KVM: GUEST_MEMFD fixes for 6.8
Sean Christopherson
reviewerMinor fixes related GUEST_MEMFD. I _just_ posted these, and they've only been in -next for one night, but I am sending this now to ensure you see it asap, as patch 1 in particular affects KVM's ABI, i.e. really should land in 6.8 before GUEST_MEMFD support is officially released.
The following changes since commit c48617fbbe831d4c80fe84056033f17b70a31136:
Merge tag 'kvmarm-fixes-6.8-3' of git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarm into HEAD (2024-02-21 05:18:56 -0500)
are available in the Git repository at:
https://github.com/kvm-x86/linux.git tags/kvm-x86-guest_memfd_fixes-6.8
for you to fetch changes up to 2dfd2383034421101300a3b7325cf339a182d218:
KVM: selftests: Add a testcase to verify GUEST_MEMFD and READONLY are exclusive (2024-02-22 17:07:06 -0800)
----------------------------------------------------------------
KVM GUEST_MEMFD fixes for 6.8:
- Make KVM_MEM_GUEST_MEMFD mutually exclusive with KVM_MEM_READONLY to avoid creating ABI that KVM can't sanely support.
- Update documentation for KVM_SW_PROTECTED_VM to make it abundantly clear that such VMs are purely a development and testing vehicle, and come with zero guarantees.
- Limit KVM_SW_PROTECTED_VM guests to the TDP MMU, as the long term plan is to support confidential VMs with deterministic private memory (SNP and TDX) only in the TDP MMU.
- Fix a bug in a GUEST_MEMFD negative test that resulted in false passes when verifying that KVM_MEM_GUEST_MEMFD memslots can't be dirty logged.
----------------------------------------------------------------
Sean Christopherson (5):
KVM: Make KVM_MEM_GUEST_MEMFD mutually exclusive with KVM_MEM_READONLY
KVM: x86: Update KVM_SW_PROTECTED_VM docs to make it clear they're a WIP
KVM: x86/mmu: Restrict KVM_SW_PROTECTED_VM to the TDP MMU
KVM: selftests: Create GUEST_MEMFD for relevant invalid flags testcases
KVM: selftests: Add a testcase to verify GUEST_MEMFD and READONLY are exclusive
Documentation/virt/kvm/api.rst | 5 +++++ arch/x86/kvm/Kconfig | 7 ++++--- arch/x86/kvm/x86.c | 2 +- tools/testing/selftests/kvm/set_memory_region_test.c | 12 +++++++++++- virt/kvm/kvm_main.c | 8 +++++++- 5 files changed, 28 insertions(+), 6 deletions(-)
Sean Christopherson
reviewerQuoted 31 lines · Sean Christopherson
Minor fixes related
GUEST_MEMFD. I _just_ posted these, and they've only been in -next for one night, but I am sending this now to ensure you see it asap, as patch 1 in particular affects KVM's ABI, i.e. really should land in 6.8 beforeGUEST_MEMFDsupport is officially released.The following changes since commit
c48617fbbe831d4c80fe84056033f17b70a31136:Merge tag 'kvmarm-fixes-6.8-3' of
git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarminto HEAD (2024-02-21 05:18:56 -0500)are available in the Git repository at:
https://github.com/kvm-x86/linux.git tags/kvm-x86-guest_memfd_fixes-6.8
for you to fetch changes up to
2dfd2383034421101300a3b7325cf339a182d218:KVM: selftests: Add a testcase to verify
GUEST_MEMFDand READONLY are exclusive (2024-02-22 17:07:06 -0800)----------------------------------------------------------------
KVMGUEST_MEMFDfixes for 6.8:- Make
KVM_MEM_GUEST_MEMFDmutually exclusive withKVM_MEM_READONLYto avoid creating ABI that KVM can't sanely support.- Update documentation for
KVM_SW_PROTECTED_VMto make it abundantly clear that such VMs are purely a development and testing vehicle, and come with zero guarantees.- Limit
KVM_SW_PROTECTED_VMguests to the TDP MMU, as the long term plan is to support confidential VMs with deterministic private memory (SNP and TDX) only in the TDP MMU.
