Received-SPF: pass (google.com: domain of linux-kernel+bounces-104740-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Date: Fri, 15 Mar 2024 10:45:04 -0700
In-Reply-To: <88920c598dcb55c15219642f27d0781af6d0c044.1708933498.git.isaku.yamahata@intel.com>
Precedence: bulk
Mime-Version: 1.0
References: <cover.1708933498.git.isaku.yamahata@intel.com> <88920c598dcb55c15219642f27d0781af6d0c044.1708933498.git.isaku.yamahata@intel.com>
Message-ID: <ZfSJIDOJzGJ4lPjX@google.com>
Subject: Re: [PATCH v19 098/130] KVM: TDX: Add a place holder to handle TDX VM exit
From: Sean Christopherson <seanjc@google.com>
To: isaku.yamahata@intel.com
Cc: kvm@vger.kernel.org, linux-kernel@vger.kernel.org, 
	isaku.yamahata@gmail.com, Paolo Bonzini <pbonzini@redhat.com>, erdemaktas@google.com, 
	Sagi Shahar <sagis@google.com>, Kai Huang <kai.huang@intel.com>, chen.bo@intel.com, 
	hang.yuan@intel.com, tina.zhang@intel.com
Content-Type: text/plain; charset="us-ascii"

On Mon, Feb 26, 2024, isaku.yamahata@intel.com wrote:
> +int tdx_handle_exit(struct kvm_vcpu *vcpu, fastpath_t fastpath)
> +{
> +	union tdx_exit_reason exit_reason = to_tdx(vcpu)->exit_reason;
> +
> +	/* See the comment of tdh_sept_seamcall(). */
> +	if (unlikely(exit_reason.full == (TDX_OPERAND_BUSY | TDX_OPERAND_ID_SEPT)))
> +		return 1;
> +
> +	/*
> +	 * TDH.VP.ENTRY checks TD EPOCH which contend with TDH.MEM.TRACK and
> +	 * vcpu TDH.VP.ENTER.
> +	 */
> +	if (unlikely(exit_reason.full == (TDX_OPERAND_BUSY | TDX_OPERAND_ID_TD_EPOCH)))
> +		return 1;
> +
> +	if (unlikely(exit_reason.full == TDX_SEAMCALL_UD)) {
> +		kvm_spurious_fault();
> +		/*
> +		 * In the case of reboot or kexec, loop with TDH.VP.ENTER and
> +		 * TDX_SEAMCALL_UD to avoid unnecessarily activity.
> +		 */
> +		return 1;

No.  This is unnecessarily risky.  KVM_BUG_ON() and exit to userspace.  The
response to "SEAMCALL faulted" should never be, "well, let's try again!".

Also, what about #GP on SEAMCALL?  In general, the error handling here seems
lacking.