* [GIT PULL] KVM/arm64 fixes for 7.4, take #3
@ 2026-11-19 12:56 Marc Zyngier
2026-11-19 17:02 ` Paolo Bonzini
0 siblings, 1 reply; 2+ messages in thread
From: Marc Zyngier @ 2026-11-19 12:56 UTC (permalink / raw)
To: Paolo Bonzini
Cc: Catalin Marinas, Colton Lewis, Dan Carpenter, Fuad Tabba,
Michael Reeves, Mostafa Saleh, Myria Sarvay, Oliver Upton,
Vincent Donnefort, Will Deacon, Xuheng Li, Yureka Lilian,
Steffen Eiden, Joey Gouly, Suzuki K Poulose, Zenghui Yu, kvmarm,
linux-arm-kernel, kvm
Paolo,
As promissed in my previous missive, I'm back with a few more
presents...
This round of fixes is mostly a Dan Carpenter special, as they seem to
have hit the jackpot with some new formal analysis tool, and found a
number of issues in NV. I am very surprised (/s).
The rest is a bunch of pKVM fixes and workarounds for ugly hardware.
I will be back next week same time for more of the same!
Please pull,
M.
The following changes since commit 181a701fc14d3d468c603b5ecfb9f12115c5fb72:
Merge branch kvm-arm64/misc-7.4-1 into kvmarm-master/fixes (2026-11-03 18:14:37 +0000)
are available in the Git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm.git tags/kvmarm-fixes-7.4-3
for you to fetch changes up to 3385a832525527801012f5993d1f6a7a3d23092c:
Merge branch kvm-arm64/mmu-fixes-7.4-2 into kvmarm-master/fixes (2026-11-19 11:16:29 +0000)
----------------------------------------------------------------
KVM/arm64 fixes for 7.4, take #3
* MMU fixes:
- Fix arbitrary truncation to 32bit when dealing with pool size
- Correctly order page pin refconting
- Prevent a HYP block mapping from being replaced with a table
- Fix NV Range invalidation by extending the boundaries to the
next block mapping
- Make sure the correct hw_mmu gets picked on the first run,
should userspace restore a nested context between init and
the first vcpu_load().
- Reject TLBI IPAS2E1 that fall outside of the S2 range
* System register fixes:
- Correctly forward POE/PIE/TVAL EL2 register traps
- Add a sanity check for missing trap forwarding
- Correctly hide generic PAuth for pKVM protected guests when
the feature PAuth has been exluced from the feature set
- Make sure that HCR_EL2.NV1 is always RES0
* Misc fixes:
- Flag Apple M3, M4 and M6 CPUs as having a broken vgic and
a wonky PMUv3 emulation.
- Correctly handle 64bit offsets into an MMIO region from
userspace
- Ensure that a vcpu is loaded when injecting an event
----------------------------------------------------------------
Colton Lewis (2):
KVM: arm64: Force HCR_EL2.NV1 to RES0 unconditionally
KVM: arm64: selftests: Verify HCR_EL2.NV1 is RES0 when E2H0 is NI_NV1
Dan Carpenter (9):
KVM: arm64: nv: Invalidate whole L1 blocks on TLBI RIPAS2E1
KVM: arm64: vgic: Don't truncate offsets in vgic_find_mmio_region()
KVM: arm64: nv: Initialize hw_mmu to NULL for nested vCPUs
KVM: arm64: nv: Forward PIR_EL2, PIRE0_EL2 and POR_EL2 traps
KVM: arm64: nv: Forward CNTP_TVAL_EL02 and CNTV_TVAL_EL02 traps
KVM: arm64: nv: Report EL2 registers with no trap forwarding
KVM: arm64: Load the vcpu in KVM_SET_VCPU_EVENTS
KVM: arm64: nv: Fix NULL dereference in kvm_pmu_nested_transition()
KVM: arm64: nv: Ignore TLBI IPAS2E1 above the IPA limit
Fuad Tabba (1):
KVM: arm64: Hide generic ptrauth from protected VMs without ptrauth
Marc Zyngier (3):
Merge branch kvm-arm64/misc-7.4-2 into kvmarm-master/fixes
Merge branch kvm-arm64/sysreg-fixes-7.4-2 into kvmarm-master/fixes
Merge branch kvm-arm64/mmu-fixes-7.4-2 into kvmarm-master/fixes
Michael Reeves (1):
