[V8,00/15] KVM: MIPS: Add Loongson-3 support (Host Side)

Message

Huacai Chen May 24, 2020, 11:13 a.m. UTC

We are preparing to add KVM support for Loongson-3. VZ extension is
fully supported in Loongson-3A R4+, and we will not care about old CPUs
(at least now). We already have a full functional Linux kernel (based
on Linux-5.4.x LTS) and QEMU (based on 5.0.0) and their git repositories
are here:

QEMU: https://github.com/chenhuacai/qemu
Kernel: https://github.com/chenhuacai/linux

Of course these two repositories need to be rework and not suitable for
upstream (especially the commits need to be splitted). We show them here
is just to tell others what we have done, and how KVM/Loongson will look
like.

Our plan is make the KVM host side be upstream first, and after that,
we will make the KVM guest side and QEMU emulator be upstream.

V1 -> V2:
1, Remove "mips: define pud_index() regardless of page table folding"
   because it has been applied.
2, Make Loongson-specific code be guarded by CONFIG_CPU_LOONGSON64.

V2 -> V3:
1, Emulate a reduced feature list of CPUCFG.
2, Fix all possible checkpatch.pl errors and warnings.

V3 -> V4:
1, Emulate LOONGSON_CFG0/LOONGSON_CFG3 in CPUCFG correctly.
2, Update commit messages to explain Loongson-3 Virtual IPI.
3, Add Reviewed-by: Aleksandar Markovic <[e-mail]>.

V4 -> V5:
1, Fix a typo.
2, Update MAINTAINERS.

V5 -> V6:
1, Fix a mismatch during rebasing.
2, Add Acked-by: Thomas Bogendoerfer <[e-mail]>.

V6 -> V7:
1, Rebase on latest mips-next (Config6 feature bits definition updated).

V7 -> V8:
1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
   https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
   need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
   of page table folding"), which has already been in mips tree but not
   in kvm tree.

Xing Li(2):
 KVM: MIPS: Define KVM_ENTRYHI_ASID to cpu_asid_mask(&boot_cpu_data)
 KVM: MIPS: Fix VPN2_MASK definition for variable cpu_vmbits

Huacai Chen(13):
 KVM: MIPS: Increase KVM_MAX_VCPUS and KVM_USER_MEM_SLOTS to 16
 KVM: MIPS: Add EVENTFD support which is needed by VHOST
 KVM: MIPS: Use lddir/ldpte instructions to lookup gpa_mm.pgd
 KVM: MIPS: Introduce and use cpu_guest_has_ldpte
 KVM: MIPS: Use root tlb to control guest's CCA for Loongson-3
 KVM: MIPS: Let indexed cacheops cause guest exit on Loongson-3
 KVM: MIPS: Add more types of virtual interrupts
 KVM: MIPS: Add Loongson-3 Virtual IPI interrupt support
 KVM: MIPS: Add CPUCFG emulation for Loongson-3
 KVM: MIPS: Add CONFIG6 and DIAG registers emulation
 KVM: MIPS: Add more MMIO load/store instructions emulation
 KVM: MIPS: Enable KVM support for Loongson-3
 MAINTAINERS: Update KVM/MIPS maintainers

Signed-off-by: Huacai Chen <[e-mail]>
---
 MAINTAINERS                          |   4 +-
 arch/mips/Kconfig                    |   1 +
 arch/mips/include/asm/cpu-features.h |   3 +
 arch/mips/include/asm/kvm_host.h     |  52 +++-
 arch/mips/include/asm/mipsregs.h     |   7 +
 arch/mips/include/uapi/asm/inst.h    |  37 +++
 arch/mips/kernel/cpu-probe.c         |   2 +
 arch/mips/kvm/Kconfig                |   1 +
 arch/mips/kvm/Makefile               |   5 +-
 arch/mips/kvm/emulate.c              | 503 ++++++++++++++++++++++++++++++++++-
 arch/mips/kvm/entry.c                |  19 +-
 arch/mips/kvm/interrupt.c            |  93 +------
 arch/mips/kvm/interrupt.h            |  14 +-
 arch/mips/kvm/loongson_ipi.c         | 214 +++++++++++++++
 arch/mips/kvm/mips.c                 |  49 +++-
 arch/mips/kvm/tlb.c                  |  41 +++
 arch/mips/kvm/trap_emul.c            |   3 +
 arch/mips/kvm/vz.c                   | 237 ++++++++++++-----
 18 files changed, 1118 insertions(+), 167 deletions(-)
 create mode 100644 arch/mips/kvm/loongson_ipi.c
--
2.7.0

