[libvirt] PATCH: Fix KVM maximum vCPU count
Libvirt currently returns '1' when asked for the maximum number of VCPUs supported for KVM guests. KVM long long ago gained SMP support, so this patch fixes it to return 16, whcih is the current supported number. I don't bother trying to detect old versions of KVM which are UP only because I find it hard to believe anyone will be using them. qemu_driver.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) Dan. diff -r b644023b5657 src/qemu_driver.c --- a/src/qemu_driver.c Sat May 10 13:12:04 2008 -0400 +++ b/src/qemu_driver.c Sat May 10 13:12:18 2008 -0400 @@ -1572,10 +1572,8 @@ if (STRCASEEQ(type, "qemu")) return 16; - /* XXX future KVM will support SMP. Need to probe - kernel to figure out KVM module version i guess */ if (STRCASEEQ(type, "kvm")) - return 1; + return 16; if (STRCASEEQ(type, "kqemu")) return 1; -- |: Red Hat, Engineering, Boston -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
Hi, Dan I think the qemu_driver should check KVM_API_VERSION for SMP support. How do you think? I am worrying about MAX_VCPUS=16 changes to 32 in future. Thanks Atsushi SAKAI "Daniel P. Berrange" <berrange@redhat.com> wrote:
I don't bother trying to detect old versions of KVM which are UP only because I find it hard to believe anyone will be using them.
On Tue, May 13, 2008 at 01:29:35PM +0900, Atsushi SAKAI wrote:
Hi, Dan
I think the qemu_driver should check KVM_API_VERSION for SMP support. How do you think? I am worrying about MAX_VCPUS=16 changes to 32 in future.
If it increases to 32, then we should certainly check based on something like KVM_API_VERSION. I just don't think its worth checking for the ancient versions which only support 1 cpu because no one will be seriously using them. Regards, Daniel. -- |: Red Hat, Engineering, Boston -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
Daniel P. Berrange wrote:
On Tue, May 13, 2008 at 01:29:35PM +0900, Atsushi SAKAI wrote:
Hi, Dan
I think the qemu_driver should check KVM_API_VERSION for SMP support. How do you think? I am worrying about MAX_VCPUS=16 changes to 32 in future.
If it increases to 32, then we should certainly check based on something like KVM_API_VERSION. I just don't think its worth checking for the ancient versions which only support 1 cpu because no one will be seriously using them.
It's worth noting that kvm-60 in f8 doesn't support 16 vcpus, it only supports 4. 16 vcpu support is only present in kvm-62 and later. So if this is ever backported to f8 and kvm isn't updated, we just replace one bug with another. - Cole
Hi, Dan Thank you for commenting. Of course, I agree your point and that you are commiting it. I am commenting it just because 2 month is not old ago in my feeling. Here Is the memo for my understanding. (Since I am not familiar with KVM) 2007/07/19 kvm-30 Guest SMP support (Parameter(KVM_MAX_VCPUS) changes from 1 to 4 in KVM-28) 2008/02/26 kvm-62 VCPUS changes 4 to 16 Thanks Atsushi SAKAI "Daniel P. Berrange" <berrange@redhat.com> wrote:
On Tue, May 13, 2008 at 01:29:35PM +0900, Atsushi SAKAI wrote:
Hi, Dan
I think the qemu_driver should check KVM_API_VERSION for SMP support. How do you think? I am worrying about MAX_VCPUS=16 changes to 32 in future.
If it increases to 32, then we should certainly check based on something like KVM_API_VERSION. I just don't think its worth checking for the ancient versions which only support 1 cpu because no one will be seriously using them.
Regards, Daniel. -- |: Red Hat, Engineering, Boston -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
On Wed, May 14, 2008 at 11:23:08AM +0900, Atsushi SAKAI wrote:
Hi, Dan
Thank you for commenting. Of course, I agree your point and that you are commiting it.
I am commenting it just because 2 month is not old ago in my feeling.
Here Is the memo for my understanding. (Since I am not familiar with KVM) 2007/07/19 kvm-30 Guest SMP support (Parameter(KVM_MAX_VCPUS) changes from 1 to 4 in KVM-28) 2008/02/26 kvm-62 VCPUS changes 4 to 16
I didn't realize it had changed from 4 -> 16 cpus so recently. I thought we'd gone straight from 1 -> 16. With this in mind, it is worth trying to decide a way to detect it right now. Dan. -- |: Red Hat, Engineering, Boston -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
Daniel P. Berrange wrote:
On Wed, May 14, 2008 at 11:23:08AM +0900, Atsushi SAKAI wrote:
Hi, Dan
Thank you for commenting. Of course, I agree your point and that you are commiting it.
I am commenting it just because 2 month is not old ago in my feeling.
Here Is the memo for my understanding. (Since I am not familiar with KVM) 2007/07/19 kvm-30 Guest SMP support (Parameter(KVM_MAX_VCPUS) changes from 1 to 4 in KVM-28) 2008/02/26 kvm-62 VCPUS changes 4 to 16
I didn't realize it had changed from 4 -> 16 cpus so recently. I thought we'd gone straight from 1 -> 16. With this in mind, it is worth trying to decide a way to detect it right now.
Dan.
I actually have a patch mostly complete that pulls the kvm version from qemu-kvm similar to how we grab the version from the qemu binary. I can clean it up and post it if that seems like the way to go. - Cole
On Wed, May 14, 2008 at 09:32:54AM -0400, Cole Robinson wrote:
Daniel P. Berrange wrote:
On Wed, May 14, 2008 at 11:23:08AM +0900, Atsushi SAKAI wrote:
Hi, Dan
Thank you for commenting. Of course, I agree your point and that you are commiting it.
I am commenting it just because 2 month is not old ago in my feeling.
Here Is the memo for my understanding. (Since I am not familiar with KVM) 2007/07/19 kvm-30 Guest SMP support (Parameter(KVM_MAX_VCPUS) changes from 1 to 4 in KVM-28) 2008/02/26 kvm-62 VCPUS changes 4 to 16
I didn't realize it had changed from 4 -> 16 cpus so recently. I thought we'd gone straight from 1 -> 16. With this in mind, it is worth trying to decide a way to detect it right now.
Dan.
I actually have a patch mostly complete that pulls the kvm version from qemu-kvm similar to how we grab the version from the qemu binary. I can clean it up and post it if that seems like the way to go.
Yes, that would be useful. Regards, Dan. -- |: Red Hat, Engineering, Boston -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|
participants (3)
-
Atsushi SAKAI -
Cole Robinson -
Daniel P. Berrange