mbox series

[v2,0/4] KVM: arm64: Fix underallocation of storage for SVE state

Message ID 20240606-kvm-arm64-fix-pkvm-sve-vl-v2-0-c88f4eb4b14b@kernel.org (mailing list archive)
Headers show
Series KVM: arm64: Fix underallocation of storage for SVE state | expand

Message

Mark Brown June 6, 2024, 3:21 p.m. UTC
As observed during review the pKVM support for saving host SVE state is
broken if an asymmetric system has VLs larger than the maximum shared
VL, fix this by discovering then using the maximum VL for allocations
and using RDVL during the save/restore process.

Signed-off-by: Mark Brown <broonie@kernel.org>
---
Changes in v2:
- Downgrade check for a late CPU increasing maximum VL to a warning only
  but do it unconditionally since pKVM prevents late CPUs anyway.
- Commit log tweaks.
- Link to v1: https://lore.kernel.org/r/20240605-kvm-arm64-fix-pkvm-sve-vl-v1-0-680d6b43b4c1@kernel.org

---
Mark Brown (4):
      arm64/fpsimd: Introduce __bit_to_vl() helper
      arm64/fpsimd: Discover maximum vector length implemented by any CPU
      KVM: arm64: Fix FFR offset calculation for pKVM host state save and restore
      KVM: arm64: Avoid underallocating storage for host SVE state

 arch/arm64/include/asm/fpsimd.h         | 17 +++++++++++++++
 arch/arm64/include/asm/kvm_host.h       |  2 +-
 arch/arm64/include/asm/kvm_hyp.h        |  3 ++-
 arch/arm64/include/asm/kvm_pkvm.h       |  2 +-
 arch/arm64/kernel/fpsimd.c              | 38 +++++++++++++++++++++++++++------
 arch/arm64/kvm/hyp/fpsimd.S             |  5 +++++
 arch/arm64/kvm/hyp/include/hyp/switch.h |  2 +-
 arch/arm64/kvm/hyp/nvhe/hyp-main.c      |  6 +++---
 arch/arm64/kvm/hyp/nvhe/pkvm.c          |  2 +-
 arch/arm64/kvm/reset.c                  |  6 +++---
 10 files changed, 65 insertions(+), 18 deletions(-)
---
base-commit: afb91f5f8ad7af172d993a34fde1947892408f53
change-id: 20240604-kvm-arm64-fix-pkvm-sve-vl-13cd71fd7db0

Best regards,

Comments

Catalin Marinas July 4, 2024, 2:20 p.m. UTC | #1
On Thu, Jun 06, 2024 at 04:21:42PM +0100, Mark Brown wrote:
> As observed during review the pKVM support for saving host SVE state is
> broken if an asymmetric system has VLs larger than the maximum shared
> VL, fix this by discovering then using the maximum VL for allocations
> and using RDVL during the save/restore process.
> 
> Signed-off-by: Mark Brown <broonie@kernel.org>
> ---
> Changes in v2:
> - Downgrade check for a late CPU increasing maximum VL to a warning only
>   but do it unconditionally since pKVM prevents late CPUs anyway.
> - Commit log tweaks.
> - Link to v1: https://lore.kernel.org/r/20240605-kvm-arm64-fix-pkvm-sve-vl-v1-0-680d6b43b4c1@kernel.org
> 
> ---
> Mark Brown (4):
>       arm64/fpsimd: Introduce __bit_to_vl() helper
>       arm64/fpsimd: Discover maximum vector length implemented by any CPU
>       KVM: arm64: Fix FFR offset calculation for pKVM host state save and restore
>       KVM: arm64: Avoid underallocating storage for host SVE state

Since the patches address a KVM issue, I think it should go via
Oliver/Marc's tree. But happy to queue it via the arm64 tree if people
think otherwise.

Thanks.