All of lore.kernel.org
 help / color / mirror / Atom feed
From: Oliver Upton <oupton@kernel.org>
To: kvmarm@lists.linux.dev
Cc: Marc Zyngier <maz@kernel.org>, Joey Gouly <joey.gouly@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	Zenghui Yu <yuzenghui@huawei.com>,
	Wei-Lin Chang <weilin.chang@arm.com>,
	Steffen Eiden <seiden@linux.ibm.com>,
	Fuad Tabba <fuad.tabba@linux.dev>,
	Leonardo Bras <leo.bras@arm.com>,
	Oliver Upton <oupton@kernel.org>
Subject: [PATCH v2 18/28] KVM: arm64: nv: Emulate CMOs via writable-clean translations as NOPs
Date: Fri, 23 Oct 2026 13:56:00 -0700	[thread overview]
Message-ID: <20261023205610.1044532-19-oupton@kernel.org> (raw)
In-Reply-To: <20261023205610.1044532-1-oupton@kernel.org>

Virtualizing FEAT_HAFDBS for an L1 hypervisor has a pretty large
architectural deficiency by way of cache maintenance instructions.
CMOs (e.g. DC IVAC) require write permission but do not set the dirty
state on the descriptor.

There's no way to shadow a writable-clean translation in such a way that
these CMOs pass the permission check and subsequent writes generate a
permission fault. Worse yet, data aborts taken on CMOs do not have
sufficient syndrome information to perform instruction emulation and
trapping all CMOs is unpalatable.

So what do we do? Absolutely nothing. Treat CMOs to writable-clean
translations as NOPs in anticipation of KVM only exposing FEAT_HAFDBS
when the VM is fully coherent (FEAT_S2FWB, CTR_EL0.{DIC,IDC}={1,1}, no
noncoherent DMA).

Signed-off-by: Oliver Upton <oupton@kernel.org>
---
 arch/arm64/kvm/mmu.c    |  4 ++++
 arch/arm64/kvm/nested.c | 20 ++++++++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/arch/arm64/kvm/mmu.c b/arch/arm64/kvm/mmu.c
index 7b29be33a1ea..a21b674a1981 100644
--- a/arch/arm64/kvm/mmu.c
+++ b/arch/arm64/kvm/mmu.c
@@ -2548,6 +2548,10 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu)
 		}
 
 		ret = kvm_s2_handle_perm_fault(vcpu, &nested_trans);
+		if (ret == -EAGAIN) {
+			ret = 1;
+			goto out_unlock;
+		}
 		if (ret) {
 			esr = nested_trans.esr;
 			kvm_inject_s2_fault(vcpu, esr);
diff --git a/arch/arm64/kvm/nested.c b/arch/arm64/kvm/nested.c
index 7dda95d37925..46fcc75f4c20 100644
--- a/arch/arm64/kvm/nested.c
+++ b/arch/arm64/kvm/nested.c
@@ -1150,6 +1150,26 @@ int kvm_s2_handle_perm_fault(struct kvm_vcpu *vcpu, struct kvm_s2_trans *trans)
 	} else {
 		bool write_fault = kvm_is_write_fault(vcpu);
 
+		/*
+		 * Plug your nose for a moment.
+		 *
+		 * Writable-clean translations are shadowed using a read-only
+		 * mapping so KVM can set the dirty state on the L1 descriptor
+		 * before installing a writable mapping.
+		 *
+		 * Certain cache maintenance instructions require a write
+		 * permission check but leave the translation in a writable-clean
+		 * state, meaning KVM is on the hook to emulate CMOs that use a
+		 * writable-clean translation.
+		 *
+		 * "Emulate" the CMO as a NOP, as KVM only exposes FEAT_HAFDBS
+		 * when FEAT_S2FWB is present and the I-cache is coherent.
+		 */
+		if (kvm_vcpu_dabt_is_cm(vcpu) && trans->writable && !trans->dirty) {
+			kvm_incr_pc(vcpu);
+			return -EAGAIN;
+		}
+
 		forward_fault = ((write_fault && !trans->writable) ||
 				 (!write_fault && !trans->readable));
 	}
-- 
2.47.3


  parent reply	other threads:[~2026-10-23 20:56 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-23 20:55 [PATCH v2 00/28] KVM: arm64: nv: Implement FEAT_HAFDBS, FEAT_HAFT Oliver Upton
2026-10-23 20:55 ` [PATCH v2 01/28] KVM: arm64: nv: Introduce struct for stage-2 walk step Oliver Upton
2026-10-23 20:55 ` [PATCH v2 02/28] KVM: arm64: nv: Consolidate computation of stage-2 permissions Oliver Upton
2026-10-23 20:55 ` [PATCH v2 03/28] KVM: arm64: nv: Get rid of kvm_s2_trans*() accessors Oliver Upton
2026-10-23 20:55 ` [PATCH v2 04/28] KVM: arm64: nv: Only shadow writable-dirty guest descs as writable Oliver Upton
2026-10-23 21:25   ` sashiko-bot
2026-10-24  6:26     ` Oliver Upton
2026-10-23 20:55 ` [PATCH v2 05/28] KVM: arm64: nv: Pass an access descriptor for stage-2 walks Oliver Upton
2026-10-23 20:55 ` [PATCH v2 06/28] KVM: arm64: nv: Use a helper for stage-2 descriptor updates Oliver Upton
2026-10-23 20:55 ` [PATCH v2 07/28] KVM: arm64: nv: Set dirty state at stage-2 Oliver Upton
2026-10-23 20:55 ` [PATCH v2 08/28] KVM: arm64: nv: Treat DBM as writable " Oliver Upton
2026-10-23 20:55 ` [PATCH v2 09/28] KVM: arm64: Compute S1 permissions as part of s1_walk() Oliver Upton
2026-10-23 20:55 ` [PATCH v2 10/28] KVM: arm64: Plumb through access descriptor for stage-1 Oliver Upton
2026-10-23 21:06   ` sashiko-bot
2026-10-23 20:55 ` [PATCH v2 11/28] KVM: arm64: Use a struct for stage-1 walk context Oliver Upton
2026-10-23 20:55 ` [PATCH v2 12/28] KVM: arm64: Create helper for stage-1 descriptor updates Oliver Upton
2026-10-23 20:55 ` [PATCH v2 13/28] KVM: arm64: Set dirty state at stage-1 Oliver Upton
2026-10-23 20:55 ` [PATCH v2 14/28] KVM: arm64: nv: Respect HA when computing translated TCR Oliver Upton
2026-10-23 20:55 ` [PATCH v2 15/28] KVM: arm64: nv: Respect HD " Oliver Upton
2026-10-23 20:55 ` [PATCH v2 16/28] KVM: arm64: Grant write permission when DBM is set at S1 Oliver Upton
2026-10-23 20:55 ` [PATCH v2 17/28] KVM: arm64: Don't update descriptors for "non-arch" access Oliver Upton
2026-10-23 20:56 ` Oliver Upton [this message]
2026-10-23 20:56 ` [PATCH v2 19/28] arm64: Add command line switch to disable FEAT_S2FWB Oliver Upton
2026-10-23 20:56 ` [PATCH v2 20/28] KVM: arm64: Reject non-coherent DMA when using forced writeback Oliver Upton
2026-10-23 21:16   ` sashiko-bot
2026-10-23 20:56 ` [PATCH v2 21/28] KVM: arm64: nv: Expose FEAT_HAFDBS Oliver Upton
2026-10-23 20:56 ` [PATCH v2 22/28] KVM: arm64: Set Access flag on table descriptors at stage-1 Oliver Upton
2026-10-23 21:22   ` sashiko-bot
2026-10-24  6:24     ` Oliver Upton
2026-10-23 20:56 ` [PATCH v2 23/28] KVM: arm64: nv: Set access flag on table descriptors at stage-2 Oliver Upton
2026-10-23 20:56 ` [PATCH v2 24/28] KVM: arm64: nv: Expose FEAT_HAFT Oliver Upton
2026-10-23 20:56 ` [PATCH v2 25/28] KVM: arm64: selftests: Only test AF behavior for emulated AT insns Oliver Upton
2026-10-23 20:56 ` [PATCH v2 26/28] KVM: arm64: selftests: Test AT emulation for FEAT_HAFT Oliver Upton
2026-10-23 20:56 ` [PATCH v2 27/28] KVM: arm64: Add tracepoints for S1 page table walk Oliver Upton
2026-10-23 20:56 ` [PATCH v2 28/28] KVM: arm64: Add tracepoints for S2 " Oliver Upton

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20261023205610.1044532-19-oupton@kernel.org \
    --to=oupton@kernel.org \
    --cc=fuad.tabba@linux.dev \
    --cc=joey.gouly@arm.com \
    --cc=kvmarm@lists.linux.dev \
    --cc=leo.bras@arm.com \
    --cc=maz@kernel.org \
    --cc=seiden@linux.ibm.com \
    --cc=suzuki.poulose@arm.com \
    --cc=weilin.chang@arm.com \
    --cc=yuzenghui@huawei.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.