netdev - [PATCH v2 3/3] net: macb: add missing barriers when reading descriptors

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20181217130541.3261-4-anssi.hannula@bitwise.fi>
Date:   Mon, 17 Dec 2018 15:05:41 +0200
From:   Anssi Hannula <anssi.hannula@...wise.fi>
To:     davem@...emloft.net, netdev@...r.kernel.org
Cc:     Nicolas Ferre <nicolas.ferre@...rochip.com>,
        Claudiu.Beznea@...rochip.com
Subject: [PATCH v2 3/3] net: macb: add missing barriers when reading descriptors

When reading buffer descriptors on RX or on TX completion, an
RX_USED/TX_USED bit is checked first to ensure that the descriptors have
been populated, i.e. the ownership has been transferred. However, there
are no memory barriers to ensure that the data protected by the
RX_USED/TX_USED bit is up-to-date with respect to that bit.

Specifically:

- TX timestamp descriptors may be loaded before ctrl is loaded for the
  TX_USED check, which is racy as the descriptors may be updated between
  the loads, causing old timestamp descriptor data to be used.

- RX ctrl may be loaded before addr is loaded for the RX_USED check,
  which is racy as a new frame may be written between the loads, causing
  old ctrl descriptor data to be used.
  This issue exists for both macb_rx() and gem_rx() variants.

Fix the races by adding DMA read memory barriers on those paths and
reordering the reads in macb_rx().

I have not observed any actual problems in practice caused by these
being missing, though.

Tested on a ZynqMP based system.

Signed-off-by: Anssi Hannula <anssi.hannula@...wise.fi>
Fixes: 89e5785fc8a6 ("[PATCH] Atmel MACB ethernet driver")
Cc: Nicolas Ferre <nicolas.ferre@...rochip.com>
---
 drivers/net/ethernet/cadence/macb_main.c | 13 ++++++++++---
 drivers/net/ethernet/cadence/macb_ptp.c  |  2 ++
 2 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
index 430b7a0f5436..10a2bb44612b 100644
--- a/drivers/net/ethernet/cadence/macb_main.c
+++ b/drivers/net/ethernet/cadence/macb_main.c
@@ -1001,11 +1001,15 @@ static int gem_rx(struct macb_queue *queue, int budget)
 
 		rxused = (desc->addr & MACB_BIT(RX_USED)) ? true : false;
 		addr = macb_get_addr(bp, desc);
-		ctrl = desc->ctrl;
 
 		if (!rxused)
 			break;
 
+		/* Ensure ctrl is at least as up-to-date as rxused */
+		dma_rmb();
+
+		ctrl = desc->ctrl;
+
 		queue->rx_tail++;
 		count++;
 
@@ -1180,11 +1184,14 @@ static int macb_rx(struct macb_queue *queue, int budget)
 		/* Make hw descriptor updates visible to CPU */
 		rmb();
 
-		ctrl = desc->ctrl;
-
 		if (!(desc->addr & MACB_BIT(RX_USED)))
 			break;
 
+		/* Ensure ctrl is at least as up-to-date as addr */
+		dma_rmb();
+
+		ctrl = desc->ctrl;
+
 		if (ctrl & MACB_BIT(RX_SOF)) {
 			if (first_frag != -1)
 				discard_partial_frame(queue, first_frag, tail);
diff --git a/drivers/net/ethernet/cadence/macb_ptp.c b/drivers/net/ethernet/cadence/macb_ptp.c
index cd5296b84229..a6dc47edc4cf 100644
--- a/drivers/net/ethernet/cadence/macb_ptp.c
+++ b/drivers/net/ethernet/cadence/macb_ptp.c
@@ -319,6 +319,8 @@ int gem_ptp_txstamp(struct macb_queue *queue, struct sk_buff *skb,
 	desc_ptp = macb_ptp_desc(queue->bp, desc);
 	tx_timestamp = &queue->tx_timestamps[head];
 	tx_timestamp->skb = skb;
+	/* ensure ts_1/ts_2 is loaded after ctrl (TX_USED check) */
+	dma_rmb();
 	tx_timestamp->desc_ptp.ts_1 = desc_ptp->ts_1;
 	tx_timestamp->desc_ptp.ts_2 = desc_ptp->ts_2;
 	/* move head */
-- 
2.17.2