KVM: x86: Fixes for (benign?) truncation bugs
Subsystems (4)
Sean Christopherson
maintainerv2 cover KVM: x86: Fixes for (benign?) truncation bugs
Patches 01 and 02 fix theoretical bugs related to loading CRs through the emulator. The rest of the patches are a bunch of small fixes for cases where KVM reads/writes a 64-bit register outside of 64-bit mode.
I stumbled on this when puzzling over commit 0107973a80ad ("KVM: x86:
Introduce cr3_lm_rsvd_bits in kvm_vcpu_arch"), which stated that SEV guests failed to boot on PCID-enabled hosts. Why only PCID hosts?
After much staring, I realized that the initial CR3 load in rsm_enter_protected_mode() would skip the MAXPHYADDR check due to the vCPU not being in long mode. But due to the ordering problems with PCID, when PCID is enabled in the guest, the second load of CR3 would be done with long mode enabled and thus hit the SEV C-bit bug.
Changing kvm_set_cr3() made me look at the callers, and seeing that SVM didn't properly truncate the value made me look at everything else, and here we are.
Note, I strongly suspect the emulator still has bugs. But, unless the guest is deliberately trying to hit these types of bugs, even the ones fixed here, they're likely benign. I figured I was more likely to break something than I was to fix something by diving into the emulator, so I left it alone. For now. :-)
v2: Rebase to kvm/queue, commit 89a22e37c8c2 ("KVM: avoid "deadlock"
between install_new_memslots and MMU notifier")
v1: https://lkml.kernel.org/r/[email protected]
Sean Christopherson (9):
KVM: x86: Remove emulator's broken checks on CR0/CR3/CR4 loads
KVM: x86: Check CR3 GPA for validity regardless of vCPU mode
KVM: SVM: Truncate GPR value for DR and CR accesses in !64-bit mode
KVM: VMX: Truncate GPR value for DR and CR reads in !64-bit mode
KVM: nVMX: Truncate bits 63:32 of VMCS field on nested check in
!64-bit
KVM: nVMX: Truncate base/index GPR value on address calc in !64-bit
KVM: x86/xen: Drop RAX[63:32] when processing hypercall
KVM: SVM: Use default rAX size for INVLPGA emulation
KVM: x86: Rename GPR accessors to make mode-aware variants the
defaults
arch/x86/kvm/emulate.c | 68 +---------------------------------- arch/x86/kvm/kvm_cache_regs.h | 19 ++++++---- arch/x86/kvm/svm/svm.c | 12 +++++-- arch/x86/kvm/vmx/nested.c | 14 ++++---- arch/x86/kvm/vmx/vmx.c | 6 ++-- arch/x86/kvm/x86.c | 19 ++++++---- arch/x86/kvm/x86.h | 8 ++--- 7 files changed, 48 insertions(+), 98 deletions(-)
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 1/9 KVM: x86: Remove emulator's broken checks on CR0/CR3/CR4 loads
Remove the emulator's checks for illegal CR0, CR3, and CR4 values, as the checks are redundant, outdated, and in the case of SEV's C-bit, broken. The emulator manually calculates MAXPHYADDR from CPUID and neglects to mask off the C-bit. For all other checks, kvm_set_cr*() are a superset of the emulator checks, e.g. see CR4.LA57.
--- arch/x86/kvm/emulate.c | 68 +----------------------------------------- 1 file changed, 1 insertion(+), 67 deletions(-)
arch/x86/kvm/emulate.c +1 -67
@@ -4230,75 +4230,9 @@ static int check_cr_read(struct x86_emulate_ctxt *ctxt)static int check_cr_write (struct x86_emulate_ctxt *ctxt) { -u64 new_val = ctxt->src.val64; -int cr = ctxt->modrm_reg; -u64 efer =0 ; - -static u64 cr_reserved_bits[] = { -0xffffffff00000000ULL , -0 ,0 ,0 ,/* CR3 checked later */ -CR4_RESERVED_BITS , -0 ,0 ,0 , -CR8_RESERVED_BITS , - }; - -if (!valid_cr (cr)) +if (!valid_cr (ctxt->modrm_reg))return emulate_ud (ctxt); -if (new_val & cr_reserved_bits[cr]) -return emulate_gp (ctxt,0 ); - -switch (cr) { -case 0 : { -u64 cr4; -if (((new_val &X86_CR0_PG ) && !(new_val &X86_CR0_PE )) || - ((new_val &X86_CR0_NW ) && !(new_val &X86_CR0_CD ))) -return emulate_gp (ctxt,0 ); - - cr4 = ctxt->ops->get_cr (ctxt,4 ); - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); - -if ((new_val &X86_CR0_PG ) && (efer &EFER_LME ) && - !(cr4 &X86_CR4_PAE )) -return emulate_gp (ctxt,0 ); - -break ; - } -case 3 : { -u64 rsvd =0 ; - - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); -if (efer &EFER_LMA ) { -u64 maxphyaddr; -u32 eax, ebx, ecx, edx; - - eax =0x80000008 ; - ecx =0 ; -if (ctxt->ops->get_cpuid (ctxt, &eax, &ebx, &ecx, - &edx, true)) - maxphyaddr = eax &0xff ; -else - maxphyaddr =36 ; - rsvd =rsvd_bits (maxphyaddr,63 ); -if (ctxt->ops->get_cr (ctxt,4 ) &X86_CR4_PCIDE ) - rsvd &= ~X86_CR3_PCID_NOFLUSH ; - } - -if (new_val & rsvd) -return emulate_gp (ctxt,0 ); - -break ; - } -case 4 : { - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); - -if ((efer &EFER_LMA ) && !(new_val &X86_CR4_PAE )) -return emulate_gp (ctxt,0 ); - -break ; - } - } -return X86EMUL_CONTINUE ; }
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Paolo Bonzini
maintainerRe: [PATCH v2 1/9] KVM: x86: Remove emulator's broken checks on CR0/CR3/CR4 loads
Quoted 6 lines · Sean Christopherson
Remove the emulator's checks for illegal CR0, CR3, and CR4 values, as the checks are redundant, outdated, and in the case of SEV's C-bit, broken. The emulator manually calculates MAXPHYADDR from CPUID and neglects to mask off the C-bit. For all other checks, kvm_set_cr*() are a superset of the emulator checks, e.g. see CR4.LA57.
