/libvirt_private.syms | 1 + src/qemu/qemu_driver.c | 7 +++++++ 2 files changed, 8 insertions(+)
[...]
diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 542d13c..44ec42d 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -4898,6 +4898,13 @@ qemuDomainSetVcpusFlags(virDomainPtr dom, unsigned int nvcpus,
if (persistentDef) { if (flags & VIR_DOMAIN_VCPU_MAXIMUM) { + if (virDomainNumaGetCPUCountTotal(persistentDef->numa) > nvcpus) { + virReportError(VIR_ERR_INVALID_ARG, "%s", + _("Number of CPUs in <numa> exceeds the desired " + "maximum vcpu count")); + goto endjob; + } +
We should probably make sure that maxcpus == numacpus. But that's for the future. ACK to this one.
I plan to phase out the ability to set cpus for numa topologies using this API once we have a better approach. Thanks; Pushed. Peter
participants (2)
-
Martin Kletzander -
Peter Krempa