Skip to contentSkip to conversation

net: atm: Fix potential Spectre v1

PATCH In review Gustavo A. R. Silva 2 participants 2 messages May 3, 2018 – May 4, 2018

Subsystems (3)
Networking Network drivers ATM
lore
Share & tools
lore

Gustavo A. R. Silva

author
#0 lore raw

PATCH 1/1 net: atm: Fix potential Spectre v1

ioc_data.dev_num can be controlled by user-space, hence leading to a potential exploitation of the Spectre variant 1 vulnerability.

This issue was detected with the help of Smatch:
net/atm/lec.c:702 lec_vcc_attach() warn: potential spectre issue 'dev_lec'

Fix this by sanitizing ioc_data.dev_num before using it to index dev_lec. Also, notice that there is another instance in which array dev_lec is being indexed using ioc_data.dev_num at line 705:
lec_vcc_added(netdev_priv(dev_lec[ioc_data.dev_num]),

Notice that given that speculation windows are large, the policy is to kill the speculation on the first load and not worry if it can be completed with a dependent load/store [1].

[1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2

Cc [email protected]
Signed-off-by Gustavo A. R. Silva <[email protected]>
---
 net/atm/lec.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
net/atm/lec.c +7 -2
index 01d5d20..3138a86 100644
@@ -41,6 +41,9 @@ static unsigned char bridge_ula_lec[] = { 0x01, 0x80, 0xc2, 0x00, 0x00 };
 #include <linux/module.h>
 #include <linux/init.h>
 
+/* Hardening for Spectre-v1 */
+#include <linux/nospec.h>
+
 #include "lec.h"
 #include "lec_arpc.h"
 #include "resources.h"

@@ -687,8 +690,10 @@ static int lec_vcc_attach(struct atm_vcc *vcc, void __user *arg)
 	bytes_left = copy_from_user(&ioc_data, arg, sizeof(struct atmlec_ioc));
 	if (bytes_left != 0)
 		pr_info("copy from user failed for %d bytes\n", bytes_left);
-	if (ioc_data.dev_num < 0 || ioc_data.dev_num >= MAX_LEC_ITF ||
-	    !dev_lec[ioc_data.dev_num])
+	if (ioc_data.dev_num < 0 || ioc_data.dev_num >= MAX_LEC_ITF)
+		return -EINVAL;
+	ioc_data.dev_num = array_index_nospec(ioc_data.dev_num, MAX_LEC_ITF);
+	if (!dev_lec[ioc_data.dev_num])
 		return -EINVAL;
 	vpriv = kmalloc(sizeof(struct lec_vcc_priv), GFP_KERNEL);
 	if (!vpriv)
Signature · 1 line
2.7.4
maintainer
#1 lore raw

From: "Gustavo A. R. Silva" <[email protected]>
Date: Thu, 3 May 2018 13:45:58 -0500

Quoted 12 lines

ioc_data.dev_num can be controlled by user-space, hence leading to a potential exploitation of the Spectre variant 1 vulnerability.

This issue was detected with the help of Smatch:
net/atm/lec.c:702 lec_vcc_attach() warn: potential spectre issue 'dev_lec'

Fix this by sanitizing ioc_data.dev_num before using it to index dev_lec. Also, notice that there is another instance in which array dev_lec is being indexed using ioc_data.dev_num at line 705:
lec_vcc_added(netdev_priv(dev_lec[ioc_data.dev_num]),

Quoted 8 lines

Notice that given that speculation windows are large, the policy is to kill the speculation on the first load and not worry if it can be completed with a dependent load/store [1].

[1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2

Cc [email protected]
Signed-off-by Gustavo A. R. Silva <[email protected]>

Applied and queued up for -stable, thanks.