linux-kernel - [PATCH 4.19 33/41] Input: matrix_keypad - check for errors from of_get_named

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: <20181206142952.905396475@linuxfoundation.org>
Date:   Thu,  6 Dec 2018 15:39:13 +0100
From:   Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To:     linux-kernel@...r.kernel.org
Cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        stable@...r.kernel.org, Christian Hoff <christian_hoff@....net>,
        Sebastian Reichel <sre@...nel.org>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>
Subject: [PATCH 4.19 33/41] Input: matrix_keypad - check for errors from of_get_named_gpio()

4.19-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Christian Hoff <christian_hoff@....net>

commit d55bda1b3e7c5a87f10da54fdda866a9a9cef30b upstream.

"of_get_named_gpio()" returns a negative error value if it fails
and drivers should check for this. This missing check was now
added to the matrix_keypad driver.

In my case "of_get_named_gpio()" returned -EPROBE_DEFER because
the referenced GPIOs belong to an I/O expander, which was not yet
probed at the point in time when the matrix_keypad driver was
loading. Because the driver did not check for errors from the
"of_get_named_gpio()" routine, it was assuming that "-EPROBE_DEFER"
is actually a GPIO number and continued as usual, which led to further
errors like this later on:

WARNING: CPU: 3 PID: 167 at drivers/gpio/gpiolib.c:114
gpio_to_desc+0xc8/0xd0
invalid GPIO -517

Note that the "GPIO number" -517 in the error message above is
actually "-EPROBE_DEFER".

As part of the patch a misleading error message "no platform data defined"
was also removed. This does not lead to information loss because the other
error paths in matrix_keypad_parse_dt() already print an error.

Signed-off-by: Christian Hoff <christian_hoff@....net>
Suggested-by: Sebastian Reichel <sre@...nel.org>
Reviewed-by: Sebastian Reichel <sre@...nel.org>
Cc: stable@...r.kernel.org
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@...il.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@...uxfoundation.org>

---
 drivers/input/keyboard/matrix_keypad.c |   23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

--- a/drivers/input/keyboard/matrix_keypad.c
+++ b/drivers/input/keyboard/matrix_keypad.c
@@ -407,7 +407,7 @@ matrix_keypad_parse_dt(struct device *de
 	struct matrix_keypad_platform_data *pdata;
 	struct device_node *np = dev->of_node;
 	unsigned int *gpios;
-	int i, nrow, ncol;
+	int ret, i, nrow, ncol;
 
 	if (!np) {
 		dev_err(dev, "device lacks DT data\n");
@@ -452,12 +452,19 @@ matrix_keypad_parse_dt(struct device *de
 		return ERR_PTR(-ENOMEM);
 	}
 
-	for (i = 0; i < pdata->num_row_gpios; i++)
-		gpios[i] = of_get_named_gpio(np, "row-gpios", i);
+	for (i = 0; i < nrow; i++) {
+		ret = of_get_named_gpio(np, "row-gpios", i);
+		if (ret < 0)
+			return ERR_PTR(ret);
+		gpios[i] = ret;
+	}
 
-	for (i = 0; i < pdata->num_col_gpios; i++)
-		gpios[pdata->num_row_gpios + i] =
-			of_get_named_gpio(np, "col-gpios", i);
+	for (i = 0; i < ncol; i++) {
+		ret = of_get_named_gpio(np, "col-gpios", i);
+		if (ret < 0)
+			return ERR_PTR(ret);
+		gpios[nrow + i] = ret;
+	}
 
 	pdata->row_gpios = gpios;
 	pdata->col_gpios = &gpios[pdata->num_row_gpios];
@@ -484,10 +491,8 @@ static int matrix_keypad_probe(struct pl
 	pdata = dev_get_platdata(&pdev->dev);
 	if (!pdata) {
 		pdata = matrix_keypad_parse_dt(&pdev->dev);
-		if (IS_ERR(pdata)) {
-			dev_err(&pdev->dev, "no platform data defined\n");
+		if (IS_ERR(pdata))
 			return PTR_ERR(pdata);
-		}
 	} else if (!pdata->keymap_data) {
 		dev_err(&pdev->dev, "no keymap data defined\n");
 		return -EINVAL;