Comments

Aleksandar Markovic May 25, 2020, 12:26 a.m. UTC | #1

> V7 -> V8:
> 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
>    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
>    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
>    of page table folding"), which has already been in mips tree but not
>    in kvm tree.
>

Huacai,

I do support and salute the series (as I always did), as I see it as a
giant step forward for KVM for MIPS.

However, in general, I think any series should not depend on "pick
that patch from another tree", and should be a stand-alone unit that
yields to successful build and desired functionality. If there is a
dependency like you described, the patch in question, in my opinion,
should be integrated into the series in question. Git is even smart
enough that it recognizes the same patch has been applied before, so
integration of another tree would not be exposed to problems.

From the point of view of synchronizing with QEMU part, and the timing
issues wrt kernel and QEMU releases, I want to stress that it is
better that this series is integrated sooner rather than later. In
other words, I think that potential Paolo's KVM pull request should
happen before Thomas' mips-next pull request (Paolo could include
"mips: define pud_index() regardless of page table folding", and
Thomas could simply omit it).

But, that said, I don't feel I should impose my opinion to others
here. Take my statements just as advises. I defer the decision on how
to proceed with the integration of this series entirely to Paolo and
Thomas.

Yours,
Aleksandar

Huacai Chen May 25, 2020, 1:23 a.m. UTC | #2

Hi, Aleksandar,

On Mon, May 25, 2020 at 8:26 AM Aleksandar Markovic
<[e-mail]> wrote:
>
> > V7 -> V8:
> > 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
> >    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
> >    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
> >    of page table folding"), which has already been in mips tree but not
> >    in kvm tree.
> >
>
> Huacai,
>
> I do support and salute the series (as I always did), as I see it as a
> giant step forward for KVM for MIPS.
>
> However, in general, I think any series should not depend on "pick
> that patch from another tree", and should be a stand-alone unit that
> yields to successful build and desired functionality. If there is a
> dependency like you described, the patch in question, in my opinion,
> should be integrated into the series in question. Git is even smart
> enough that it recognizes the same patch has been applied before, so
> integration of another tree would not be exposed to problems.
>
> From the point of view of synchronizing with QEMU part, and the timing
> issues wrt kernel and QEMU releases, I want to stress that it is
> better that this series is integrated sooner rather than later. In
> other words, I think that potential Paolo's KVM pull request should
> happen before Thomas' mips-next pull request (Paolo could include
> "mips: define pud_index() regardless of page table folding", and
> Thomas could simply omit it).
>
> But, that said, I don't feel I should impose my opinion to others
> here. Take my statements just as advises. I defer the decision on how
> to proceed with the integration of this series entirely to Paolo and
> Thomas.
I think I lack some experience of cross-subsystem development, so in
V8 I only adjust the context of my own patches to let the series be
applied on kvm tree, but didn't consider patches already in other
trees (Moreover, that patch is not mine). So, should I send a V9 that
take commit 3fbfb4585bfd4ff3 together?

Thanks,
Huacai
>
> Yours,
> Aleksandar

Aleksandar Markovic May 25, 2020, 6:37 a.m. UTC | #3

пон, 25. мај 2020. у 03:23 Huacai Chen <[e-mail]> је написао/ла:
>
> Hi, Aleksandar,
>
> On Mon, May 25, 2020 at 8:26 AM Aleksandar Markovic
> <[e-mail]> wrote:
> >
> > > V7 -> V8:
> > > 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
> > >    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
> > >    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
> > >    of page table folding"), which has already been in mips tree but not
> > >    in kvm tree.
> > >
> >
> > Huacai,
> >
> > I do support and salute the series (as I always did), as I see it as a
> > giant step forward for KVM for MIPS.
> >
> > However, in general, I think any series should not depend on "pick
> > that patch from another tree", and should be a stand-alone unit that
> > yields to successful build and desired functionality. If there is a
> > dependency like you described, the patch in question, in my opinion,
> > should be integrated into the series in question. Git is even smart
> > enough that it recognizes the same patch has been applied before, so
> > integration of another tree would not be exposed to problems.
> >
> > From the point of view of synchronizing with QEMU part, and the timing
> > issues wrt kernel and QEMU releases, I want to stress that it is
> > better that this series is integrated sooner rather than later. In
> > other words, I think that potential Paolo's KVM pull request should
> > happen before Thomas' mips-next pull request (Paolo could include
> > "mips: define pud_index() regardless of page table folding", and
> > Thomas could simply omit it).
> >
> > But, that said, I don't feel I should impose my opinion to others
> > here. Take my statements just as advises. I defer the decision on how
> > to proceed with the integration of this series entirely to Paolo and
> > Thomas.
> I think I lack some experience of cross-subsystem development, so in
> V8 I only adjust the context of my own patches to let the series be
> applied on kvm tree, but didn't consider patches already in other
> trees (Moreover, that patch is not mine). So, should I send a V9 that
> take commit 3fbfb4585bfd4ff3 together?
>

No, I think we should wait for Paolo's feedback.

> Thanks,
> Huacai
> >
> > Yours,
> > Aleksandar

Aleksandar Markovic May 26, 2020, 11:44 a.m. UTC | #4

пон, 25. мај 2020. у 08:37 Aleksandar Markovic
<[e-mail]> је написао/ла:
>
> пон, 25. мај 2020. у 03:23 Huacai Chen <[e-mail]> је написао/ла:
> >
> > Hi, Aleksandar,
> >
> > On Mon, May 25, 2020 at 8:26 AM Aleksandar Markovic
> > <[e-mail]> wrote:
> > >
> > > > V7 -> V8:
> > > > 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
> > > >    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
> > > >    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
> > > >    of page table folding"), which has already been in mips tree but not
> > > >    in kvm tree.
> > > >
> > >
> > > Huacai,
> > >
> > > I do support and salute the series (as I always did), as I see it as a
> > > giant step forward for KVM for MIPS.
> > >
> > > However, in general, I think any series should not depend on "pick
> > > that patch from another tree", and should be a stand-alone unit that
> > > yields to successful build and desired functionality. If there is a
> > > dependency like you described, the patch in question, in my opinion,
> > > should be integrated into the series in question. Git is even smart
> > > enough that it recognizes the same patch has been applied before, so
> > > integration of another tree would not be exposed to problems.
> > >
> > > From the point of view of synchronizing with QEMU part, and the timing
> > > issues wrt kernel and QEMU releases, I want to stress that it is
> > > better that this series is integrated sooner rather than later. In
> > > other words, I think that potential Paolo's KVM pull request should
> > > happen before Thomas' mips-next pull request (Paolo could include
> > > "mips: define pud_index() regardless of page table folding", and
> > > Thomas could simply omit it).
> > >
> > > But, that said, I don't feel I should impose my opinion to others
> > > here. Take my statements just as advises. I defer the decision on how
> > > to proceed with the integration of this series entirely to Paolo and
> > > Thomas.
> > I think I lack some experience of cross-subsystem development, so in
> > V8 I only adjust the context of my own patches to let the series be
> > applied on kvm tree, but didn't consider patches already in other
> > trees (Moreover, that patch is not mine). So, should I send a V9 that
> > take commit 3fbfb4585bfd4ff3 together?
> >
>
> No, I think we should wait for Paolo's feedback.
>

Paolo, can you please give us some guidance on integration of this
series, and on future workflow?

Since, the reality is, you are the most experienced between us on
these topics, so you know what is the optimal organization. I gather,
it is very important for us in KVM/MIPS to adopt the efficient and
no-nonsens practices, that you know how to choose the best.

I truly appreciate any future response of yours!

Aleksandar



> > Thanks,
> > Huacai
> > >
> > > Yours,
> > > Aleksandar

Aleksandar Markovic May 28, 2020, 7:35 a.m. UTC | #5

нед, 24. мај 2020. у 13:13 Huacai Chen <[e-mail]> је написао/ла:
>
> We are preparing to add KVM support for Loongson-3. VZ extension is
> fully supported in Loongson-3A R4+, and we will not care about old CPUs
> (at least now). We already have a full functional Linux kernel (based
> on Linux-5.4.x LTS) and QEMU (based on 5.0.0) and their git repositories
> are here:
>
> QEMU: https://github.com/chenhuacai/qemu
> Kernel: https://github.com/chenhuacai/linux
>
> Of course these two repositories need to be rework and not suitable for
> upstream (especially the commits need to be splitted). We show them here
> is just to tell others what we have done, and how KVM/Loongson will look
> like.
>
> Our plan is make the KVM host side be upstream first, and after that,
> we will make the KVM guest side and QEMU emulator be upstream.
>

Huacai, Paolo, Thomas,

As you know, there are number of Huacai's patches to QEMU, related
to this series, actually we could call them counterparts to this series.

Regarding QEMU upstreaming, unless someone objects, I plan to send
pull request that will include some of Huacai-s QEMU patches in next few
days - those that are of general nature, do not rely on any kernel changes,
and passed review process. The rest of Huacai's QEMU patches will wait
for this series to be integrated in kernel upstream.

Please give Huacai and myself some direction and path to go forward with
this series. Please take into account that we inherit KVM for MIPS in an
"orphaned" state, and do not know exactly all details how to "adopt" it.
Forgive us if we made some missteps along the route.

We would like to establish a long-lasting, and high-quality, workflow that
will fit well to both KVM and MIPS kernel development.

Please help us do it.

Thanks in advance!

Aleksandar



> V1 -> V2:
> 1, Remove "mips: define pud_index() regardless of page table folding"
>    because it has been applied.
> 2, Make Loongson-specific code be guarded by CONFIG_CPU_LOONGSON64.
>
> V2 -> V3:
> 1, Emulate a reduced feature list of CPUCFG.
> 2, Fix all possible checkpatch.pl errors and warnings.
>
> V3 -> V4:
> 1, Emulate LOONGSON_CFG0/LOONGSON_CFG3 in CPUCFG correctly.
> 2, Update commit messages to explain Loongson-3 Virtual IPI.
> 3, Add Reviewed-by: Aleksandar Markovic <[e-mail]>.
>
> V4 -> V5:
> 1, Fix a typo.
> 2, Update MAINTAINERS.
>
> V5 -> V6:
> 1, Fix a mismatch during rebasing.
> 2, Add Acked-by: Thomas Bogendoerfer <[e-mail]>.
>
> V6 -> V7:
> 1, Rebase on latest mips-next (Config6 feature bits definition updated).
>
> V7 -> V8:
> 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
>    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
>    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
>    of page table folding"), which has already been in mips tree but not
>    in kvm tree.
>
> Xing Li(2):
>  KVM: MIPS: Define KVM_ENTRYHI_ASID to cpu_asid_mask(&boot_cpu_data)
>  KVM: MIPS: Fix VPN2_MASK definition for variable cpu_vmbits
>
> Huacai Chen(13):
>  KVM: MIPS: Increase KVM_MAX_VCPUS and KVM_USER_MEM_SLOTS to 16
>  KVM: MIPS: Add EVENTFD support which is needed by VHOST
>  KVM: MIPS: Use lddir/ldpte instructions to lookup gpa_mm.pgd
>  KVM: MIPS: Introduce and use cpu_guest_has_ldpte
>  KVM: MIPS: Use root tlb to control guest's CCA for Loongson-3
>  KVM: MIPS: Let indexed cacheops cause guest exit on Loongson-3
>  KVM: MIPS: Add more types of virtual interrupts
>  KVM: MIPS: Add Loongson-3 Virtual IPI interrupt support
>  KVM: MIPS: Add CPUCFG emulation for Loongson-3
>  KVM: MIPS: Add CONFIG6 and DIAG registers emulation
>  KVM: MIPS: Add more MMIO load/store instructions emulation
>  KVM: MIPS: Enable KVM support for Loongson-3
>  MAINTAINERS: Update KVM/MIPS maintainers
>
> Signed-off-by: Huacai Chen <[e-mail]>
> ---
>  MAINTAINERS                          |   4 +-
>  arch/mips/Kconfig                    |   1 +
>  arch/mips/include/asm/cpu-features.h |   3 +
>  arch/mips/include/asm/kvm_host.h     |  52 +++-
>  arch/mips/include/asm/mipsregs.h     |   7 +
>  arch/mips/include/uapi/asm/inst.h    |  37 +++
>  arch/mips/kernel/cpu-probe.c         |   2 +
>  arch/mips/kvm/Kconfig                |   1 +
>  arch/mips/kvm/Makefile               |   5 +-
>  arch/mips/kvm/emulate.c              | 503 ++++++++++++++++++++++++++++++++++-
>  arch/mips/kvm/entry.c                |  19 +-
>  arch/mips/kvm/interrupt.c            |  93 +------
>  arch/mips/kvm/interrupt.h            |  14 +-
>  arch/mips/kvm/loongson_ipi.c         | 214 +++++++++++++++
>  arch/mips/kvm/mips.c                 |  49 +++-
>  arch/mips/kvm/tlb.c                  |  41 +++
>  arch/mips/kvm/trap_emul.c            |   3 +
>  arch/mips/kvm/vz.c                   | 237 ++++++++++++-----
>  18 files changed, 1118 insertions(+), 167 deletions(-)
>  create mode 100644 arch/mips/kvm/loongson_ipi.c
> --
> 2.7.0

