wireguard-linux/arch/x86, branch update-toolchain

KVM: x86: Disallow read-only memslots for SEV-ES and SEV-SNP (and TDX)

2024-08-14T16:28:24Z

Disallow read-only memslots for SEV-{ES,SNP} VM types, as KVM can't directly emulate instructions for ES/SNP, and instead the guest must explicitly request emulation. Unless the guest explicitly requests emulation without accessing memory, ES/SNP relies on KVM creating an MMIO SPTE, with the subsequent #NPF being reflected into the guest as a #VC. But for read-only memslots, KVM deliberately doesn't create MMIO SPTEs, because except for ES/SNP, doing so requires setting reserved bits in the SPTE, i.e. the SPTE can't be readable while also generating a #VC on writes. Because KVM never creates MMIO SPTEs and jumps directly to emulation, the guest never gets a #VC. And since KVM simply resumes the guest if ES/SNP guests trigger emulation, KVM effectively puts the vCPU into an infinite #NPF loop if the vCPU attempts to write read-only memory. Disallow read-only memory for all VMs with protected state, i.e. for upcoming TDX VMs as well as ES/SNP VMs. For TDX, it's actually possible to support read-only memory, as TDX uses EPT Violation #VE to reflect the fault into the guest, e.g. KVM could configure read-only SPTEs with RX protections and SUPPRESS_VE=0. But there is no strong use case for supporting read-only memslots on TDX, e.g. the main historical usage is to emulate option ROMs, but TDX disallows executing from shared memory. And if someone comes along with a legitimate, strong use case, the restriction can always be lifted for TDX. Don't bother trying to retroactively apply the restriction to SEV-ES VMs that are created as type KVM_X86_DEFAULT_VM. Read-only memslots can't possibly work for SEV-ES, i.e. disallowing such memslots is really just means reporting an error to userspace instead of silently hanging vCPUs. Trying to deal with the ordering between KVM_SEV_INIT and memslot creation isn't worth the marginal benefit it would provide userspace. Fixes: 26c44aa9e076 ("KVM: SEV: define VM types for SEV and SEV-ES") Fixes: 1dfe571c12cf ("KVM: SEV: Add initial SEV-SNP support") Cc: Peter Gonda Cc: Michael Roth Cc: Vishal Annapurve Cc: Ackerly Tng Signed-off-by: Sean Christopherson Message-ID: <20240809190319.1710470-2-seanjc@google.com> Signed-off-by: Paolo Bonzini

KVM: x86: Make x2APIC ID 100% readonly

2024-08-13T16:01:46Z

Ignore the userspace provided x2APIC ID when fixing up APIC state for KVM_SET_LAPIC, i.e. make the x2APIC fully readonly in KVM. Commit a92e2543d6a8 ("KVM: x86: use hardware-compatible format for APIC ID register"), which added the fixup, didn't intend to allow userspace to modify the x2APIC ID. In fact, that commit is when KVM first started treating the x2APIC ID as readonly, apparently to fix some race: static inline u32 kvm_apic_id(struct kvm_lapic *apic) { - return (kvm_lapic_get_reg(apic, APIC_ID) >> 24) & 0xff; + /* To avoid a race between apic_base and following APIC_ID update when + * switching to x2apic_mode, the x2apic mode returns initial x2apic id. + */ + if (apic_x2apic_mode(apic)) + return apic->vcpu->vcpu_id; + + return kvm_lapic_get_reg(apic, APIC_ID) >> 24; } Furthermore, KVM doesn't support delivering interrupts to vCPUs with a modified x2APIC ID, but KVM *does* return the modified value on a guest RDMSR and for KVM_GET_LAPIC. I.e. no remotely sane setup can actually work with a modified x2APIC ID. Making the x2APIC ID fully readonly fixes a WARN in KVM's optimized map calculation, which expects the LDR to align with the x2APIC ID. WARNING: CPU: 2 PID: 958 at arch/x86/kvm/lapic.c:331 kvm_recalculate_apic_map+0x609/0xa00 [kvm] CPU: 2 PID: 958 Comm: recalc_apic_map Not tainted 6.4.0-rc3-vanilla+ #35 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Arch Linux 1.16.2-1-1 04/01/2014 RIP: 0010:kvm_recalculate_apic_map+0x609/0xa00 [kvm] Call Trace: kvm_apic_set_state+0x1cf/0x5b0 [kvm] kvm_arch_vcpu_ioctl+0x1806/0x2100 [kvm] kvm_vcpu_ioctl+0x663/0x8a0 [kvm] __x64_sys_ioctl+0xb8/0xf0 do_syscall_64+0x56/0x80 entry_SYSCALL_64_after_hwframe+0x46/0xb0 RIP: 0033:0x7fade8b9dd6f Unfortunately, the WARN can still trigger for other CPUs than the current one by racing against KVM_SET_LAPIC, so remove it completely. Reported-by: Michal Luczaj Closes: https://lore.kernel.org/all/814baa0c-1eaa-4503-129f-059917365e80@rbox.co Reported-by: Haoyu Wu Closes: https://lore.kernel.org/all/20240126161633.62529-1-haoyuwu254@gmail.com Reported-by: syzbot+545f1326f405db4e1c3e@syzkaller.appspotmail.com Closes: https://lore.kernel.org/all/000000000000c2a6b9061cbca3c3@google.com Signed-off-by: Sean Christopherson Message-ID: <20240802202941.344889-2-seanjc@google.com> Signed-off-by: Paolo Bonzini

