linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Quentin Perret <qperret@google.com>
To: Marc Zyngier <maz@kernel.org>, James Morse <james.morse@arm.com>,
	 Alexandru Elisei <alexandru.elisei@arm.com>,
	Suzuki K Poulose <suzuki.poulose@arm.com>,
	 Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will@kernel.org>, Fuad Tabba <tabba@google.com>,
	David Brazdil <dbrazdil@google.com>
Cc: linux-arm-kernel@lists.infradead.org,
	kvmarm@lists.cs.columbia.edu,  linux-kernel@vger.kernel.org,
	kernel-team@android.com
Subject: [PATCH 13/16] KVM: arm64: Move double-sharing logic into hyp-specific function
Date: Wed, 13 Oct 2021 16:58:28 +0100	[thread overview]
Message-ID: <20211013155831.943476-14-qperret@google.com> (raw)
In-Reply-To: <20211013155831.943476-1-qperret@google.com>

From: Will Deacon <will@kernel.org>

Strictly speaking, double-sharing a page is an invalid transition and
should be rejected, however we allow this in order to simplify the
book-keeping when KVM metadata (such as vcpu structures) co-exists in
the same page.

Given that double-sharing is only required for pages shared with the
hypervisor by the host, move the handling into a hyp-specific function
to check incoming shares, therefore preventing double-sharing outside
of this particular transition.

Signed-off-by: Will Deacon <will@kernel.org>
Signed-off-by: Quentin Perret <qperret@google.com>
---
 arch/arm64/kvm/hyp/nvhe/mem_protect.c | 57 +++++++++++++++++++--------
 1 file changed, 41 insertions(+), 16 deletions(-)

diff --git a/arch/arm64/kvm/hyp/nvhe/mem_protect.c b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
index 909e60f71b06..3378117d010c 100644
--- a/arch/arm64/kvm/hyp/nvhe/mem_protect.c
+++ b/arch/arm64/kvm/hyp/nvhe/mem_protect.c
@@ -536,6 +536,33 @@ static int ack_share(struct pkvm_page_share_ack *ack,
 	}
 }
 
+static int hyp_check_incoming_share(struct pkvm_page_req *req,
+				    struct pkvm_page_share_ack *ack,
+				    enum pkvm_component_id initiator,
+				    enum kvm_pgtable_prot prot)
+{
+	/*
+	 * We allow the host to share the same page twice, but that means we
+	 * have to check that the states really do match exactly.
+	 */
+	if (initiator != PKVM_ID_HOST)
+		return -EPERM;
+
+	if (req->initiator.state != PKVM_PAGE_SHARED_OWNED)
+		return -EPERM;
+
+	if (ack->completer.state != PKVM_PAGE_SHARED_BORROWED)
+		return -EPERM;
+
+	if (ack->completer.phys != req->phys)
+		return -EPERM;
+
+	if (ack->completer.prot != prot)
+		return -EPERM;
+
+	return 0;
+}
+
 /*
  * Check that the page states in the initiator and the completer are compatible
  * for the requested page-sharing operation to go ahead.
@@ -544,6 +571,8 @@ static int check_share(struct pkvm_page_req *req,
 		       struct pkvm_page_share_ack *ack,
 		       struct pkvm_mem_share *share)
 {
+	struct pkvm_mem_transition *tx = &share->tx;
+
 	if (!addr_is_memory(req->phys))
 		return -EINVAL;
 
@@ -552,25 +581,22 @@ static int check_share(struct pkvm_page_req *req,
 		return 0;
 	}
 
-	if (req->initiator.state != PKVM_PAGE_SHARED_OWNED)
-		return -EPERM;
-
-	if (ack->completer.state != PKVM_PAGE_SHARED_BORROWED)
-		return -EPERM;
-
-	if (ack->completer.phys != req->phys)
-		return -EPERM;
-
-	if (ack->completer.prot != share->prot)
+	switch (tx->completer.id) {
+	case PKVM_ID_HYP:
+		return hyp_check_incoming_share(req, ack, tx->initiator.id,
+						share->prot);
+	default:
 		return -EPERM;
-
-	return 0;
+	}
 }
 
 static int host_initiate_share(struct pkvm_page_req *req)
 {
 	enum kvm_pgtable_prot prot;
 