arm64: Apply existing Apple Silicon CPU workarounds to newer CPUs
Mostafa Saleh (1):
KVM: arm64: Use consistent type for pool size
Myria Sarvay (1):
arm64: Add CPU part numbers for newer Apple Silicon CPUs
Xuheng Li (2):
KVM: arm64: Don't replace a valid hyp block mapping with a table
KVM: arm64: Order a later pinner's accesses after the hyp mapping
arch/arm64/include/asm/cputype.h | 28 +++++++++++++++++++
arch/arm64/include/asm/kvm_host.h | 2 +-
arch/arm64/include/asm/kvm_pgtable.h | 2 +-
arch/arm64/kernel/cpu_errata.c | 12 +++++++++
arch/arm64/kernel/cpufeature.c | 12 +++++++++
arch/arm64/kvm/arm.c | 8 ++++--
arch/arm64/kvm/config.c | 7 +----
arch/arm64/kvm/emulate-nested.c | 10 ++++++-
arch/arm64/kvm/hyp/include/nvhe/early_alloc.h | 2 +-
arch/arm64/kvm/hyp/include/nvhe/gfp.h | 4 +--
arch/arm64/kvm/hyp/nvhe/early_alloc.c | 2 +-
arch/arm64/kvm/hyp/nvhe/mem_protect.c | 8 ++++++
arch/arm64/kvm/hyp/nvhe/page_alloc.c | 6 ++---
arch/arm64/kvm/hyp/nvhe/sys_regs.c | 4 +--
arch/arm64/kvm/hyp/pgtable.c | 10 ++++++-
arch/arm64/kvm/nested.c | 8 ++++++
arch/arm64/kvm/pmu-emul.c | 3 +++
arch/arm64/kvm/sys_regs.c | 39 ++++++++++++++++++++++++---
arch/arm64/kvm/vgic/vgic-mmio.c | 4 +--
arch/arm64/kvm/vgic/vgic-mmio.h | 2 +-
tools/testing/selftests/kvm/arm64/hello_el2.c | 15 +++++++++++
21 files changed, 161 insertions(+), 27 deletions(-)
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [GIT PULL] KVM/arm64 fixes for 7.4, take #3
2026-11-19 12:56 [GIT PULL] KVM/arm64 fixes for 7.4, take #3 Marc Zyngier
@ 2026-11-19 17:02 ` Paolo Bonzini
0 siblings, 0 replies; 2+ messages in thread
From: Paolo Bonzini @ 2026-11-19 17:02 UTC (permalink / raw)
To: Marc Zyngier
Cc: Catalin Marinas, Colton Lewis, Dan Carpenter, Fuad Tabba,
Michael Reeves, Mostafa Saleh, Myria Sarvay, Oliver Upton,
Vincent Donnefort, Will Deacon, Xuheng Li, Yureka Lilian,
Steffen Eiden, Joey Gouly, Suzuki K Poulose, Zenghui Yu, kvmarm,
linux-arm-kernel, kvm
On Thu, Nov 19, 2026 at 1:56 PM Marc Zyngier <maz@kernel.org> wrote:
>
> Paolo,
>
> As promissed in my previous missive, I'm back with a few more
> presents...
>
> This round of fixes is mostly a Dan Carpenter special, as they seem to
> have hit the jackpot with some new formal analysis tool, and found a
> number of issues in NV. I am very surprised (/s).
>
> The rest is a bunch of pKVM fixes and workarounds for ugly hardware.
Pulled, thanks.
> I will be back next week same time for more of the same!
I can't wait for it!
Paolo
> Please pull,
>
> M.
>
> The following changes since commit 181a701fc14d3d468c603b5ecfb9f12115c5fb72:
>
> Merge branch kvm-arm64/misc-7.4-1 into kvmarm-master/fixes (2026-11-03 18:14:37 +0000)
>
> are available in the Git repository at:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm.git tags/kvmarm-fixes-7.4-3
>
> for you to fetch changes up to 3385a832525527801012f5993d1f6a7a3d23092c:
>
> Merge branch kvm-arm64/mmu-fixes-7.4-2 into kvmarm-master/fixes (2026-11-19 11:16:29 +0000)
>
> ----------------------------------------------------------------
> KVM/arm64 fixes for 7.4, take #3
>
> * MMU fixes:
>
> - Fix arbitrary truncation to 32bit when dealing with pool size
>
> - Correctly order page pin refconting
>
> - Prevent a HYP block mapping from being replaced with a table
>
> - Fix NV Range invalidation by extending the boundaries to the
> next block mapping
>
> - Make sure the correct hw_mmu gets picked on the first run,
> should userspace restore a nested context between init and
> the first vcpu_load().