KVM: x86: Use this_cpu_ptr() instead of per_cpu_ptr(smp_processor_id())

2024-08-13T14:24:37Z

Use this_cpu_ptr() instead of open coding the equivalent in various user return MSR helpers. Signed-off-by: Isaku Yamahata Reviewed-by: Chao Gao Reviewed-by: Yuan Yao [sean: massage changelog] Signed-off-by: Sean Christopherson Reviewed-by: Pankaj Gupta Message-ID: <20240802201630.339306-1-seanjc@google.com> Signed-off-by: Paolo Bonzini

KVM: x86: hyper-v: Remove unused inline function kvm_hv_free_pa_page()

2024-08-13T13:28:48Z

There is no caller in tree since introduction in commit b4f69df0f65e ("KVM: x86: Make Hyper-V emulation optional") Signed-off-by: Yue Haibing Message-ID: <20240803113233.128185-1-yuehaibing@huawei.com> Signed-off-by: Paolo Bonzini

KVM: SVM: Fix an error code in sev_gmem_post_populate()

2024-08-13T10:08:40Z

The copy_from_user() function returns the number of bytes which it was not able to copy. Return -EFAULT instead. Fixes: dee5a47cc7a4 ("KVM: SEV: Add KVM_SEV_SNP_LAUNCH_UPDATE command") Signed-off-by: Dan Carpenter Message-ID: <20240612115040.2423290-4-dan.carpenter@linaro.org> Signed-off-by: Paolo Bonzini

KVM: SVM: Fix uninitialized variable bug

2024-08-13T10:05:10Z

If snp_lookup_rmpentry() fails then "assigned" is printed in the error message but it was never initialized. Initialize it to false. Fixes: dee5a47cc7a4 ("KVM: SEV: Add KVM_SEV_SNP_LAUNCH_UPDATE command") Signed-off-by: Dan Carpenter Message-ID: <20240612115040.2423290-3-dan.carpenter@linaro.org> Signed-off-by: Paolo Bonzini

x86/mtrr: Check if fixed MTRRs exist before saving them

2024-08-08T15:03:12Z

MTRRs have an obsolete fixed variant for fine grained caching control of the 640K-1MB region that uses separate MSRs. This fixed variant has a separate capability bit in the MTRR capability MSR. So far all x86 CPUs which support MTRR have this separate bit set, so it went unnoticed that mtrr_save_state() does not check the capability bit before accessing the fixed MTRR MSRs. Though on a CPU that does not support the fixed MTRR capability this results in a #GP. The #GP itself is harmless because the RDMSR fault is handled gracefully, but results in a WARN_ON(). Add the missing capability check to prevent this. Fixes: 2b1f6278d77c ("[PATCH] x86: Save the MTRRs of the BSP before booting an AP") Signed-off-by: Andi Kleen Signed-off-by: Thomas Gleixner Cc: stable@vger.kernel.org Link: https://lore.kernel.org/all/20240808000244.946864-1-ak@linux.intel.com