+	if (req->initiator.state == PKVM_PAGE_SHARED_OWNED)
+		return 0;
+
 	prot = pkvm_mkstate(PKVM_HOST_MEM_PROT, PKVM_PAGE_SHARED_OWNED);
 	return host_stage2_idmap_locked(req->initiator.addr, PAGE_SIZE, prot);
 }
@@ -595,6 +621,9 @@ static int hyp_complete_share(struct pkvm_page_req *req,
 	void *start = (void *)req->completer.addr, *end = start + PAGE_SIZE;
 	enum kvm_pgtable_prot prot;
 
+	if (req->initiator.state == PKVM_PAGE_SHARED_OWNED)
+		return 0;
+
 	prot = pkvm_mkstate(perms, PKVM_PAGE_SHARED_BORROWED);
 	return pkvm_create_mappings_locked(start, end, prot);
 }
@@ -653,10 +682,6 @@ static int do_share(struct pkvm_mem_share *share)
 		if (ret)
 			break;
 
-		/* Allow double-sharing by skipping over the page */
-		if (req.initiator.state == PKVM_PAGE_SHARED_OWNED)
-			continue;
-
 		ret = initiate_share(&req, share);
 		if (ret)
 			break;
-- 
2.33.0.882.g93a45727a2-goog


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  parent reply	other threads:[~2021-10-13 16:07 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-13 15:58 [PATCH 00/16] KVM: arm64: Implement unshare hypercall for pkvm Quentin Perret
2021-10-13 15:58 ` [PATCH 01/16] KVM: arm64: Introduce do_share() helper for memory sharing between components Quentin Perret
2021-10-15 15:11   ` Andrew Walbran
2021-10-19 10:37     ` Quentin Perret
2021-10-13 15:58 ` [PATCH 02/16] KVM: arm64: Implement __pkvm_host_share_hyp() using do_share() Quentin Perret
2021-10-13 15:58 ` [PATCH 03/16] KVM: arm64: Avoid remapping the SVE state in the hyp stage-1 Quentin Perret
2021-10-16 11:04   ` Marc Zyngier
2021-10-18 10:36     ` Quentin Perret
2021-10-13 15:58 ` [PATCH 04/16] KVM: arm64: Introduce kvm_share_hyp() Quentin Perret
2021-10-17 10:41   ` Marc Zyngier
2021-10-18 10:51     ` Quentin Perret
2021-10-13 15:58 ` [PATCH 05/16] KVM: arm64: Accept page ranges in pkvm share hypercall Quentin Perret
2021-10-13 15:58 ` [PATCH 06/16] KVM: arm64: Provide {get,put}_page() stubs for early hyp allocator Quentin Perret
2021-10-13 15:58 ` [PATCH 07/16] KVM: arm64: Refcount hyp stage-1 pgtable pages Quentin Perret
2021-10-13 15:58 ` [PATCH 08/16] KVM: arm64: Fixup hyp stage-1 refcount Quentin Perret
2021-10-13 15:58 ` [PATCH 09/16] KVM: arm64: Hook up ->page_count() for hypervisor stage-1 page-table Quentin Perret
2021-10-13 15:58 ` [PATCH 10/16] KVM: arm64: Implement kvm_pgtable_hyp_unmap() at EL2 Quentin Perret
2021-10-13 15:58 ` [PATCH 11/16] KVM: arm64: Back hyp_vmemmap for all of memory Quentin Perret
2021-10-13 15:58 ` [PATCH 12/16] KVM: arm64: Move hyp refcount helpers to header files Quentin Perret
2021-10-13 15:58 ` Quentin Perret [this message]
2021-10-13 15:58 ` [PATCH 14/16] KVM: arm64: Refcount shared pages at EL2 Quentin Perret
2021-10-13 15:58 ` [PATCH 15/16] KVM: arm64: pkvm: Introduce an unshare hypercall Quentin Perret
2021-10-13 15:58 ` [PATCH 16/16] KVM: arm64: pkvm: Unshare guest structs during teardown Quentin Perret
2021-10-16 12:25   ` Marc Zyngier
2021-10-18 10:32     ` Quentin Perret
2021-10-18 14:03       ` Quentin Perret
2021-10-18 17:12         ` Marc Zyngier
2021-10-19  9:40           ` Quentin Perret

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=20211013155831.943476-14-qperret@google.com \
    --to=qperret@google.com \
    --cc=alexandru.elisei@arm.com \
    --cc=catalin.marinas@arm.com \
    --cc=dbrazdil@google.com \
    --cc=james.morse@arm.com \
    --cc=kernel-team@android.com \
    --cc=kvmarm@lists.cs.columbia.edu \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=maz@kernel.org \
    --cc=suzuki.poulose@arm.com \
    --cc=tabba@google.com \
    --cc=will@kernel.org \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).