linux-kernel - [PATCH AUTOSEL for 4.9 067/293] iio: imu: inv_mpu6050: test whoami first and against all known values

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20180409002239.163177-67-alexander.levin@microsoft.com>
Date:   Mon, 9 Apr 2018 00:23:45 +0000
From:   Sasha Levin <Alexander.Levin@...rosoft.com>
To:     "stable@...r.kernel.org" <stable@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
CC:     Jean-Baptiste Maneyrol <JManeyrol@...ensense.com>,
        Jean-Baptiste Maneyrol <jmaneyrol@...ensense.com>,
        Jonathan Cameron <jic23@...nel.org>,
        Sasha Levin <Alexander.Levin@...rosoft.com>
Subject: [PATCH AUTOSEL for 4.9 067/293] iio: imu: inv_mpu6050: test whoami
 first and against all known values

From: Jean-Baptiste Maneyrol <JManeyrol@...ensense.com>

[ Upstream commit 881b556f6cf22d37be9340f293f2db10ce7ae8bf ]

SPI bus is never generating error during transfer, so to check if
a chip is correctly connected on a SPI bus we enforce whoami check
to be correct. In this way we can assure SPI probe is failing if
there is no chip connected.

Note that this is really papering over boards that claim the device
is there when it isn't.  Not a bad thing to cope with, but not
necessarily stable material.

Fixes: cec0154556f8 ("iio: inv_mpu6050: Check WHO_AM_I register on probe")
Signed-off-by: Jean-Baptiste Maneyrol <jmaneyrol@...ensense.com>
Signed-off-by: Jonathan Cameron <jic23@...nel.org>
Signed-off-by: Sasha Levin <alexander.levin@...rosoft.com>
---
 drivers/iio/imu/inv_mpu6050/inv_mpu_core.c | 33 ++++++++++++++++++++++--------
 1 file changed, 24 insertions(+), 9 deletions(-)

diff --git a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
index 5faea370ab57..41c68d4a3f6b 100644
--- a/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
+++ b/drivers/iio/imu/inv_mpu6050/inv_mpu_core.c
@@ -797,27 +797,42 @@ static int inv_check_and_setup_chip(struct inv_mpu6050_state *st)
 {
 	int result;
 	unsigned int regval;
+	int i;
 
 	st->hw  = &hw_info[st->chip_type];
 	st->reg = hw_info[st->chip_type].reg;
 
-	/* reset to make sure previous state are not there */
-	result = regmap_write(st->map, st->reg->pwr_mgmt_1,
-			      INV_MPU6050_BIT_H_RESET);
-	if (result)
-		return result;
-	msleep(INV_MPU6050_POWER_UP_TIME);
-
 	/* check chip self-identification */
 	result = regmap_read(st->map, INV_MPU6050_REG_WHOAMI, &regval);
 	if (result)
 		return result;
 	if (regval != st->hw->whoami) {
-		dev_warn(regmap_get_device(st->map),
-				"whoami mismatch got %#02x expected %#02hhx for %s\n",
+		/* check whoami against all possible values */
+		for (i = 0; i < INV_NUM_PARTS; ++i) {
+			if (regval == hw_info[i].whoami) {
+				dev_warn(regmap_get_device(st->map),
+					"whoami mismatch got %#02x (%s)"
+					"expected %#02hhx (%s)\n",
+					regval, hw_info[i].name,
+					st->hw->whoami, st->hw->name);
+				break;
+			}
+		}
+		if (i >= INV_NUM_PARTS) {
+			dev_err(regmap_get_device(st->map),
+				"invalid whoami %#02x expected %#02hhx (%s)\n",
 				regval, st->hw->whoami, st->hw->name);
+			return -ENODEV;
+		}
 	}
 
+	/* reset to make sure previous state are not there */
+	result = regmap_write(st->map, st->reg->pwr_mgmt_1,
+			      INV_MPU6050_BIT_H_RESET);
+	if (result)
+		return result;
+	msleep(INV_MPU6050_POWER_UP_TIME);
+
 	/*
 	 * toggle power state. After reset, the sleep bit could be on
 	 * or off depending on the OTP settings. Toggling power would
-- 
2.15.1