Quoted 6 lines
Fixes a780a3ea6282 ("KVM: X86: Fix reserved bits check for MOV to CR3")Cc Babu Moger <[email protected]>Signed-off-by Sean Christopherson <[email protected]>--- arch/x86/kvm/emulate.c | 68 +----------------------------------------- 1 file changed, 1 insertion(+), 67 deletions(-)
This can be (opportunistically ;)) squashed on top:
arch/x86/kvm/emulate.c +3 -11
@@ -4220,15 +4220,7 @@ static bool valid_cr(int nr) } } -static int check_cr_read (struct x86_emulate_ctxt *ctxt) -{ -if (!valid_cr (ctxt->modrm_reg)) -return emulate_ud (ctxt); - -return X86EMUL_CONTINUE ; -} - -static int check_cr_write (struct x86_emulate_ctxt *ctxt) +static int check_cr_access (struct x86_emulate_ctxt *ctxt) {if (!valid_cr (ctxt->modrm_reg))return emulate_ud (ctxt);
@@ -4775,10 +4767,10 @@ static const struct opcode twobyte_table[256] = {D (ImplicitOps | ModRM | SrcMem | NoAccess),/* 8 * reserved NOP */ D (ImplicitOps | ModRM | SrcMem | NoAccess),/* NOP + 7 * reserved NOP */ /* 0x20 - 0x2F */ -DIP (ModRM | DstMem | Priv | Op3264 | NoMod, cr_read, check_cr_read), +DIP (ModRM | DstMem | Priv | Op3264 | NoMod, cr_read, check_cr_access),DIP (ModRM | DstMem | Priv | Op3264 | NoMod, dr_read, check_dr_read),IIP (ModRM | SrcMem | Priv | Op3264 | NoMod, em_cr_write, cr_write, - check_cr_write), + check_cr_access),IIP (ModRM | SrcMem | Priv | Op3264 | NoMod, em_dr_write, dr_write, check_dr_write), N, N, N, N,
Quoted 82 lines
arch/x86/kvm/emulate.c +1
-67@@ -4230,75 +4230,9 @@ static int check_cr_read(struct x86_emulate_ctxt *ctxt)static int check_cr_write (struct x86_emulate_ctxt *ctxt) { -u64 new_val = ctxt->src.val64; -int cr = ctxt->modrm_reg; -u64 efer =0 ; - -static u64 cr_reserved_bits[] = { -0xffffffff00000000ULL , -0 ,0 ,0 ,/* CR3 checked later */ -CR4_RESERVED_BITS , -0 ,0 ,0 , -CR8_RESERVED_BITS , - }; - -if (!valid_cr (cr)) +if (!valid_cr (ctxt->modrm_reg))return emulate_ud (ctxt); -if (new_val & cr_reserved_bits[cr]) -return emulate_gp (ctxt,0 ); - -switch (cr) { -case 0 : { -u64 cr4; -if (((new_val &X86_CR0_PG ) && !(new_val &X86_CR0_PE )) || - ((new_val &X86_CR0_NW ) && !(new_val &X86_CR0_CD ))) -return emulate_gp (ctxt,0 ); - - cr4 = ctxt->ops->get_cr (ctxt,4 ); - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); - -if ((new_val &X86_CR0_PG ) && (efer &EFER_LME ) && - !(cr4 &X86_CR4_PAE )) -return emulate_gp (ctxt,0 ); - -break ; - } -case 3 : { -u64 rsvd =0 ; - - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); -if (efer &EFER_LMA ) { -u64 maxphyaddr; -u32 eax, ebx, ecx, edx; - - eax =0x80000008 ; - ecx =0 ; -if (ctxt->ops->get_cpuid (ctxt, &eax, &ebx, &ecx, - &edx, true)) - maxphyaddr = eax &0xff ; -else - maxphyaddr =36 ; - rsvd =rsvd_bits (maxphyaddr,63 ); -if (ctxt->ops->get_cr (ctxt,4 ) &X86_CR4_PCIDE ) - rsvd &= ~X86_CR3_PCID_NOFLUSH ; - } - -if (new_val & rsvd) -return emulate_gp (ctxt,0 ); - -break ; - } -case 4 : { - ctxt->ops->get_msr (ctxt,MSR_EFER , &efer); - -if ((efer &EFER_LMA ) && !(new_val &X86_CR4_PAE )) -return emulate_gp (ctxt,0 ); - -break ; - } - } -return X86EMUL_CONTINUE ; }
Sean Christopherson
maintainerPATCH v2 2/9 KVM: x86: Check CR3 GPA for validity regardless of vCPU mode
In reply to Sean ChristophersonCheck CR3 for an invalid GPA even if the vCPU isn't in long mode. For bigger emulation flows, notably RSM, the vCPU mode may not be accurate if CR0/CR4 are loaded after CR3. For MOV CR3 and similar flows, the caller is responsible for truncating the value.
Note, SMRAM.CR3 is read-only, so this is mostly a theoretical bug since KVM will not have stored an illegal CR3 into SMRAM during SMI emulation.
--- arch/x86/kvm/x86.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)
arch/x86/kvm/x86.c +8 -3
@@ -1078,10 +1078,15 @@ int kvm_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)return 0 ; } -if (is_long_mode (vcpu) &&kvm_vcpu_is_illegal_gpa (vcpu, cr3)) +/* +* Do not condition the GPA check on long mode, this helper is used to +* stuff CR3, e.g. for RSM emulation, and there is no guarantee that +* the current vCPU mode is accurate. +*/ +if (kvm_vcpu_is_illegal_gpa (vcpu, cr3))return 1 ; -else if (is_pae_paging (vcpu) && - !load_pdptrs (vcpu, vcpu->arch.walk_mmu, cr3)) + +if (is_pae_paging (vcpu) && !load_pdptrs (vcpu, vcpu->arch.walk_mmu, cr3))return 1 ;kvm_mmu_new_pgd (vcpu, cr3, skip_tlb_flush, skip_tlb_flush);
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Paolo Bonzini
maintainerRe: [PATCH v2 2/9] KVM: x86: Check CR3 GPA for validity regardless of vCPU mode
Quoted 7 lines · Sean Christopherson
Check CR3 for an invalid GPA even if the vCPU isn't in long mode. For bigger emulation flows, notably RSM, the vCPU mode may not be accurate if CR0/CR4 are loaded after CR3. For MOV CR3 and similar flows, the caller is responsible for truncating the value.
Note, SMRAM.CR3 is read-only, so this is mostly a theoretical bug since KVM will not have stored an illegal CR3 into SMRAM during SMI emulation.
Well, the guest could have changed it...
Paolo
Quoted 31 lines
Fixes 660a5d517aaa ("KVM: x86: save/load state on SMM switch")Signed-off-by Sean Christopherson <[email protected]>--- arch/x86/kvm/x86.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-)arch/x86/kvm/x86.c +8
-3@@ -1078,10 +1078,15 @@ int kvm_set_cr3(struct kvm_vcpu *vcpu, unsigned long cr3)return 0 ; } -if (is_long_mode (vcpu) &&kvm_vcpu_is_illegal_gpa (vcpu, cr3)) +/* +* Do not condition the GPA check on long mode, this helper is used to +* stuff CR3, e.g. for RSM emulation, and there is no guarantee that +* the current vCPU mode is accurate. +*/ +if (kvm_vcpu_is_illegal_gpa (vcpu, cr3))return 1 ; -else if (is_pae_paging (vcpu) && - !load_pdptrs (vcpu, vcpu->arch.walk_mmu, cr3)) + +if (is_pae_paging (vcpu) && !load_pdptrs (vcpu, vcpu->arch.walk_mmu, cr3))return 1 ;kvm_mmu_new_pgd (vcpu, cr3, skip_tlb_flush, skip_tlb_flush);
Sean Christopherson
maintainerRe: [PATCH v2 2/9] KVM: x86: Check CR3 GPA for validity regardless of vCPU mode
Quoted 10 lines · Paolo Bonzini
Quoted 7 lines · Sean Christopherson
Check CR3 for an invalid GPA even if the vCPU isn't in long mode. For bigger emulation flows, notably RSM, the vCPU mode may not be accurate if CR0/CR4 are loaded after CR3. For MOV CR3 and similar flows, the caller is responsible for truncating the value.
Note, SMRAM.CR3 is read-only, so this is mostly a theoretical bug since KVM will not have stored an illegal CR3 into SMRAM during SMI emulation.
Well, the guest could have changed it...
That's what I tried to address with "SMRAM.CR3 is read-only". Both Intel and AMD state that modifying read-only fields will result in unpredictable behavior, i.e. KVM going into the weeds would be within spec. IIRC, there's no real danger to the host, it'll "just" fail VM-Enter.
SDM:
Some register images are read-only, and must not be modified (modifying these registers will result in unpredictable behavior)
APM:
Software should not modify offsets specified as read-only or reserved, otherwise unpredictable results can occur.
Sean Christopherson
maintainerPATCH v2 3/9 KVM: SVM: Truncate GPR value for DR and CR accesses in !64-bit mode
In reply to Sean ChristophersonDrop bits 63:32 on loads/stores to/from DRs and CRs when the vCPU is not in 64-bit mode. The APM states bits 63:32 are dropped for both DRs and CRs:
In 64-bit mode, the operand size is fixed at 64 bits without the need for a REX prefix. In non-64-bit mode, the operand size is fixed at 32 bits and the upper 32 bits of the destination are forced to 0.
--- arch/x86/kvm/svm/svm.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-)
arch/x86/kvm/svm/svm.c +4 -4
@@ -2462,7 +2462,7 @@ static int cr_interception(struct kvm_vcpu *vcpu) err =0 ;if (cr >=16 ) {/* mov to cr */ cr -=16 ; - val =kvm_register_read (vcpu, reg); + val =kvm_register_readl (vcpu, reg);trace_kvm_cr_write (cr, val);switch (cr) {case 0 :
@@ -2508,7 +2508,7 @@ static int cr_interception(struct kvm_vcpu *vcpu)kvm_queue_exception (vcpu,UD_VECTOR );return 1 ; } -kvm_register_write (vcpu, reg, val); +kvm_register_writel (vcpu, reg, val);trace_kvm_cr_read (cr, val); }return kvm_complete_insn_gp (vcpu, err);
@@ -2574,11 +2574,11 @@ static int dr_interception(struct kvm_vcpu *vcpu) dr = svm->vmcb->control.exit_code -SVM_EXIT_READ_DR0 ;if (dr >=16 ) {/* mov to DRn */ dr -=16 ; - val =kvm_register_read (vcpu, reg); + val =kvm_register_readl (vcpu, reg); err =kvm_set_dr (vcpu, dr, val); }else {kvm_get_dr (vcpu, dr, &val); -kvm_register_write (vcpu, reg, val); +kvm_register_writel (vcpu, reg, val); }return kvm_complete_insn_gp (vcpu, err);
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 4/9 KVM: VMX: Truncate GPR value for DR and CR reads in !64-bit mode
In reply to Sean ChristophersonDrop bits 63:32 when storing a DR/CR to a GPR when the vCPU is not in 64-bit mode. Per the SDM:
The operand size for these instructions is always 32 bits in non-64-bit modes, regardless of the operand-size attribute.
CR8 technically isn't affected as CR8 isn't accessible outside of 64-bit mode, but fix it up for consistency and to allow for future cleanup.
--- arch/x86/kvm/vmx/vmx.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
arch/x86/kvm/vmx/vmx.c +3 -3
@@ -5132,12 +5132,12 @@ static int handle_cr(struct kvm_vcpu *vcpu)case 3 :WARN_ON_ONCE (enable_unrestricted_guest); val =kvm_read_cr3 (vcpu); -kvm_register_write (vcpu, reg, val); +kvm_register_writel (vcpu, reg, val);trace_kvm_cr_read (cr, val);return kvm_skip_emulated_instruction (vcpu);case 8 : val =kvm_get_cr8 (vcpu); -kvm_register_write (vcpu, reg, val); +kvm_register_writel (vcpu, reg, val);trace_kvm_cr_read (cr, val);return kvm_skip_emulated_instruction (vcpu); }
@@ -5210,7 +5210,7 @@ static int handle_dr(struct kvm_vcpu *vcpu)unsigned long val;kvm_get_dr (vcpu, dr, &val); -kvm_register_write (vcpu, reg, val); +kvm_register_writel (vcpu, reg, val); err =0 ; }else { err =kvm_set_dr (vcpu, dr,kvm_register_readl (vcpu, reg));
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 5/9 KVM: nVMX: Truncate bits 63:32 of VMCS field on nested check in !64-bit
In reply to Sean ChristophersonDrop bits 63:32 of the VMCS field encoding when checking for a nested VM-Exit on VMREAD/VMWRITE in !64-bit mode. VMREAD and VMWRITE always use 32-bit operands outside of 64-bit mode.
The actual emulation of VMREAD/VMWRITE does the right thing, this bug is purely limited to incorrectly causing a nested VM-Exit if a GPR happens to have bits 63:32 set outside of 64-bit mode.
--- arch/x86/kvm/vmx/nested.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/vmx/nested.c +1 -1
@@ -5743,7 +5743,7 @@ static bool nested_vmx_exit_handled_vmcs_access(struct kvm_vcpu *vcpu,/* Decode instruction info and find the field to access */ vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO ); - field =kvm_register_read (vcpu, (((vmx_instruction_info) >>28 ) &0xf )); + field =kvm_register_readl (vcpu, (((vmx_instruction_info) >>28 ) &0xf ));/* Out-of-range fields always cause a VM exit from L2 to L1 */ if (field >>15 )
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 6/9 KVM: nVMX: Truncate base/index GPR value on address calc in !64-bit
In reply to Sean ChristophersonDrop bits 63:32 of the base and/or index GPRs when calculating the effective address of a VMX instruction memory operand. Outside of 64-bit mode, memory encodings are strictly limited to E*X and below.
--- arch/x86/kvm/vmx/nested.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
arch/x86/kvm/vmx/nested.c +2 -2
@@ -4617,9 +4617,9 @@ int get_vmx_mem_address(struct kvm_vcpu *vcpu, unsigned long exit_qualification,else if (addr_size ==0 ) off = (gva_t )sign_extend64 (off,15 );if (base_is_valid) - off +=kvm_register_read (vcpu, base_reg); + off +=kvm_register_readl (vcpu, base_reg);if (index_is_valid) - off +=kvm_register_read (vcpu, index_reg) << scaling; + off +=kvm_register_readl (vcpu, index_reg) << scaling;vmx_get_segment (vcpu, &s, seg_reg);/*
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 7/9 KVM: x86/xen: Drop RAX[63:32] when processing hypercall
In reply to Sean ChristophersonTruncate RAX to 32 bits, i.e. consume EAX, when retrieving the hypecall index for a Xen hypercall. Per Xen documentation[*], the index is EAX when the vCPU is not in 64-bit mode.
[*] http://xenbits.xenproject.org/docs/sphinx-unstable/guest-guide/x86/hypercall-abi.html
--- arch/x86/kvm/xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kvm/xen.c +1 -1
@@ -673,7 +673,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)bool longmode;u64 input, params[6 ]; - input = (u64 )kvm_register_read (vcpu,VCPU_REGS_RAX ); + input = (u64 )kvm_register_readl (vcpu,VCPU_REGS_RAX );/* Hyper-V hypercalls get bit 31 set in EAX */ if ((input & 0x80000000) &&
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Re: [PATCH v2 7/9] KVM: x86/xen: Drop RAX[63:32] when processing hypercall
Quoted 20 lines · Sean Christopherson
Truncate RAX to 32 bits, i.e. consume EAX, when retrieving the hypecall index for a Xen hypercall. Per Xen documentation[*], the index is EAX when the vCPU is not in 64-bit mode.
[*] http://xenbits.xenproject.org/docs/sphinx-unstable/guest-guide/x86/hypercall-abi.html
Fixes 23200b7a30de ("KVM: x86/xen: intercept xen hypercalls if enabled")Cc Joao Martins <[email protected]>Cc David Woodhouse <[email protected]>Signed-off-by Sean Christopherson <[email protected]>--- arch/x86/kvm/xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/x86/kvm/xen.c +0
-0
arch/x86/kvm/xen.c int kvm_xen_hypercall(struct kvm_vcpu *vcpu) · line 679
Quoted 8 lines
arch/x86/kvm/xen.c +1
-1@@ -673,7 +673,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)bool longmode;u64 input, params[6 ]; - input = (u64 )kvm_register_read (vcpu,VCPU_REGS_RAX ); + input = (u64 )kvm_register_readl (vcpu,VCPU_REGS_RAX );/* Hyper-V hypercalls get bit 31 set in EAX */ if ((input & 0x80000000) &&
Alternatively, as a minor optimization, you could've used '!longmode' check below, something like:
arch/x86/kvm/xen.c +1 -0
@@ -682,6 +682,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu) longmode =is_64_bit_mode (vcpu);if (!longmode) { + input = (u32 )input; params[0 ] = (u32 )kvm_rbx_read (vcpu); params[1 ] = (u32 )kvm_rcx_read (vcpu); params[2 ] = (u32 )kvm_rdx_read (vcpu);
Signature · 1 line
Vitaly
Paolo Bonzini
maintainerRe: [PATCH v2 7/9] KVM: x86/xen: Drop RAX[63:32] when processing hypercall
Quoted 42 lines · Vitaly Kuznetsov
Quoted 20 lines · Sean Christopherson
Truncate RAX to 32 bits, i.e. consume EAX, when retrieving the hypecall index for a Xen hypercall. Per Xen documentation[*], the index is EAX when the vCPU is not in 64-bit mode.
[*] http://xenbits.xenproject.org/docs/sphinx-unstable/guest-guide/x86/hypercall-abi.html
Fixes 23200b7a30de ("KVM: x86/xen: intercept xen hypercalls if enabled")Cc Joao Martins <[email protected]>Cc David Woodhouse <[email protected]>Signed-off-by Sean Christopherson <[email protected]>--- arch/x86/kvm/xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/x86/kvm/xen.c +0
-0arch/x86/kvm/xen.c int kvm_xen_hypercall(struct kvm_vcpu *vcpu) · line 679
Quoted 8 lines
arch/x86/kvm/xen.c +1
-1@@ -673,7 +673,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)bool longmode;u64 input, params[6 ]; - input = (u64 )kvm_register_read (vcpu,VCPU_REGS_RAX ); + input = (u64 )kvm_register_readl (vcpu,VCPU_REGS_RAX );/* Hyper-V hypercalls get bit 31 set in EAX */ if ((input & 0x80000000) &&Reviewed-by Vitaly Kuznetsov <[email protected]>Alternatively, as a minor optimization, you could've used '!longmode' check below, something like:
arch/x86/kvm/xen.c +0
-0
arch/x86/kvm/xen.c int kvm_xen_hypercall(struct kvm_vcpu *vcpu) · line 688
Quoted 7 lines
arch/x86/kvm/xen.c +1
-0@@ -682,6 +682,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu) longmode =is_64_bit_mode (vcpu);if (!longmode) { + input = (u32 )input; params[0 ] = (u32 )kvm_rbx_read (vcpu); params[1 ] = (u32 )kvm_rcx_read (vcpu); params[2 ] = (u32 )kvm_rdx_read (vcpu);
You haven't seen patch 9 yet. :)
Paolo
Re: [PATCH v2 7/9] KVM: x86/xen: Drop RAX[63:32] when processing hypercall
Quoted 43 lines · Paolo Bonzini
Quoted 42 lines · Vitaly Kuznetsov
Quoted 20 lines · Sean Christopherson
Truncate RAX to 32 bits, i.e. consume EAX, when retrieving the hypecall index for a Xen hypercall. Per Xen documentation[*], the index is EAX when the vCPU is not in 64-bit mode.
[*] http://xenbits.xenproject.org/docs/sphinx-unstable/guest-guide/x86/hypercall-abi.html
Fixes 23200b7a30de ("KVM: x86/xen: intercept xen hypercalls if enabled")Cc Joao Martins <[email protected]>Cc David Woodhouse <[email protected]>Signed-off-by Sean Christopherson <[email protected]>--- arch/x86/kvm/xen.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)arch/x86/kvm/xen.c +0
-0arch/x86/kvm/xen.c int kvm_xen_hypercall(struct kvm_vcpu *vcpu) · line 679
Quoted 8 lines
arch/x86/kvm/xen.c +1
-1@@ -673,7 +673,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)bool longmode;u64 input, params[6 ]; - input = (u64 )kvm_register_read (vcpu,VCPU_REGS_RAX ); + input = (u64 )kvm_register_readl (vcpu,VCPU_REGS_RAX );/* Hyper-V hypercalls get bit 31 set in EAX */ if ((input & 0x80000000) &&Reviewed-by Vitaly Kuznetsov <[email protected]>Alternatively, as a minor optimization, you could've used '!longmode' check below, something like:
arch/x86/kvm/xen.c +0
-0@@ -682,6 +682,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)
Quoted 10 lines
arch/x86/kvm/xen.c int kvm_xen_hypercall(struct kvm_vcpu *vcpu) · line 688
Quoted 7 lines
arch/x86/kvm/xen.c +1
-0longmode =is_64_bit_mode (vcpu);if (!longmode) { + input = (u32 )input; params[0 ] = (u32 )kvm_rbx_read (vcpu); params[1 ] = (u32 )kvm_rcx_read (vcpu); params[2 ] = (u32 )kvm_rdx_read (vcpu);You haven't seen patch 9 yet. :)
True; suggestion dismissed :-)
Signature · 1 line
Vitaly
Sean Christopherson
maintainerPATCH v2 8/9 KVM: SVM: Use default rAX size for INVLPGA emulation
In reply to Sean ChristophersonDrop bits 63:32 of RAX when grabbing the address for INVLPGA emulation outside of 64-bit mode to make KVM's emulation slightly less wrong. The address for INVLPGA is determined by the effective address size, i.e. it's not hardcoded to 64/32 bits for a given mode. Add a FIXME to call out that the emulation is wrong.
Opportunistically tweak the ASID handling to make it clear that it's defined by ECX, not rCX.
Per the APM:
The portion of rAX used to form the address is determined by the effective address size (current execution mode and optional address size prefix). The ASID is taken from ECX.
--- arch/x86/kvm/svm/svm.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-)
arch/x86/kvm/svm/svm.c +9 -3
@@ -2301,11 +2301,17 @@ static int clgi_interception(struct kvm_vcpu *vcpu)static int invlpga_interception (struct kvm_vcpu *vcpu) { -trace_kvm_invlpga (to_svm (vcpu)->vmcb->save.rip,kvm_rcx_read (vcpu), -kvm_rax_read (vcpu)); +gva_t gva =kvm_rax_read (vcpu); +u32 asid =kvm_rcx_read (vcpu); + +/* FIXME: Handle an address size prefix. */ +if (!is_long_mode (vcpu)) + gva = (u32 )gva; + +trace_kvm_invlpga (to_svm (vcpu)->vmcb->save.rip, asid, gva);/* Let's treat INVLPGA the same as INVLPG (can be optimized!) */ -kvm_mmu_invlpg (vcpu,kvm_rax_read (vcpu)); +kvm_mmu_invlpg (vcpu, gva);return kvm_skip_emulated_instruction (vcpu); }
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Sean Christopherson
maintainerPATCH v2 9/9 KVM: x86: Rename GPR accessors to make mode-aware variants the defaults
In reply to Sean ChristophersonAppend raw to the direct variants of kvm_register_read/write(), and drop the "l" from the mode-aware variants. I.e. make the mode-aware variants the default, and make the direct variants scary sounding so as to discourage use. Accessing the full 64-bit values irrespective of mode is rarely the desired behavior.
--- arch/x86/kvm/kvm_cache_regs.h | 19 ++++++++++++------- arch/x86/kvm/svm/svm.c | 8 ++++---- arch/x86/kvm/vmx/nested.c | 20 ++++++++++---------- arch/x86/kvm/vmx/vmx.c | 12 ++++++------ arch/x86/kvm/x86.c | 8 ++++---- arch/x86/kvm/x86.h | 8 ++++---- arch/x86/kvm/xen.c | 2 +- 7 files changed, 41 insertions(+), 36 deletions(-)
arch/x86/kvm/kvm_cache_regs.h +12 -7
@@ -62,7 +62,12 @@ static inline void kvm_register_mark_dirty(struct kvm_vcpu *vcpu,__set_bit (reg, (unsigned long *)&vcpu->arch.regs_dirty); } -static inline unsigned long kvm_register_read (struct kvm_vcpu *vcpu,int reg) +/* +* The "raw" register helpers are only for cases where the full 64 bits of a +* register are read/written irrespective of current vCPU mode. In other words, +* odds are good you shouldn't be using the raw variants. +*/ +static inline unsigned long kvm_register_read_raw (struct kvm_vcpu *vcpu,int reg) {if (WARN_ON_ONCE ((unsigned int )reg >=NR_VCPU_REGS ))return 0 ;
@@ -73,8 +78,8 @@ static inline unsigned long kvm_register_read(struct kvm_vcpu *vcpu, int reg)return vcpu->arch.regs[reg]; } -static inline void kvm_register_write (struct kvm_vcpu *vcpu,int reg, -unsigned long val) +static inline void kvm_register_write_raw (struct kvm_vcpu *vcpu,int reg, +unsigned long val) {if (WARN_ON_ONCE ((unsigned int )reg >=NR_VCPU_REGS ))return ;
@@ -85,22 +90,22 @@ static inline void kvm_register_write(struct kvm_vcpu *vcpu, int reg,static inline unsigned long kvm_rip_read (struct kvm_vcpu *vcpu) { -return kvm_register_read (vcpu,VCPU_REGS_RIP ); +return kvm_register_read_raw (vcpu,VCPU_REGS_RIP ); }static inline void kvm_rip_write (struct kvm_vcpu *vcpu,unsigned long val) { -kvm_register_write (vcpu,VCPU_REGS_RIP , val); +kvm_register_write_raw (vcpu,VCPU_REGS_RIP , val); }static inline unsigned long kvm_rsp_read (struct kvm_vcpu *vcpu) { -return kvm_register_read (vcpu,VCPU_REGS_RSP ); +return kvm_register_read_raw (vcpu,VCPU_REGS_RSP ); }static inline void kvm_rsp_write (struct kvm_vcpu *vcpu,unsigned long val) { -kvm_register_write (vcpu,VCPU_REGS_RSP , val); +kvm_register_write_raw (vcpu,VCPU_REGS_RSP , val); }static inline u64 kvm_pdptr_read (struct kvm_vcpu *vcpu,int index)
arch/x86/kvm/svm/svm.c +4 -4
@@ -2468,7 +2468,7 @@ static int cr_interception(struct kvm_vcpu *vcpu) err =0 ;if (cr >=16 ) {/* mov to cr */ cr -=16 ; - val =kvm_register_readl (vcpu, reg); + val =kvm_register_read (vcpu, reg);trace_kvm_cr_write (cr, val);switch (cr) {case 0 :
@@ -2514,7 +2514,7 @@ static int cr_interception(struct kvm_vcpu *vcpu)kvm_queue_exception (vcpu,UD_VECTOR );return 1 ; } -kvm_register_writel (vcpu, reg, val); +kvm_register_write (vcpu, reg, val);trace_kvm_cr_read (cr, val); }return kvm_complete_insn_gp (vcpu, err);
@@ -2580,11 +2580,11 @@ static int dr_interception(struct kvm_vcpu *vcpu) dr = svm->vmcb->control.exit_code -SVM_EXIT_READ_DR0 ;if (dr >=16 ) {/* mov to DRn */ dr -=16 ; - val =kvm_register_readl (vcpu, reg); + val =kvm_register_read (vcpu, reg); err =kvm_set_dr (vcpu, dr, val); }else {kvm_get_dr (vcpu, dr, &val); -kvm_register_writel (vcpu, reg, val); +kvm_register_write (vcpu, reg, val); }return kvm_complete_insn_gp (vcpu, err);
arch/x86/kvm/vmx/nested.c +10 -10
@@ -4617,9 +4617,9 @@ int get_vmx_mem_address(struct kvm_vcpu *vcpu, unsigned long exit_qualification,else if (addr_size ==0 ) off = (gva_t )sign_extend64 (off,15 );if (base_is_valid) - off +=kvm_register_readl (vcpu, base_reg); + off +=kvm_register_read (vcpu, base_reg);if (index_is_valid) - off +=kvm_register_readl (vcpu, index_reg) << scaling; + off +=kvm_register_read (vcpu, index_reg) << scaling;vmx_get_segment (vcpu, &s, seg_reg);/*
@@ -5021,7 +5021,7 @@ static int handle_vmread(struct kvm_vcpu *vcpu)return nested_vmx_failInvalid (vcpu);/* Decode instruction info and find the field to read */ - field =kvm_register_readl (vcpu, (((instr_info) >>28 ) &0xf )); + field =kvm_register_read (vcpu, (((instr_info) >>28 ) &0xf )); offset =vmcs_field_to_offset (field);if (offset <0 )
@@ -5039,7 +5039,7 @@ static int handle_vmread(struct kvm_vcpu *vcpu)* on the guest's mode (32 or 64 bit), not on the given field's length. */ if (instr_info &BIT (10 )) { -kvm_register_writel (vcpu, (((instr_info) >>3 ) &0xf ), value); +kvm_register_write (vcpu, (((instr_info) >>3 ) &0xf ), value); }else { len =is_64_bit_mode (vcpu) ?8 :4 ; if (get_vmx_mem_address(vcpu, exit_qualification,
@@ -5113,7 +5113,7 @@ static int handle_vmwrite(struct kvm_vcpu *vcpu)return nested_vmx_failInvalid (vcpu);if (instr_info &BIT (10 )) - value =kvm_register_readl (vcpu, (((instr_info) >>3 ) &0xf )); + value =kvm_register_read (vcpu, (((instr_info) >>3 ) &0xf ));else { len =is_64_bit_mode (vcpu) ?8 :4 ; if (get_vmx_mem_address(vcpu, exit_qualification,
@@ -5124,7 +5124,7 @@ static int handle_vmwrite(struct kvm_vcpu *vcpu)return kvm_handle_memory_failure (vcpu, r, &e); } - field =kvm_register_readl (vcpu, (((instr_info) >>28 ) &0xf )); + field =kvm_register_read (vcpu, (((instr_info) >>28 ) &0xf )); offset =vmcs_field_to_offset (field);if (offset <0 )
@@ -5321,7 +5321,7 @@ static int handle_invept(struct kvm_vcpu *vcpu)return 1 ; vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO ); - type =kvm_register_readl (vcpu, (vmx_instruction_info >>28 ) &0xf ); + type =kvm_register_read (vcpu, (vmx_instruction_info >>28 ) &0xf ); types = (vmx->nested.msrs.ept_caps >>VMX_EPT_EXTENT_SHIFT ) &6 ;
@@ -5401,7 +5401,7 @@ static int handle_invvpid(struct kvm_vcpu *vcpu)return 1 ; vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO ); - type =kvm_register_readl (vcpu, (vmx_instruction_info >>28 ) &0xf ); + type =kvm_register_read (vcpu, (vmx_instruction_info >>28 ) &0xf ); types = (vmx->nested.msrs.vpid_caps &VMX_VPID_EXTENT_SUPPORTED_MASK ) >>8 ;
@@ -5657,7 +5657,7 @@ static bool nested_vmx_exit_handled_cr(struct kvm_vcpu *vcpu,switch ((exit_qualification >>4 ) &3 ) {case 0 :/* mov to cr */ reg = (exit_qualification >>8 ) &15 ; - val =kvm_register_readl (vcpu, reg); + val =kvm_register_read (vcpu, reg);switch (cr) {case 0 : if (vmcs12->cr0_guest_host_mask &
@@ -5743,7 +5743,7 @@ static bool nested_vmx_exit_handled_vmcs_access(struct kvm_vcpu *vcpu,/* Decode instruction info and find the field to access */ vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO ); - field =kvm_register_readl (vcpu, (((vmx_instruction_info) >>28 ) &0xf )); + field =kvm_register_read (vcpu, (((vmx_instruction_info) >>28 ) &0xf ));/* Out-of-range fields always cause a VM exit from L2 to L1 */ if (field >>15 )
arch/x86/kvm/vmx/vmx.c +6 -6
@@ -5090,7 +5090,7 @@ static int handle_cr(struct kvm_vcpu *vcpu) reg = (exit_qualification >>8 ) &15 ;switch ((exit_qualification >>4 ) &3 ) {case 0 :/* mov to cr */ - val =kvm_register_readl (vcpu, reg); + val =kvm_register_read (vcpu, reg);trace_kvm_cr_write (cr, val);switch (cr) {case 0 :
@@ -5132,12 +5132,12 @@ static int handle_cr(struct kvm_vcpu *vcpu)case 3 :WARN_ON_ONCE (enable_unrestricted_guest); val =kvm_read_cr3 (vcpu); -kvm_register_writel (vcpu, reg, val); +kvm_register_write (vcpu, reg, val);trace_kvm_cr_read (cr, val);return kvm_skip_emulated_instruction (vcpu);case 8 : val =kvm_get_cr8 (vcpu); -kvm_register_writel (vcpu, reg, val); +kvm_register_write (vcpu, reg, val);trace_kvm_cr_read (cr, val);return kvm_skip_emulated_instruction (vcpu); }
@@ -5210,10 +5210,10 @@ static int handle_dr(struct kvm_vcpu *vcpu)unsigned long val;kvm_get_dr (vcpu, dr, &val); -kvm_register_writel (vcpu, reg, val); +kvm_register_write (vcpu, reg, val); err =0 ; }else { - err =kvm_set_dr (vcpu, dr,kvm_register_readl (vcpu, reg)); + err =kvm_set_dr (vcpu, dr,kvm_register_read (vcpu, reg)); } out:
@@ -5565,7 +5565,7 @@ static int handle_invpcid(struct kvm_vcpu *vcpu) } vmx_instruction_info =vmcs_read32 (VMX_INSTRUCTION_INFO ); - type =kvm_register_readl (vcpu, (vmx_instruction_info >>28 ) &0xf ); + type =kvm_register_read (vcpu, (vmx_instruction_info >>28 ) &0xf );if (type >3 ) {kvm_inject_gp (vcpu,0 );
arch/x86/kvm/x86.c +4 -4
@@ -6974,12 +6974,12 @@ static bool emulator_guest_has_fxsr(struct x86_emulate_ctxt *ctxt)static ulong emulator_read_gpr (struct x86_emulate_ctxt *ctxt,unsigned reg) { -return kvm_register_read (emul_to_vcpu (ctxt), reg); +return kvm_register_read_raw (emul_to_vcpu (ctxt), reg); }static void emulator_write_gpr (struct x86_emulate_ctxt *ctxt,unsigned reg,ulong val) { -kvm_register_write (emul_to_vcpu (ctxt), reg, val); +kvm_register_write_raw (emul_to_vcpu (ctxt), reg, val); }static void emulator_set_nmi_mask (struct x86_emulate_ctxt *ctxt,bool masked)
