Received-SPF: pass (google.com: domain of linux-kernel+bounces-207934-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Precedence: bulk
MIME-Version: 1.0
References: <20240530210714.364118-1-rick.p.edgecombe@intel.com>
 <20240530210714.364118-10-rick.p.edgecombe@intel.com> <CABgObfbzjLtzFX9wC_FU2GKGF_Wq8og+O=pSnG_yD8j1Dn3jAg@mail.gmail.com>
 <b1306914ee4ca844f9963fcd77b8bf9a30d05249.camel@intel.com>
 <CABgObfb1L4SLGLOPwUKTBusN9bVKACJp7cyvgL8LPhGz0QTNAA@mail.gmail.com> <9c5f7aae312325c0e880baf411f956d4cce3c6d1.camel@intel.com>
In-Reply-To: <9c5f7aae312325c0e880baf411f956d4cce3c6d1.camel@intel.com>
From: Paolo Bonzini <pbonzini@redhat.com>
Date: Mon, 10 Jun 2024 11:23:43 +0200
Message-ID: <CABgObfYd4TWq4ObUzkDruj_e111cTniWtXckzB_Ft7SOdv7YMQ@mail.gmail.com>
Subject: Re: [PATCH v2 09/15] KVM: x86/tdp_mmu: Support mirror root for TDP MMU
To: "Edgecombe, Rick P" <rick.p.edgecombe@intel.com>
Cc: "seanjc@google.com" <seanjc@google.com>, "Huang, Kai" <kai.huang@intel.com>, 
	"sagis@google.com" <sagis@google.com>, 
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>, "Zhao, Yan Y" <yan.y.zhao@intel.com>, 
	"Aktas, Erdem" <erdemaktas@google.com>, "kvm@vger.kernel.org" <kvm@vger.kernel.org>, 
	"dmatlack@google.com" <dmatlack@google.com>, "Yamahata, Isaku" <isaku.yamahata@intel.com>, 
	"isaku.yamahata@gmail.com" <isaku.yamahata@gmail.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Jun 10, 2024 at 2:09=E2=80=AFAM Edgecombe, Rick P
<rick.p.edgecombe@intel.com> wrote:
> Agreed that this is less TDX specific and it means that this part of the =
generic
> MMU code doesn't need to know that the mirror/direct matches to private v=
s
> shared. I don't love that it has such a complicated conditional for the n=
ormal
> VM case, though. Just for readability.
>
> The previous versions checked kvm_gfn_shared_mask() more readily in vario=
us open
> coded spots. In this v2 we tried to reduce this and instead always rely o=
n
> the "private" concept to switch between the roots in the generic code. I =
think
> it's arguably a little easier to understand if we stick to a single way o=
f
> deciding which root to use.

But there isn't any other place that relies on is_private, right? So...

> But I don't feel like any of these solutions discussed is perfectly clean=
. So
> I'm ok taking the benefits you prefer. I guess doing bitwise operations w=
hen
> possible is kind of the KVM way, haha. :)

... while I'm definitely guilty of that, :) it does seem the cleanest
option to use fault->addr to go from struct kvm_page_fault to the kind
of root.

If you prefer, you can introduce a bool kvm_is_addr_direct(struct kvm
*kvm, gpa_t gpa), and use it here as kvm_is_addr_direct(kvm,
fault->addr). Maybe that's the best of both worlds.

Paolo