Huacai Chen June 4, 2020, 6:57 a.m. UTC | #6

Hi, Paolo

On Thu, May 28, 2020 at 3:36 PM Aleksandar Markovic
<[e-mail]> wrote:
>
> нед, 24. мај 2020. у 13:13 Huacai Chen <[e-mail]> је написао/ла:
> >
> > We are preparing to add KVM support for Loongson-3. VZ extension is
> > fully supported in Loongson-3A R4+, and we will not care about old CPUs
> > (at least now). We already have a full functional Linux kernel (based
> > on Linux-5.4.x LTS) and QEMU (based on 5.0.0) and their git repositories
> > are here:
> >
> > QEMU: https://github.com/chenhuacai/qemu
> > Kernel: https://github.com/chenhuacai/linux
> >
> > Of course these two repositories need to be rework and not suitable for
> > upstream (especially the commits need to be splitted). We show them here
> > is just to tell others what we have done, and how KVM/Loongson will look
> > like.
> >
> > Our plan is make the KVM host side be upstream first, and after that,
> > we will make the KVM guest side and QEMU emulator be upstream.
> >
>
> Huacai, Paolo, Thomas,
>
> As you know, there are number of Huacai's patches to QEMU, related
> to this series, actually we could call them counterparts to this series.
>
> Regarding QEMU upstreaming, unless someone objects, I plan to send
> pull request that will include some of Huacai-s QEMU patches in next few
> days - those that are of general nature, do not rely on any kernel changes,
> and passed review process. The rest of Huacai's QEMU patches will wait
> for this series to be integrated in kernel upstream.
>
> Please give Huacai and myself some direction and path to go forward with
> this series. Please take into account that we inherit KVM for MIPS in an
> "orphaned" state, and do not know exactly all details how to "adopt" it.
> Forgive us if we made some missteps along the route.
>
> We would like to establish a long-lasting, and high-quality, workflow that
> will fit well to both KVM and MIPS kernel development.
>
> Please help us do it.
>
> Thanks in advance!
>
> Aleksandar
>
If you will take this series after MIPS tree be merged into upstream,
maybe V7 is better for you.