x86/paravirt: Fix incorrect virt spinlock setting on bare metal

2024-08-07T18:04:38Z

The kernel can change spinlock behavior when running as a guest. But this guest-friendly behavior causes performance problems on bare metal. The kernel uses a static key to switch between the two modes. In theory, the static key is enabled by default (run in guest mode) and should be disabled for bare metal (and in some guests that want native behavior or paravirt spinlock). A performance drop is reported when running encode/decode workload and BenchSEE cache sub-workload. Bisect points to commit ce0a1b608bfc ("x86/paravirt: Silence unused native_pv_lock_init() function warning"). When CONFIG_PARAVIRT_SPINLOCKS is disabled the virt_spin_lock_key is incorrectly set to true on bare metal. The qspinlock degenerates to test-and-set spinlock, which decreases the performance on bare metal. Set the default value of virt_spin_lock_key to false. If booting in a VM, enable this key. Later during the VM initialization, if other high-efficient spinlock is preferred (e.g. paravirt-spinlock), or the user wants the native qspinlock (via nopvspin boot commandline), the virt_spin_lock_key is disabled accordingly. This results in the following decision matrix: X86_FEATURE_HYPERVISOR Y Y Y N CONFIG_PARAVIRT_SPINLOCKS Y Y N Y/N PV spinlock Y N N Y/N virt_spin_lock_key N Y/N Y N Fixes: ce0a1b608bfc ("x86/paravirt: Silence unused native_pv_lock_init() function warning") Reported-by: Prem Nath Dey Reported-by: Xiaoping Zhou Suggested-by: Dave Hansen Suggested-by: Qiuxu Zhuo Suggested-by: Nikolay Borisov Signed-off-by: Chen Yu Signed-off-by: Thomas Gleixner Reviewed-by: Nikolay Borisov Cc: stable@vger.kernel.org Link: https://lore.kernel.org/all/20240806112207.29792-1-yu.c.chen@intel.com

x86/acpi: Remove __ro_after_init from acpi_mp_wake_mailbox

2024-08-07T18:04:38Z

On a platform using the "Multiprocessor Wakeup Structure"[1] to startup secondary CPUs the control processor needs to memremap() the physical address of the MP Wakeup Structure mailbox to the variable acpi_mp_wake_mailbox, which holds the virtual address of mailbox. To wake up the AP the control processor writes the APIC ID of AP, the wakeup vector and the ACPI_MP_WAKE_COMMAND_WAKEUP command into the mailbox. Current implementation doesn't consider the case which restricts boot time CPU bringup to 1 with the kernel parameter "maxcpus=1" and brings other CPUs online later from user space as it sets acpi_mp_wake_mailbox to read-only after init. So when the first AP is tried to brought online after init, the attempt to update the variable results in a kernel panic. The memremap() call that initializes the variable cannot be moved into acpi_parse_mp_wake() because memremap() is not functional at that point in the boot process. Also as the APs might never be brought up, keep the memremap() call in acpi_wakeup_cpu() so that the operation only takes place when needed. Fixes: 24dd05da8c79 ("x86/apic: Mark acpi_mp_wake_* variables as __ro_after_init") Signed-off-by: Zhiquan Li Signed-off-by: Thomas Gleixner Reviewed-by: Kirill A. Shutemov Link: https://lore.kernel.org/all/20240805103531.1230635-1-zhiquan1.li@intel.com

x86/mm: Fix PTI for i386 some more

2024-08-07T13:35:01Z

So it turns out that we have to do two passes of pti_clone_entry_text(), once before initcalls, such that device and late initcalls can use user-mode-helper / modprobe and once after free_initmem() / mark_readonly(). Now obviously mark_readonly() can cause PMD splits, and pti_clone_pgtable() doesn't like that much. Allow the late clone to split PMDs so that pagetables stay in sync. [peterz: Changelog and comments] Reported-by: Guenter Roeck Signed-off-by: Thomas Gleixner Signed-off-by: Peter Zijlstra (Intel) Tested-by: Guenter Roeck Link: https://lkml.kernel.org/r/20240806184843.GX37996@noisy.programming.kicks-ass.net