Quoted 6 lines
- Fix a bug in a
GUEST_MEMFDnegative test that resulted in false passes when verifying thatKVM_MEM_GUEST_MEMFDmemslots can't be dirty logged.----------------------------------------------------------------
Sean Christopherson (5):
KVM: MakeKVM_MEM_GUEST_MEMFDmutually exclusive withKVM_MEM_READONLY
Almost forgot, just as an FYI, this has a minor conflict with your kvm/kvm-uapi branch. I've been fixing it up in kvm-x86/next, and IIUC you don't feed kvm/master into -next, so I don't think Stephen will see a conflict?
Quoted 11 lines
KVM: x86: Update
KVM_SW_PROTECTED_VMdocs to make it clear they're a WIP
KVM: x86/mmu: RestrictKVM_SW_PROTECTED_VMto the TDP MMU
KVM: selftests: CreateGUEST_MEMFDfor relevant invalid flags testcases
KVM: selftests: Add a testcase to verifyGUEST_MEMFDand READONLY are exclusiveDocumentation/virt/kvm/api.rst | 5 +++++ arch/x86/kvm/Kconfig | 7 ++++--- arch/x86/kvm/x86.c | 2 +- tools/testing/selftests/kvm/set_memory_region_test.c | 12 +++++++++++- virt/kvm/kvm_main.c | 8 +++++++- 5 files changed, 28 insertions(+), 6 deletions(-)
Paolo Bonzini
maintainerQuoted 36 lines · Sean Christopherson
Quoted 34 lines · Sean Christopherson
Minor fixes related
GUEST_MEMFD. I _just_ posted these, and they've only been in -next for one night, but I am sending this now to ensure you see it asap, as patch 1 in particular affects KVM's ABI, i.e. really should land in 6.8 beforeGUEST_MEMFDsupport is officially released.The following changes since commit
c48617fbbe831d4c80fe84056033f17b70a31136:Merge tag 'kvmarm-fixes-6.8-3' of
git://git.kernel.org/pub/scm/linux/kernel/git/kvmarm/kvmarminto HEAD (2024-02-21 05:18:56 -0500)are available in the Git repository at:
https://github.com/kvm-x86/linux.git tags/kvm-x86-guest_memfd_fixes-6.8
for you to fetch changes up to
2dfd2383034421101300a3b7325cf339a182d218:KVM: selftests: Add a testcase to verify
GUEST_MEMFDand READONLY are exclusive (2024-02-22 17:07:06 -0800)----------------------------------------------------------------
KVMGUEST_MEMFDfixes for 6.8:- Make
KVM_MEM_GUEST_MEMFDmutually exclusive withKVM_MEM_READONLYto avoid creating ABI that KVM can't sanely support.- Update documentation for
KVM_SW_PROTECTED_VMto make it abundantly clear that such VMs are purely a development and testing vehicle, and come with zero guarantees.- Limit
KVM_SW_PROTECTED_VMguests to the TDP MMU, as the long term plan is to support confidential VMs with deterministic private memory (SNP and TDX) only in the TDP MMU.- Fix a bug in a
GUEST_MEMFDnegative test that resulted in false passes when verifying thatKVM_MEM_GUEST_MEMFDmemslots can't be dirty logged.
Quoted 7 lines
Quoted 3 lines
----------------------------------------------------------------
Sean Christopherson (5):
KVM: MakeKVM_MEM_GUEST_MEMFDmutually exclusive withKVM_MEM_READONLYAlmost forgot, just as an FYI, this has a minor conflict with your kvm/kvm-uapi branch. I've been fixing it up in kvm-x86/next, and IIUC you don't feed kvm/master into -next, so I don't think Stephen will see a conflict?
I do feed it in linux-next, so he would, but it's not a big deal. I'll pull it into kvm-next as well.
Paolo