>
> - Reject TLBI IPAS2E1 that fall outside of the S2 range
>
> * System register fixes:
>
> - Correctly forward POE/PIE/TVAL EL2 register traps
>
> - Add a sanity check for missing trap forwarding
>
> - Correctly hide generic PAuth for pKVM protected guests when
> the feature PAuth has been exluced from the feature set
>
> - Make sure that HCR_EL2.NV1 is always RES0
>
> * Misc fixes:
>
> - Flag Apple M3, M4 and M6 CPUs as having a broken vgic and
> a wonky PMUv3 emulation.
>
> - Correctly handle 64bit offsets into an MMIO region from
> userspace
>
> - Ensure that a vcpu is loaded when injecting an event
>
> ----------------------------------------------------------------
> Colton Lewis (2):
> KVM: arm64: Force HCR_EL2.NV1 to RES0 unconditionally
> KVM: arm64: selftests: Verify HCR_EL2.NV1 is RES0 when E2H0 is NI_NV1
>
> Dan Carpenter (9):
> KVM: arm64: nv: Invalidate whole L1 blocks on TLBI RIPAS2E1
> KVM: arm64: vgic: Don't truncate offsets in vgic_find_mmio_region()
> KVM: arm64: nv: Initialize hw_mmu to NULL for nested vCPUs
> KVM: arm64: nv: Forward PIR_EL2, PIRE0_EL2 and POR_EL2 traps
> KVM: arm64: nv: Forward CNTP_TVAL_EL02 and CNTV_TVAL_EL02 traps
> KVM: arm64: nv: Report EL2 registers with no trap forwarding
> KVM: arm64: Load the vcpu in KVM_SET_VCPU_EVENTS
> KVM: arm64: nv: Fix NULL dereference in kvm_pmu_nested_transition()
> KVM: arm64: nv: Ignore TLBI IPAS2E1 above the IPA limit
>
> Fuad Tabba (1):
> KVM: arm64: Hide generic ptrauth from protected VMs without ptrauth
>
> Marc Zyngier (3):
> Merge branch kvm-arm64/misc-7.4-2 into kvmarm-master/fixes
> Merge branch kvm-arm64/sysreg-fixes-7.4-2 into kvmarm-master/fixes
> Merge branch kvm-arm64/mmu-fixes-7.4-2 into kvmarm-master/fixes
>
> Michael Reeves (1):
> arm64: Apply existing Apple Silicon CPU workarounds to newer CPUs
>
> Mostafa Saleh (1):
> KVM: arm64: Use consistent type for pool size
>
> Myria Sarvay (1):
> arm64: Add CPU part numbers for newer Apple Silicon CPUs
>
> Xuheng Li (2):
> KVM: arm64: Don't replace a valid hyp block mapping with a table
> KVM: arm64: Order a later pinner's accesses after the hyp mapping
>
> arch/arm64/include/asm/cputype.h | 28 +++++++++++++++++++
> arch/arm64/include/asm/kvm_host.h | 2 +-
> arch/arm64/include/asm/kvm_pgtable.h | 2 +-
> arch/arm64/kernel/cpu_errata.c | 12 +++++++++
> arch/arm64/kernel/cpufeature.c | 12 +++++++++
> arch/arm64/kvm/arm.c | 8 ++++--
> arch/arm64/kvm/config.c | 7 +----
> arch/arm64/kvm/emulate-nested.c | 10 ++++++-
> arch/arm64/kvm/hyp/include/nvhe/early_alloc.h | 2 +-
> arch/arm64/kvm/hyp/include/nvhe/gfp.h | 4 +--
> arch/arm64/kvm/hyp/nvhe/early_alloc.c | 2 +-
> arch/arm64/kvm/hyp/nvhe/mem_protect.c | 8 ++++++
> arch/arm64/kvm/hyp/nvhe/page_alloc.c | 6 ++---
> arch/arm64/kvm/hyp/nvhe/sys_regs.c | 4 +--
> arch/arm64/kvm/hyp/pgtable.c | 10 ++++++-
> arch/arm64/kvm/nested.c | 8 ++++++
> arch/arm64/kvm/pmu-emul.c | 3 +++
> arch/arm64/kvm/sys_regs.c | 39 ++++++++++++++++++++++++---
> arch/arm64/kvm/vgic/vgic-mmio.c | 4 +--
> arch/arm64/kvm/vgic/vgic-mmio.h | 2 +-
> tools/testing/selftests/kvm/arm64/hello_el2.c | 15 +++++++++++
> 21 files changed, 161 insertions(+), 27 deletions(-)
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2026-11-19 17:02 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-19 12:56 [GIT PULL] KVM/arm64 fixes for 7.4, take #3 Marc Zyngier
2026-11-19 17:02 ` Paolo Bonzini
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox