From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-yx2-f13.google.com (mail-yx2-f13.google.com [74.125.224.141]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1F76E47207E for ; Sun, 20 Sep 2026 19:37:43 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=74.125.224.141 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789933066; cv=none; b=TmnPUfU1YP1zTQgc8xTEu/eq0EVUxLPWiQfuu5skD7Y5UBOFxC+SbKHqgl5Vu2j9+i77z4AS3r75TZ0MfLzidqb3b8hUMKzKsOz8uzYLxw2mYuruTRqnQYbjlsdH5Y7UE4NMqfYmdlNRm+7U4LtfetsCabB1KlG2tAjGzep5p7o= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789933066; c=relaxed/simple; bh=hErW7kyLcp3jWuujp7J2W5IYfLUdG0x+oJfbxhieGso=; h=From:To:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=L0KmBcm8P+HRqC4ONOvS/oQcmACmkzVPcvGGW/zAvImpTIcA61Tst8kc0BnDqgrvGa5K2EGorkYWMhT20i6TH3F+RgRqtj0usd9iEK+QhoYkV9m0vgLHA33xTfPihNXQJGqCVBM8DLQfNtz+1WXU5yjIc15TFXjo8B2fHHaFUVU= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=soleen.com; spf=pass smtp.mailfrom=soleen.com; dkim=pass (2048-bit key) header.d=soleen.com header.i=@soleen.com header.b=Gz/DtfA4; arc=none smtp.client-ip=74.125.224.141 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=soleen.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=soleen.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=soleen.com header.i=@soleen.com header.b="Gz/DtfA4" Received: by mail-yx2-f13.google.com with SMTP id 956f58d0204a3-66e4aac92dbso1961060d50.2 for ; Sun, 20 Sep 2026 12:37:43 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=soleen.com; s=google; t=1789933062; x=1790537862; darn=vger.kernel.org; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:to:from:from:to:cc:subject:date:message-id :reply-to:content-type; bh=LP9MAkgaZIZB/G//n4eIZm11yhyh4okaQRNwKZew5U0=; b=Gz/DtfA4MWPqRZSqsmTbvlzwHNsrog7u/PBBXoYbpWK8BEvTMzihlHbg23V7OagX35 yMvcPmwQ4IWuHvigulwofn9U/cvE+pXHKDtNxfXyOQQr0Yc+iqWRcuJE4Zf8yThNG+/4 KoiYfA1OvjwquqpxeaV7oOHWW03Gx0BJeA7dWl3Mih6V3KBeUsXaXSrpYSwv3M5BytAJ +LmwokhPaEAMeRR9PAw43SxWpSq2AIRkeyi5mjJfb/S77DGk6MeowIG+zI9qXHI4X05k l9jiaf8OIZuxtU9RRSN6Rzmu7VjJmSDG9XvGSMBLmSZBmERkBDR2k7F6V1Un9zyj6Vxf m34g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20260707; t=1789933062; x=1790537862; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:to:from:x-gm-gg:x-gm-message-state:from:to :cc:subject:date:message-id:reply-to:content-type; bh=LP9MAkgaZIZB/G//n4eIZm11yhyh4okaQRNwKZew5U0=; b=bLN403RE53Jd+2v72NumntZP8KF4cSrAkqNG17ASuQ01+SckgCoc45UWv71TPzgDap J5MBZ9cbHGF1FsLeJNV5Z8DudsCeIaKuotC1SXvKal9yD0Lx0eUrqOQW6pVzwwCnVCBz wHqUGRVqPNdxHKDyJBjk1v53cLuSEDONOtmiZGDs6e25zt/oRrIwcgz5a0EH4nxvTSDB jXINju2efsJyNL/J29o+awCa6Un2+jgviPMf1PtGnz5AQQ6d0hojO+iN3Tpc7OMeeQMt 4LjKCcj5jvvEgx7fQ1UZtZFy8DEOlfyEizcQP2c/gSvVmHov9WeeQdBv7ZrqxZ7oKeSS Q22g== X-Forwarded-Encrypted: i=1; AKwUvBz/15iNPmpKFz1BNlTkCWZRvFkdYKnZTo9qlYwk2OmQPd1YE0ZNlk/vrkCexWVCcevuGZrrTJURxbw=@vger.kernel.org X-Gm-Message-State: AFuF++n98go93zFPoZgxG+FohqvsHRWpl0tYnuqmkNFgOH4aNuztdOOZ 6KY5ic1WEvpFh5d6OqI/5WwnhMkw68ZyKVbAaizpXc3d5ImJtkrwIBQ3HyGRU//lPHQ= X-Gm-Gg: AYBFou2gDd3HcWV8ASMD/SmBvyKhRGX4RnQHCrhwp71EhqW0o5biyqAPmOnEk3qeuTp Y91zbwZPhIBViw5d+zIZqqXZCjyyVPubjyUP7T1LC7IlOr6EMulhx/5Gu200Howz3VnTosTomD8 