@@ -8660,7 +8660,7 @@ static void enter_smm_save_state_32(struct kvm_vcpu *vcpu, char *buf)put_smstate (u32, buf,0x7ff0 ,kvm_rip_read (vcpu));for (i =0 ; i <8 ; i++) -put_smstate (u32, buf,0x7fd0 + i *4 ,kvm_register_read (vcpu, i)); +put_smstate (u32, buf,0x7fd0 + i *4 ,kvm_register_read_raw (vcpu, i));kvm_get_dr (vcpu,6 , &val);put_smstate (u32, buf,0x7fcc , (u32 )val);
@@ -8706,7 +8706,7 @@ static void enter_smm_save_state_64(struct kvm_vcpu *vcpu, char *buf)int i;for (i =0 ; i <16 ; i++) -put_smstate (u64, buf,0x7ff8 - i *8 ,kvm_register_read (vcpu, i)); +put_smstate (u64, buf,0x7ff8 - i *8 ,kvm_register_read_raw (vcpu, i));put_smstate (u64, buf,0x7f78 ,kvm_rip_read (vcpu));put_smstate (u32, buf,0x7f70 ,kvm_get_rflags (vcpu));
arch/x86/kvm/x86.h +4 -4
@@ -232,19 +232,19 @@ static inline bool vcpu_match_mmio_gpa(struct kvm_vcpu *vcpu, gpa_t gpa)return false; } -static inline unsigned long kvm_register_readl (struct kvm_vcpu *vcpu,int reg) +static inline unsigned long kvm_register_read (struct kvm_vcpu *vcpu,int reg) { -unsigned long val =kvm_register_read (vcpu, reg); +unsigned long val =kvm_register_read_raw (vcpu, reg);return is_64_bit_mode (vcpu) ? val : (u32 )val; } -static inline void kvm_register_writel (struct kvm_vcpu *vcpu, +static inline void kvm_register_write (struct kvm_vcpu *vcpu,int reg,unsigned long val) {if (!is_64_bit_mode (vcpu)) val = (u32 )val; -return kvm_register_write (vcpu, reg, val); +return kvm_register_write_raw (vcpu, reg, val); }static inline bool kvm_check_has_quirk (struct kvm *kvm,u64 quirk)
arch/x86/kvm/xen.c +1 -1
@@ -673,7 +673,7 @@ int kvm_xen_hypercall(struct kvm_vcpu *vcpu)bool longmode;u64 input, params[6 ]; - input = (u64 )kvm_register_readl (vcpu,VCPU_REGS_RAX ); + input = (u64 )kvm_register_read (vcpu,VCPU_REGS_RAX );/* Hyper-V hypercalls get bit 31 set in EAX */ if ((input & 0x80000000) &&
Signature · 1 line
2.31.1.498.g6c1eba8ee3d-goog
Quoted 42 lines · Sean Christopherson
Patches 01 and 02 fix theoretical bugs related to loading CRs through the emulator. The rest of the patches are a bunch of small fixes for cases where KVM reads/writes a 64-bit register outside of 64-bit mode.
I stumbled on this when puzzling over commit
0107973a80ad("KVM: x86:
Introducecr3_lm_rsvd_bitsinkvm_vcpu_arch"), which stated that SEV guests failed to boot on PCID-enabled hosts. Why only PCID hosts?After much staring, I realized that the initial CR3 load in
rsm_enter_protected_mode()would skip the MAXPHYADDR check due to the vCPU not being in long mode. But due to the ordering problems with PCID, when PCID is enabled in the guest, the second load of CR3 would be done with long mode enabled and thus hit the SEV C-bit bug.Changing
kvm_set_cr3()made me look at the callers, and seeing that SVM didn't properly truncate the value made me look at everything else, and here we are.Note, I strongly suspect the emulator still has bugs. But, unless the guest is deliberately trying to hit these types of bugs, even the ones fixed here, they're likely benign. I figured I was more likely to break something than I was to fix something by diving into the emulator, so I left it alone. For now. :-)
v2: Rebase to kvm/queue, commit
89a22e37c8c2("KVM: avoid "deadlock"
betweeninstall_new_memslotsand MMU notifier")v1: https://lkml.kernel.org/r/[email protected]
Sean Christopherson (9):
KVM: x86: Remove emulator's broken checks on CR0/CR3/CR4 loads
KVM: x86: Check CR3 GPA for validity regardless of vCPU mode
KVM: SVM: Truncate GPR value for DR and CR accesses in !64-bit mode
KVM: VMX: Truncate GPR value for DR and CR reads in !64-bit mode
KVM: nVMX: Truncate bits 63:32 of VMCS field on nested check in!64-bitKVM: nVMX: Truncate base/index GPR value on address calc in !64-bit
KVM: x86/xen: Drop RAX[63:32] when processing hypercall
KVM: SVM: Use default rAX size for INVLPGA emulation
KVM: x86: Rename GPR accessors to make mode-aware variants the
defaults
Quoted 9 lines
arch/x86/kvm/emulate.c | 68 +---------------------------------- arch/x86/kvm/kvm_cache_regs.h | 19 ++++++---- arch/x86/kvm/svm/svm.c | 12 +++++-- arch/x86/kvm/vmx/nested.c | 14 ++++---- arch/x86/kvm/vmx/vmx.c | 6 ++-- arch/x86/kvm/x86.c | 19 ++++++---- arch/x86/kvm/x86.h | 8 ++--- 7 files changed, 48 insertions(+), 98 deletions(-)
Queued, thanks.
Paolo