Huacai
>
>
> > V1 -> V2:
> > 1, Remove "mips: define pud_index() regardless of page table folding"
> >    because it has been applied.
> > 2, Make Loongson-specific code be guarded by CONFIG_CPU_LOONGSON64.
> >
> > V2 -> V3:
> > 1, Emulate a reduced feature list of CPUCFG.
> > 2, Fix all possible checkpatch.pl errors and warnings.
> >
> > V3 -> V4:
> > 1, Emulate LOONGSON_CFG0/LOONGSON_CFG3 in CPUCFG correctly.
> > 2, Update commit messages to explain Loongson-3 Virtual IPI.
> > 3, Add Reviewed-by: Aleksandar Markovic <[e-mail]>.
> >
> > V4 -> V5:
> > 1, Fix a typo.
> > 2, Update MAINTAINERS.
> >
> > V5 -> V6:
> > 1, Fix a mismatch during rebasing.
> > 2, Add Acked-by: Thomas Bogendoerfer <[e-mail]>.
> >
> > V6 -> V7:
> > 1, Rebase on latest mips-next (Config6 feature bits definition updated).
> >
> > V7 -> V8:
> > 1, Rebase to be applied on kvm tree, i.e., the linux-next branch of
> >    https://git.kernel.org/pub/scm/virt/kvm/kvm.git/. Building KVM/MIPS
> >    need commit 3fbfb4585bfd4ff34e ("mips: define pud_index() regardless
> >    of page table folding"), which has already been in mips tree but not
> >    in kvm tree.
> >
> > Xing Li(2):
> >  KVM: MIPS: Define KVM_ENTRYHI_ASID to cpu_asid_mask(&boot_cpu_data)
> >  KVM: MIPS: Fix VPN2_MASK definition for variable cpu_vmbits
> >
> > Huacai Chen(13):
> >  KVM: MIPS: Increase KVM_MAX_VCPUS and KVM_USER_MEM_SLOTS to 16
> >  KVM: MIPS: Add EVENTFD support which is needed by VHOST
> >  KVM: MIPS: Use lddir/ldpte instructions to lookup gpa_mm.pgd
> >  KVM: MIPS: Introduce and use cpu_guest_has_ldpte
> >  KVM: MIPS: Use root tlb to control guest's CCA for Loongson-3
> >  KVM: MIPS: Let indexed cacheops cause guest exit on Loongson-3
> >  KVM: MIPS: Add more types of virtual interrupts
> >  KVM: MIPS: Add Loongson-3 Virtual IPI interrupt support
> >  KVM: MIPS: Add CPUCFG emulation for Loongson-3
> >  KVM: MIPS: Add CONFIG6 and DIAG registers emulation
> >  KVM: MIPS: Add more MMIO load/store instructions emulation
> >  KVM: MIPS: Enable KVM support for Loongson-3
> >  MAINTAINERS: Update KVM/MIPS maintainers
> >
> > Signed-off-by: Huacai Chen <[e-mail]>
> > ---
> >  MAINTAINERS                          |   4 +-
> >  arch/mips/Kconfig                    |   1 +
> >  arch/mips/include/asm/cpu-features.h |   3 +
> >  arch/mips/include/asm/kvm_host.h     |  52 +++-
> >  arch/mips/include/asm/mipsregs.h     |   7 +
> >  arch/mips/include/uapi/asm/inst.h    |  37 +++
> >  arch/mips/kernel/cpu-probe.c         |   2 +
> >  arch/mips/kvm/Kconfig                |   1 +
> >  arch/mips/kvm/Makefile               |   5 +-
> >  arch/mips/kvm/emulate.c              | 503 ++++++++++++++++++++++++++++++++++-
> >  arch/mips/kvm/entry.c                |  19 +-
> >  arch/mips/kvm/interrupt.c            |  93 +------
> >  arch/mips/kvm/interrupt.h            |  14 +-
> >  arch/mips/kvm/loongson_ipi.c         | 214 +++++++++++++++
> >  arch/mips/kvm/mips.c                 |  49 +++-
> >  arch/mips/kvm/tlb.c                  |  41 +++
> >  arch/mips/kvm/trap_emul.c            |   3 +
> >  arch/mips/kvm/vz.c                   | 237 ++++++++++++-----
> >  18 files changed, 1118 insertions(+), 167 deletions(-)
> >  create mode 100644 arch/mips/kvm/loongson_ipi.c
> > --
> > 2.7.0