aun78Z4VnYiyR/+odRfOQrJBeWuZwIkM5uEFQV6O3B9jx6vw3CYPsSHTMX7WHCiRPJaDe4ccEpt GU8OBmmtMU0GyGMbe8dV7biwUsOjLOJQQICFyGsAphnvmZ6q7LYSnurEJbrVbF8zo9D4Y+mcoKh +oX29nvyXA1IPTvkt0SipLTAFSeYIOmCnLRyGpNhQYw2kaL0YoXmud+tYZPmyxLhWc5fnssC/ZA uvWm23SHtpgr2ve8d2ykKHkE9kXRC8iXMVFON7B9pwa16HDE7Kjqsl4fBKuj/R9LlZBXkH+vmek OIDNbN0n81W4pffZ7ur8X/j6p7VJWrQGMVzxSk2+CQhYIlYo60FUf8qx8JD4PUEim/e6pRTv4Pb ibXI9Oss4MVPyNSq4Cr3aGrGEH+plYkU+5oZy26UIfzGorl66wFXPeb2clCy4CMRkwlWQqt+2ZI oyuWWaYkBn9a1KRQUMS+ X-Received: by 2002:a05:690e:128f:b0:668:8dc0:bcc3 with SMTP id 956f58d0204a3-6717fadb1fcmr2667214d50.0.1789933061983; Sun, 20 Sep 2026 12:37:41 -0700 (PDT) Received: from soleen.c.googlers.com.com (16.49.199.35.bc.googleusercontent.com. [35.199.49.16]) by smtp.gmail.com with ESMTPSA id 00721157ae682-89a489aef57sm18632047b3.43.2026.09.20.12.37.40 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Sun, 20 Sep 2026 12:37:41 -0700 (PDT) From: Pasha Tatashin To: linux-kselftest@vger.kernel.org, legion@kernel.org, kees@kernel.org, will@kernel.org, ruanjinjie@huawei.com, atomlin@atomlin.com, rppt@kernel.org, jani.nikula@intel.com, hamzamahfooz@linux.microsoft.com, joey.gouly@arm.com, tglx@kernel.org, nsc@kernel.org, alexandre.chartre@oracle.com, james.morse@arm.com, dianders@chromium.org, bp@alien8.de, jpoimboe@kernel.org, shuah@kernel.org, catalin.marinas@arm.com, linux-kbuild@vger.kernel.org, linux-arch@vger.kernel.org, kvmarm@lists.linux.dev, jaredwhite@microsoft.com, johan@kernel.org, pbonzini@redhat.com, mingo@redhat.com, linux-mm@kvack.org, seanjc@google.com, mark.rutland@arm.com, vdonnefort@google.com, tarunsahu@google.com, gshan@redhat.com, skhan@linuxfoundation.org, linux-doc@vger.kernel.org, xur@google.com, djbw@kernel.org, oupton@kernel.org, nogikh@google.com, sumitg@nvidia.com, linux-kernel@vger.kernel.org, zengheng4@huawei.com, peterz@infradead.org, corbet@lwn.net, suzuki.poulose@arm.com, luto@kernel.org, hpa@zytor.com, zhangpengjie2@huawei.com, x86@kernel.org, yuzenghui@huawei.com, jic23@kernel.org, ardb@kernel.org, pasha.tatashin@soleen.com, petr.pavlu@suse.com, ryan.roberts@arm.com, kexec@lists.infradead.org, pratyush@kernel.org, dave.hansen@linux.intel.com, rdunlap@infradead.org, kvm@vger.kernel.org, fuad.tabba@linux.dev, maz@kernel.org, mbenes@suse.cz, jgross@suse.com, seiden@linux.ibm.com, pierre.gondois@arm.com, song@kernel.org, nathan@kernel.org, pmladek@suse.com, graf@amazon.com, chao.gao@intel.com, zhenglifeng1@huawei.com, arnd@arndb.de, sidnayyar@google.com, linux-arm-kernel@lists.infradead.org, vladimir.murzin@arm.com, kas@kernel.org Subject: [RFC PATCH 08/46] KVM: x86: Add x86 vCPU LUO preservation ABI and register helpers Date: Sun, 20 Sep 2026 15:36:12 -0400 Message-ID: <20260920193650.3373435-9-pasha.tatashin@soleen.com> X-Mailer: git-send-email 2.55.0.1082.g2b9226bbc0-goog In-Reply-To: <20260920193650.3373435-1-pasha.tatashin@soleen.com> References: <20260920193650.3373435-1-pasha.tatashin@soleen.com> Precedence: bulk X-Mailing-List: linux-doc@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Add struct kvm_vcpu_arch_ser in include/linux/kho/abi/kvm_x86.h and export x86 register and MSR access helpers for LUO state save/restore. Signed-off-by: Pasha Tatashin --- arch/x86/kvm/cpuid.c | 4 +- arch/x86/kvm/cpuid.h | 1 + arch/x86/kvm/msrs.c | 15 ++++++ arch/x86/kvm/msrs.h | 2 + arch/x86/kvm/regs.c | 8 ++-- arch/x86/kvm/regs.h | 4 ++ arch/x86/kvm/x86.c | 8 ++-- arch/x86/kvm/x86.h | 4 ++ include/linux/kho/abi/kvm_x86.h | 84 +++++++++++++++++++++++++++++++++ 9 files changed, 120 insertions(+), 10 deletions(-) create mode 100644 include/linux/kho/abi/kvm_x86.h diff --git a/arch/x86/kvm/cpuid.c b/arch/x86/kvm/cpuid.c index 851f151efb35..5aba16e0610b 100644 --- a/arch/x86/kvm/cpuid.c +++ b/arch/x86/kvm/cpuid.c @@ -507,8 +507,8 @@ u64 kvm_vcpu_reserved_gpa_bits_raw(struct kvm_vcpu *vcpu) return rsvd_bits(cpuid_maxphyaddr(vcpu), 63); } -static int kvm_set_cpuid(struct kvm_vcpu *vcpu, struct kvm_cpuid_entry2 *e2, - int nent) +int kvm_set_cpuid(struct kvm_vcpu *vcpu, struct kvm_cpuid_entry2 *e2, + int nent) { u32 vcpu_caps[NR_KVM_CPU_CAPS]; int r; diff --git a/arch/x86/kvm/cpuid.h b/arch/x86/kvm/cpuid.h index 8d863f45585d..e5ec112d39cc 100644 --- a/arch/x86/kvm/cpuid.h +++ b/arch/x86/kvm/cpuid.h @@ -21,6 +21,7 @@ static inline void kvm_finalize_cpu_caps(void) } void kvm_vcpu_after_set_cpuid(struct kvm_vcpu *vcpu); +int kvm_set_cpuid(struct kvm_vcpu *vcpu, struct kvm_cpuid_entry2 *e2, int nent); struct kvm_cpuid_entry2 *kvm_find_cpuid_entry2(struct kvm_cpuid_entry2 *entries, int nent, u32 function, u64 index); /* diff --git a/arch/x86/kvm/msrs.c b/arch/x86/kvm/msrs.c index dd3bb04878ca..ae849e0b7d14 100644 --- a/arch/x86/kvm/msrs.c +++ b/arch/x86/kvm/msrs.c @@ -339,6 +339,21 @@ static u32 msr_based_features[ARRAY_SIZE(msr_based_features_all_except_vmx) + (KVM_LAST_EMULATED_VMX_MSR - KVM_FIRST_EMULATED_VMX_MSR + 1)]; static unsigned int num_msr_based_features; +unsigned int kvm_num_msrs_to_save(void) +{ + return num_msrs_to_save + num_emulated_msrs; +} + +u32 kvm_get_msr_to_save_index(unsigned int i) +{ + if (i < num_msrs_to_save) + return msrs_to_save[i]; + i -= num_msrs_to_save; + if (i < num_emulated_msrs) + return emulated_msrs[i]; + return 0; +} + int kvm_get_msr_index_list(struct kvm_msr_list __user *user_msr_list) { struct kvm_msr_list msr_list; diff --git a/arch/x86/kvm/msrs.h b/arch/x86/kvm/msrs.h index 7cc182a15b3b..3a5e39041d99 100644 --- a/arch/x86/kvm/msrs.h +++ b/arch/x86/kvm/msrs.h @@ -65,6 +65,8 @@ int __kvm_emulate_msr_read(struct kvm_vcpu *vcpu, u32 index, u64 *data); int __kvm_emulate_msr_write(struct kvm_vcpu *vcpu, u32 index, u64 data); int kvm_msr_read(struct kvm_vcpu *vcpu, u32 index, u64 *data); int kvm_msr_write(struct kvm_vcpu *vcpu, u32 index, u64 data); +unsigned int kvm_num_msrs_to_save(void); +u32 kvm_get_msr_to_save_index(unsigned int i); int kvm_emulate_rdmsr(struct kvm_vcpu *vcpu); int kvm_emulate_rdmsr_imm(struct kvm_vcpu *vcpu, u32 msr, int reg); int kvm_emulate_wrmsr(struct kvm_vcpu *vcpu); diff --git a/arch/x86/kvm/regs.c b/arch/x86/kvm/regs.c index 8f66438989e4..02f041c355cc 100644 --- a/arch/x86/kvm/regs.c +++ b/arch/x86/kvm/regs.c @@ -51,7 +51,7 @@ void kvm_set_rflags(struct kvm_vcpu *vcpu, unsigned long rflags) } EXPORT_SYMBOL_FOR_KVM_INTERNAL(kvm_set_rflags); -static void __get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +void __get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) { if (vcpu->arch.emulate_regs_need_sync_to_vcpu) { /* @@ -99,7 +99,7 @@ int kvm_arch_vcpu_ioctl_get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) return 0; } -static void __set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) +void __set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs) { vcpu->arch.emulate_regs_need_sync_from_vcpu = true; vcpu->arch.emulate_regs_need_sync_to_vcpu = false; @@ -498,7 +498,7 @@ static void __get_sregs_common(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs) sregs->apic_base = vcpu->arch.apic_base; } -static void __get_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs) +void __get_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs) { __get_sregs_common(vcpu, sregs); @@ -635,7 +635,7 @@ static int __set_sregs_common(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs, return 0; } -static int __set_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs) +int __set_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs) { int pending_vec, max_bits; int mmu_reset_needed = 0; diff --git a/arch/x86/kvm/regs.h b/arch/x86/kvm/regs.h index 447f0ec3e63e..40947b1be404 100644 --- a/arch/x86/kvm/regs.h +++ b/arch/x86/kvm/regs.h @@ -517,5 +517,9 @@ int kvm_vcpu_ioctl_x86_get_debugregs(struct kvm_vcpu *vcpu, int kvm_vcpu_ioctl_x86_set_debugregs(struct kvm_vcpu *vcpu, struct kvm_debugregs *dbgregs); +void __get_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs); +void __set_regs(struct kvm_vcpu *vcpu, struct kvm_regs *regs); +void __get_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs); +int __set_sregs(struct kvm_vcpu *vcpu, struct kvm_sregs *sregs); #endif diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 79468ddfe473..69e0d67a309b 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -2959,8 +2959,8 @@ void kvm_handle_exception_payload_quirk(struct kvm_vcpu *vcpu) kvm_deliver_exception_payload(vcpu, ex); } -static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu, - struct kvm_vcpu_events *events) +void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu, + struct kvm_vcpu_events *events) { struct kvm_queued_exception *ex = kvm_get_exception_to_save(vcpu); @@ -3028,8 +3028,8 @@ static void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu, } } -static int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu, - struct kvm_vcpu_events *events) +int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu, + struct kvm_vcpu_events *events) { if (events->flags & ~(KVM_VCPUEVENT_VALID_NMI_PENDING | KVM_VCPUEVENT_VALID_SIPI_VECTOR diff --git a/arch/x86/kvm/x86.h b/arch/x86/kvm/x86.h index 0f5919b092e4..936ee660ccec 100644 --- a/arch/x86/kvm/x86.h +++ b/arch/x86/kvm/x86.h @@ -492,6 +492,10 @@ int kvm_task_switch(struct kvm_vcpu *vcpu, u16 tss_selector, int idt_index, int reason, bool has_error_code, u32 error_code); int __kvm_set_xcr(struct kvm_vcpu *vcpu, u32 index, u64 xcr); +void kvm_vcpu_ioctl_x86_get_vcpu_events(struct kvm_vcpu *vcpu, + struct kvm_vcpu_events *events); +int kvm_vcpu_ioctl_x86_set_vcpu_events(struct kvm_vcpu *vcpu, + struct kvm_vcpu_events *events); int kvm_emulate_xsetbv(struct kvm_vcpu *vcpu); int kvm_emulate_rdpmc(struct kvm_vcpu *vcpu); diff --git a/include/linux/kho/abi/kvm_x86.h