stable.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	stable@vger.kernel.org,
	Demi Marie Obenour <demi@invisiblethingslab.com>,
	Juergen Gross <jgross@suse.com>, Jan Beulich <jbeulich@suse.com>
Subject: [PATCH 4.9 38/38] xen/netfront: react properly to failing gnttab_end_foreign_access_ref()
Date: Thu, 10 Mar 2022 15:13:51 +0100	[thread overview]
Message-ID: <20220310140809.242977132@linuxfoundation.org> (raw)
In-Reply-To: <20220310140808.136149678@linuxfoundation.org>

From: Juergen Gross <jgross@suse.com>

Commit 66e3531b33ee51dad17c463b4d9c9f52e341503d upstream.

When calling gnttab_end_foreign_access_ref() the returned value must
be tested and the reaction to that value should be appropriate.

In case of failure in xennet_get_responses() the reaction should not be
to crash the system, but to disable the network device.

The calls in setup_netfront() can be replaced by calls of
gnttab_end_foreign_access(). While at it avoid double free of ring
pages and grant references via xennet_disconnect_backend() in this case.

This is CVE-2022-23042 / part of XSA-396.

Reported-by: Demi Marie Obenour <demi@invisiblethingslab.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 drivers/net/xen-netfront.c |   48 +++++++++++++++++++++++++++++----------------
 1 file changed, 31 insertions(+), 17 deletions(-)

--- a/drivers/net/xen-netfront.c
+++ b/drivers/net/xen-netfront.c
@@ -838,7 +838,6 @@ static int xennet_get_responses(struct n
 	int max = XEN_NETIF_NR_SLOTS_MIN + (rx->status <= RX_COPY_THRESHOLD);
 	int slots = 1;
 	int err = 0;
-	unsigned long ret;
 
 	if (rx->flags & XEN_NETRXF_extra_info) {
 		err = xennet_get_extras(queue, extras, rp);
@@ -869,8 +868,13 @@ static int xennet_get_responses(struct n
 			goto next;
 		}
 
-		ret = gnttab_end_foreign_access_ref(ref, 0);
-		BUG_ON(!ret);
+		if (!gnttab_end_foreign_access_ref(ref, 0)) {
+			dev_alert(dev,
+				  "Grant still in use by backend domain\n");
+			queue->info->broken = true;
+			dev_alert(dev, "Disabled for further use\n");
+			return -EINVAL;
+		}
 
 		gnttab_release_grant_reference(&queue->gref_rx_head, ref);
 
@@ -1074,6 +1078,10 @@ static int xennet_poll(struct napi_struc
 		err = xennet_get_responses(queue, &rinfo, rp, &tmpq);
 
 		if (unlikely(err)) {
+			if (queue->info->broken) {
+				spin_unlock(&queue->rx_lock);
+				return 0;
+			}
 err:
 			while ((skb = __skb_dequeue(&tmpq)))
 				__skb_queue_tail(&errq, skb);
@@ -1671,7 +1679,7 @@ static int setup_netfront(struct xenbus_
 			struct netfront_queue *queue, unsigned int feature_split_evtchn)
 {
 	struct xen_netif_tx_sring *txs;
-	struct xen_netif_rx_sring *rxs;
+	struct xen_netif_rx_sring *rxs = NULL;
 	grant_ref_t gref;
 	int err;
 
@@ -1691,21 +1699,21 @@ static int setup_netfront(struct xenbus_
 
 	err = xenbus_grant_ring(dev, txs, 1, &gref);
 	if (err < 0)
-		goto grant_tx_ring_fail;
+		goto fail;
 	queue->tx_ring_ref = gref;
 
 	rxs = (struct xen_netif_rx_sring *)get_zeroed_page(GFP_NOIO | __GFP_HIGH);
 	if (!rxs) {
 		err = -ENOMEM;
 		xenbus_dev_fatal(dev, err, "allocating rx ring page");
-		goto alloc_rx_ring_fail;
+		goto fail;
 	}
 	SHARED_RING_INIT(rxs);
 	FRONT_RING_INIT(&queue->rx, rxs, XEN_PAGE_SIZE);
 
 	err = xenbus_grant_ring(dev, rxs, 1, &gref);
 	if (err < 0)
-		goto grant_rx_ring_fail;
+		goto fail;
 	queue->rx_ring_ref = gref;
 
 	if (feature_split_evtchn)
@@ -1718,22 +1726,28 @@ static int setup_netfront(struct xenbus_
 		err = setup_netfront_single(queue);
 
 	if (err)
-		goto alloc_evtchn_fail;
+		goto fail;
 
 	return 0;
 
