linux-kernel - [PATCH v2 2/2] platform/chrome: cros_ec_proto: check for missing EC_CMD_HOST_EVENT_GET_WAKE

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: <20200724190841.3112365-2-briannorris@chromium.org>
Date:   Fri, 24 Jul 2020 12:08:41 -0700
From:   Brian Norris <briannorris@...omium.org>
To:     Benson Leung <bleung@...omium.org>,
        Enric Balletbo i Serra <enric.balletbo@...labora.com>,
        Guenter Roeck <groeck@...omium.org>
Cc:     <linux-kernel@...r.kernel.org>,
        Brian Norris <briannorris@...omium.org>
Subject: [PATCH v2 2/2] platform/chrome: cros_ec_proto: check for missing EC_CMD_HOST_EVENT_GET_WAKE_MASK

As with cros_ec_cmd_xfer_status(), etc., it's not enough to simply check
for the return status of send_command() -- that only covers transport or
other similarly-fatal errors. One must also check the ->result field, to
see whether the command really succeeded. If not, we can't use the data
it returns.

The caller of cros_ec_get_host_event_wake_mask() ignores this, and so
for example, on EC's where the command is not implemented, we're using
junk (or in practice, all zeros) for our wake-mask. We should be using a
non-zero default (currently, it's supposed to be all-1's).

Fix this by checking the ->result field and returning -EPROTO for
errors.

I might label this as fixing commit 29d99b966d60 ("cros_ec: Don't signal
wake event for non-wake host events"), except that this fix alone
actually may make things worse, as it now allows for a lot more spurious
wakeups. The patch "platform/chrome: cros_ec_proto: ignore battery/AC
wakeups on old ECs" helps to mitigate this.

Signed-off-by: Brian Norris <briannorris@...omium.org>
---
v2:
 * EOPNOTSUPP, not ENOTSUPP
---
 drivers/platform/chrome/cros_ec_proto.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)

diff --git a/drivers/platform/chrome/cros_ec_proto.c b/drivers/platform/chrome/cros_ec_proto.c
index e93024b55ce8..31ca0af62388 100644
--- a/drivers/platform/chrome/cros_ec_proto.c
+++ b/drivers/platform/chrome/cros_ec_proto.c
@@ -208,6 +208,12 @@ static int cros_ec_get_host_event_wake_mask(struct cros_ec_device *ec_dev,
 	msg->insize = sizeof(*r);

 	ret = send_command(ec_dev, msg);
+	if (ret >= 0) {
+		if (msg->result == EC_RES_INVALID_COMMAND)
+			return -EOPNOTSUPP;
+		if (msg->result != EC_RES_SUCCESS)
+			return -EPROTO;
+	}
 	if (ret > 0) {
 		r = (struct ec_response_host_event_mask *)msg->data;
 		*mask = r->mask;
@@ -488,6 +494,13 @@ int cros_ec_query_all(struct cros_ec_device *ec_dev)
 			  BIT(EC_HOST_EVENT_BATTERY_CRITICAL) |
 			  BIT(EC_HOST_EVENT_PD_MCU) |
 			  BIT(EC_HOST_EVENT_BATTERY_STATUS));
+		/*
+		 * Old ECs may not support this command. Complain about all
+		 * other errors.
+		 */
+		if (ret != -EOPNOTSUPP)
+			dev_err(ec_dev->dev,
+				"failed to retrieve wake mask: %d\n", ret);
 	}

 	ret = 0;
-- 
2.28.0.rc0.142.g3c755180ce-goog