b/include/linux/kho/abi/kvm_x86.h new file mode 100644 index 000000000000..316250482db5 --- /dev/null +++ b/include/linux/kho/abi/kvm_x86.h @@ -0,0 +1,84 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright (c) 2026, Google LLC. + * Pasha Tatashin + */ +#ifndef _LINUX_KHO_ABI_KVM_X86_H +#define _LINUX_KHO_ABI_KVM_X86_H + +#ifdef CONFIG_X86_64 + +#include +#include +#include +#include +#include +#include + +/** + * DOC: x86 KVM Live Update ABI + * + * x86 KVM uses the ABI defined below for preserving architectural VM and vCPU + * state across a kexec reboot using LUO. + * + * The vCPU-level architectural state contains the compact register sets and + * CPUID table for each vCPU. + * + * All sub-structures (struct kvm_regs, struct kvm_sregs, struct kvm_mp_state, + * struct kvm_cpuid_entry2) are uAPI contracts. + */ + +/** + * struct kvm_vcpu_arch_ser - Preserved x86 architectural vCPU state in RAM. + * @regs: General-purpose registers (uAPI struct kvm_regs). + * @sregs: Segment and control registers (uAPI struct kvm_sregs). + * @mp_state: Multiprocessor state (uAPI struct kvm_mp_state). + * @pad: Padding to maintain 64-bit alignment after mp_state. + * @xcrs: Extended control registers including XCR0 (uAPI struct kvm_xcrs). + * @lapic: In-kernel local APIC register state (uAPI struct kvm_lapic_state). + * @pad_xsave: Padding that aligns @xsave to 64 bytes; see the static_assert + * below. Must be zero. + * @xsave: Extended processor state (uAPI struct kvm_xsave). This is the + * only copy of the guest FPU state; there is no kvm_fpu twin. + * @events: vCPU exception, interrupt, NMI and SMI events (uAPI struct kvm_vcpu_events). + * @debugregs: Hardware debug registers DR0-DR7 (uAPI struct kvm_debugregs). + * @num_msrs: Number of valid MSR entries in msrs. + * @cpuid_nent: Number of valid CPUID entries immediately following msrs[num_msrs]. + * @msrs: Array of preserved architectural and paravirtual MSR entries, + * followed by @cpuid_nent struct kvm_cpuid_entry2 entries. + */ +struct kvm_vcpu_arch_ser { + struct kvm_regs regs; + struct kvm_sregs sregs; + struct kvm_mp_state mp_state; + u32 pad; + struct kvm_xcrs xcrs; + struct kvm_lapic_state lapic; + u8 pad_xsave[40]; + struct kvm_xsave xsave; + struct kvm_vcpu_events events; + struct kvm_debugregs debugregs; + u32 num_msrs; + u32 cpuid_nent; + struct kvm_msr_entry msrs[]; +} __packed; + +static_assert(offsetof(struct kvm_vcpu_arch_ser, msrs) % sizeof(u64) == 0, + "msrs must be 64-bit aligned"); + +/* + * @xsave must be a legal XSAVE destination. The structure is __packed, so + * without @pad_xsave the member lands at offset 1880, which is not even + * 16-byte aligned, and the XSAVE/XRSTOR family faults on anything less than + * 64. The allocation itself is fine: kho_alloc_preserve() returns a folio + * address, so the base is page aligned. + * + * If this assert fires, a member above @xsave changed size; adjust + * @pad_xsave rather than deleting the assert. + */ +static_assert(offsetof(struct kvm_vcpu_arch_ser, xsave) % 64 == 0, + "xsave must be 64-byte aligned to be XSAVE-able in place"); + +#endif /* CONFIG_X86_64 */ + +#endif /* _LINUX_KHO_ABI_KVM_X86_H */ -- 2.55.0.1082.g2b9226bbc0-goog