Re: [PATCH v9 5/6] KVM: x86: Untag address when LAM applicable

[Sean Christopherson]

On Tue, Jun 06, 2023, Binbin Wu wrote:
> Untag address for 64-bit memory/MMIO operand in instruction emulations
> and VMExit handlers when LAM is applicable.
> 
> For instruction emulation, untag address in __linearize() before
> canonical check. LAM doesn't apply to addresses used for instruction
> fetches or to those that specify the targets of jump and call instructions,
> use X86EMUL_F_SKIPLAM to skip LAM untag.
> 
> For VMExit handlers related to 64-bit linear address:
> - Cases need to untag address
>   Operand(s) of VMX instructions and INVPCID.
>   Operand(s) of SGX ENCLS.
> - Cases LAM doesn't apply to
>   Operand of INVLPG.
>   Linear address in INVPCID descriptor (no change needed).
>   Linear address in INVVPID descriptor (it has been confirmed, although it is
>   not called out in LAM spec, no change needed).
>   BASEADDR specified in SESC of ECREATE (no change needed).
> 
> Note:
> LAM doesn't apply to the writes to control registers or MSRs.
> LAM masking applies before paging, so the faulting linear address in CR2
> doesn't contain the metadata.
> The guest linear address saved in VMCS doesn't contain metadata.
> 
> Co-developed-by: Robert Hoo <robert.hu@xxxxxxxxxxxxxxx>
> Signed-off-by: Robert Hoo <robert.hu@xxxxxxxxxxxxxxx>
> Signed-off-by: Binbin Wu <binbin.wu@xxxxxxxxxxxxxxx>
> Reviewed-by: Chao Gao <chao.gao@xxxxxxxxx>
> Tested-by: Xuelian Guo <xuelian.guo@xxxxxxxxx>
> ---
>  arch/x86/kvm/emulate.c     | 16 +++++++++++++---
>  arch/x86/kvm/kvm_emulate.h |  2 ++
>  arch/x86/kvm/vmx/nested.c  |  2 ++
>  arch/x86/kvm/vmx/sgx.c     |  1 +
>  arch/x86/kvm/x86.c         |  7 +++++++
>  5 files changed, 25 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
> index e89afc39e56f..c135adb26f1e 100644
> --- a/arch/x86/kvm/emulate.c
> +++ b/arch/x86/kvm/emulate.c
> @@ -701,6 +701,7 @@ static __always_inline int __linearize(struct x86_emulate_ctxt *ctxt,
>  	*max_size = 0;
>  	switch (mode) {
>  	case X86EMUL_MODE_PROT64:
> +		ctxt->ops->untag_addr(ctxt, &la, flags);
>  		*linear = la;

Ha!  Returning the untagged address does help:

		*linear = ctx->ops->get_untagged_address(ctxt, la, flags);

>  		va_bits = ctxt_virt_addr_bits(ctxt);
>  		if (!__is_canonical_address(la, va_bits))
> @@ -771,8 +772,12 @@ static inline int assign_eip(struct x86_emulate_ctxt *ctxt, ulong dst)
>  
>  	if (ctxt->op_bytes != sizeof(unsigned long))
>  		addr.ea = dst & ((1UL << (ctxt->op_bytes << 3)) - 1);
> +	/*
> +	 * LAM doesn't apply to addresses that specify the targets of jump and
> +	 * call instructions.
> +	 */
>  	rc = __linearize(ctxt, addr, &max_size, 1, ctxt->mode, &linear,
> -			 X86EMUL_F_FETCH);
> +			 X86EMUL_F_FETCH | X86EMUL_F_SKIPLAM);

No need for anything LAM specific here, just skip all FETCH access (unlike LASS
which skips checks only for branch targets).

> -	rc = linearize(ctxt, ctxt->src.addr.mem, 1, false, &linear);
> +	/* LAM doesn't apply to invlpg */

Comment unneeded if X86EMUL_F_INVLPG is added.