 	/* If we fail to setup netfront, it is safe to just revoke access to
 	 * granted pages because backend is not accessing it at this point.
 	 */
-alloc_evtchn_fail:
-	gnttab_end_foreign_access_ref(queue->rx_ring_ref, 0);
-grant_rx_ring_fail:
-	free_page((unsigned long)rxs);
-alloc_rx_ring_fail:
-	gnttab_end_foreign_access_ref(queue->tx_ring_ref, 0);
-grant_tx_ring_fail:
-	free_page((unsigned long)txs);
-fail:
+ fail:
+	if (queue->rx_ring_ref != GRANT_INVALID_REF) {
+		gnttab_end_foreign_access(queue->rx_ring_ref, 0,
+					  (unsigned long)rxs);
+		queue->rx_ring_ref = GRANT_INVALID_REF;
+	} else {
+		free_page((unsigned long)rxs);
+	}
+	if (queue->tx_ring_ref != GRANT_INVALID_REF) {
+		gnttab_end_foreign_access(queue->tx_ring_ref, 0,
+					  (unsigned long)txs);
+		queue->tx_ring_ref = GRANT_INVALID_REF;
+	} else {
+		free_page((unsigned long)txs);
+	}
 	return err;
 }
 



  parent reply	other threads:[~2022-03-10 14:20 UTC|newest]

Thread overview: 45+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-03-10 14:13 [PATCH 4.9 00/38] 4.9.306-rc2 review Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 01/38] x86/speculation: Add RETPOLINE_AMD support to the inline asm CALL_NOSPEC variant Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 02/38] x86/retpoline: Make CONFIG_RETPOLINE depend on compiler support Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 03/38] x86/retpoline: Remove minimal retpoline support Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 04/38] Documentation: Add section about CPU vulnerabilities for Spectre Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 05/38] Documentation: Add swapgs description to the Spectre v1 documentation Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 06/38] Documentation: refer to config RANDOMIZE_BASE for kernel address-space randomization Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 07/38] x86/speculation: Merge one test in spectre_v2_user_select_mitigation() Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 08/38] x86,bugs: Unconditionally allow spectre_v2=retpoline,amd Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 09/38] x86/speculation: Rename RETPOLINE_AMD to RETPOLINE_LFENCE Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 10/38] x86/speculation: Add eIBRS + Retpoline options Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 11/38] Documentation/hw-vuln: Update spectre doc Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 12/38] x86/speculation: Include unprivileged eBPF status in Spectre v2 mitigation reporting Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 13/38] x86/speculation: Use generic retpoline by default on AMD Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 14/38] x86/speculation: Update link to AMD speculation whitepaper Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 15/38] x86/speculation: Warn about Spectre v2 LFENCE mitigation Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 16/38] x86/speculation: Warn about eIBRS + LFENCE + Unprivileged eBPF + SMT Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 17/38] arm/arm64: Provide a wrapper for SMCCC 1.1 calls Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 18/38] arm/arm64: smccc/psci: add arm_smccc_1_1_get_conduit() Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 19/38] ARM: report Spectre v2 status through sysfs Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 20/38] ARM: early traps initialisation Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 21/38] ARM: use LOADADDR() to get load address of sections Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 22/38] ARM: Spectre-BHB workaround Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 23/38] ARM: include unprivileged BPF status in Spectre V2 reporting Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 24/38] ARM: fix build error when BPF_SYSCALL is disabled Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 25/38] ARM: fix co-processor register typo Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 26/38] ARM: Do not use NOCROSSREFS directive with ld.lld Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 27/38] x86/build: Fix compiler support check for CONFIG_RETPOLINE Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 28/38] x86, modpost: Replace last remnants of RETPOLINE with CONFIG_RETPOLINE Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 29/38] ARM: fix build warning in proc-v7-bugs.c Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 30/38] xen/xenbus: dont let xenbus_grant_ring() remove grants in error case Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 31/38] xen/grant-table: add gnttab_try_end_foreign_access() Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 32/38] xen/blkfront: dont use gnttab_query_foreign_access() for mapped status Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 33/38] xen/netfront: " Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 34/38] xen/scsifront: " Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 35/38] xen/gntalloc: dont use gnttab_query_foreign_access() Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 36/38] xen: remove gnttab_query_foreign_access() Greg Kroah-Hartman
2022-03-10 14:13 ` [PATCH 4.9 37/38] xen/gnttab: fix gnttab_end_foreign_access() without page specified Greg Kroah-Hartman
2022-03-10 14:13 ` Greg Kroah-Hartman [this message]
2022-03-10 18:00 ` [PATCH 4.9 00/38] 4.9.306-rc2 review Pavel Machek
2022-03-10 18:48 ` Jon Hunter
2022-03-10 19:35 ` Shuah Khan
2022-03-10 20:14 ` Florian Fainelli
2022-03-11  1:13 ` Guenter Roeck
2022-03-11 13:35 ` Naresh Kamboju

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=20220310140809.242977132@linuxfoundation.org \
    --to=gregkh@linuxfoundation.org \
    --cc=demi@invisiblethingslab.com \
    --cc=jbeulich@suse.com \
    --cc=jgross@suse.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.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).