Re: [Qemu-devel] [PATCH v4 00/13] pc: q35: x2APIC support in kvm_apic mode

[Igor Mammedov]

On Wed, 19 Oct 2016 14:29:22 +0100
"Daniel P. Berrange" <address@hidden> wrote:

> On Wed, Oct 19, 2016 at 02:05:30PM +0200, Igor Mammedov wrote:
> > 
> > Changes since v3:
> >   - fix endianness issues when filling MADT/SRAT entries (Eduardo)
> >   - squash "acpi: cphp: support x2APIC entry in cpu._MAT" into
> >     "pc: acpi: x2APIC support for MADT table and _MAT method" (Eduardo)
> >   - keep assert() as it doesn't affect x2APIC cpus in bochs_bios_init() 
> > (Eduardo)
> >   - restore kvm_has_x2apic_api() and use it to avoid side-effects
> >     of kvm_enable_x2apic(). x2APIC API will be enabled by iommu
> >     if it's present or not enabled at all. (Radim)
> >   - add intel iommu example to error message (Eduardo)
> > 
> > Changes since v2:
> >   - rebase on top of EIM fixes
> >   - drop kvm_has_x2apic_ids() and reuse kvm_enable_x2apic()
> >     from Radim's EIM fixes
> >   - fix hang on reboot in BIOS due to not updated 'etc/boot-cpus' fwcfg file
> >     after CPU hotplug
> >   - drop not used anymore pc_present_cpus_count() and incrementally count
> >     present VCPUs as they are added/removed at (un)plug callbacks time
> > 
> > Changes since v1:
> >   - rebase on top of 2.7
> >   - drop add 2.8 machine and linux headers update patches
> >   - drop numa related patches (will post separately as unrelated)
> >   - change default mc->maxcpus only for q35
> > 
> > Changes since RFC:
> >   - use new KVM_CAP_X2APIC_API to detect x2APIC IDs support
> >   - rebase on top of 2.7-rc1, since many deps were merged
> >   - fix etc/boot-cpus to account for -device provided cpus
> >   - include not yet merged _PXM fix as prereq
> >   - add 2.8 machine type and bump up maxcpus count since it
> > 
> > Series extends current CPU/kvm_apic/Q35 machine
> > code to support x2APIC and upto 288 VCPUs when QEMU
> > is used with KVM's lapic.
> > 
> > Due to FW_CFG_MAX_CPUS (which is actually apic_id_limit)
> > being limited to uint16_t, the max possible APIC ID is
> > limitted to 2^16 with this series but that should
> > be sufficient for bumping VCPUs number for quite a while.  
> 
> Not a problem, just curiosity - 2^16 gives us 65536, so where
> is the 288 limit coming from ?
It comes from KVM but it basically limited to HW we can test it on
(even though it might need a little bit of extra work to handle
large number of CPUs).

> 
> Regards,
> Daniel