From: "Radim Krčmář" <rkrcmar@redhat.com>
To: kvm@vger.kernel.org
Cc: Paolo Bonzini <pbonzini@redhat.com>,
"Lan, Tianyu" <tianyu.lan@intel.com>,
Igor Mammedov <imammedo@redhat.com>,
Jan Kiszka <jan.kiszka@web.de>, Peter Xu <peterx@redhat.com>
Subject: [RFC 0/9] KVM: x86: break the xAPIC barrier
Date: Fri, 6 May 2016 22:53:56 +0200 [thread overview]
Message-ID: <1462568045-31085-1-git-send-email-rkrcmar@redhat.com> (raw)
Please give attention to [5/9], which introduces a new userpace
interface, and to [7/9], which codifies a very unfourtunate interface
choice that was made when introducing paravirtual x2APIC.
The rest makes APIC ID > 255 work in KVM.
I've tested latest upstream and rhel7 kernels as guests in physical and
cluster x2APIC modes and there were no issues that could be tracked to
these patches. If interrupt remapping + split irqchip didn't screw up
EOI, then everything useful would have worked.
It was a setup from Igor's latest x2APIC QEMU series, so 2 VCPUs in
total, first has id 0, the second has 280. I used v4 of Peter's IR
patches with Jan's EIM on top, because newer versions got stuck at boot
even without other patches.
See the qemu-devel-list for potential uses,
"[RFC 0/4] APIC, IOMMU, KVM: add x2APIC interface".
Radim Krčmář (9):
KVM: x86: add kvm_apic_map_get_dest_lapic
KVM: x86: dynamic kvm_apic_map
KVM: x86: use u16 for logical VCPU mask in lapic
KVM: x86: use generic function for MSI parsing
KVM: support x2APIC ID in userspace routes
KVM: x86: directly call recalculate_apic_map on lapic restore
KVM: x86: use proper format of APIC ID register
KVM: x86: reset lapic base in kvm_lapic_reset
KVM: bump MAX_VCPUS
Documentation/virtual/kvm/api.txt | 17 +-
arch/x86/include/asm/kvm_host.h | 11 +-
arch/x86/kvm/irq_comm.c | 26 +--
arch/x86/kvm/lapic.c | 354 ++++++++++++++++++--------------------
arch/x86/kvm/lapic.h | 9 +-
arch/x86/kvm/vmx.c | 7 +-
arch/x86/kvm/x86.c | 3 +
include/uapi/linux/kvm.h | 5 +
virt/kvm/irqchip.c | 6 +-
9 files changed, 228 insertions(+), 210 deletions(-)
--
2.8.2
next reply other threads:[~2016-05-06 20:54 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-06 20:53 Radim Krčmář [this message]
2016-05-06 20:53 ` [PATCH 1/9] KVM: x86: add kvm_apic_map_get_dest_lapic Radim Krčmář
2016-05-19 6:36 ` Peter Xu
2016-05-25 16:02 ` Radim Krčmář
2016-05-26 11:58 ` Peter Xu
2016-05-06 20:53 ` [PATCH 2/9] KVM: x86: dynamic kvm_apic_map Radim Krčmář
2016-05-23 8:04 ` Peter Xu
2016-05-25 16:15 ` Radim Krčmář
2016-05-30 5:24 ` Peter Xu
2016-05-06 20:53 ` [PATCH 3/9] KVM: x86: use u16 for logical VCPU mask in lapic Radim Krčmář
2016-05-06 20:54 ` [PATCH 4/9] KVM: x86: use generic function for MSI parsing Radim Krčmář
2016-05-06 20:54 ` [PATCH 5/9] KVM: support x2APIC ID in userspace routes Radim Krčmář
2016-05-06 20:54 ` [PATCH 6/9] KVM: x86: directly call recalculate_apic_map on lapic restore Radim Krčmář
2016-05-23 8:30 ` Peter Xu
2016-05-06 20:54 ` [PATCH 7/9] KVM: x86: use proper format of APIC ID register Radim Krčmář
2016-05-17 15:34 ` Paolo Bonzini
2016-05-25 16:30 ` Radim Krčmář
2016-05-06 20:54 ` [PATCH 8/9] KVM: x86: reset lapic base in kvm_lapic_reset Radim Krčmář
2016-05-06 20:54 ` [PATCH 9/9] KVM: bump MAX_VCPUS Radim Krčmář
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1462568045-31085-1-git-send-email-rkrcmar@redhat.com \
--to=rkrcmar@redhat.com \
--cc=imammedo@redhat.com \
--cc=jan.kiszka@web.de \
--cc=kvm@vger.kernel.org \
--cc=pbonzini@redhat.com \
--cc=peterx@redhat.com \
--cc=tianyu.lan@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).