linux-kernel - [PATCH] gpiolib: fix oops on reading sysfs exported GPIOs

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-next>] [day] [month] [year] [list]

Date:	Thu, 16 Oct 2008 18:45:03 +0400
From:	Anton Vorontsov <avorontsov@...mvista.com>
To:	David Brownell <dbrownell@...rs.sourceforge.net>
Cc:	Andrew Morton <akpm@...ux-foundation.org>,
	linux-kernel@...r.kernel.org
Subject: [PATCH] gpiolib: fix oops on reading sysfs exported GPIOs

We can get the following oops when a GPIO controller doesn't provide
.direction_input and .get callbacks:

root@b1:~# cat /sys/class/gpio/gpio255/value
Unable to handle kernel paging request for instruction fetch
Faulting instruction address: 0x00000000
Oops: Kernel access of bad area, sig: 11 [#1]
[...]
NIP [00000000] 0x0
LR [c0182fb0] gpio_get_value_cansleep+0x40/0x50
Call Trace:
[c7b79e80] [c0183f28] gpio_value_show+0x5c/0x94
[c7b79ea0] [c01a584c] dev_attr_show+0x30/0x7c
[c7b79eb0] [c00d6b48] fill_read_buffer+0x68/0xe0
[c7b79ed0] [c00d6c54] sysfs_read_file+0x94/0xbc
[c7b79ef0] [c008f24c] vfs_read+0xb4/0x16c
[c7b79f10] [c008f580] sys_read+0x4c/0x90
[c7b79f40] [c0013a14] ret_from_syscall+0x0/0x38

GPIO users should always issue the gpio_direction_input() call and
check its return value prior to trying gpio_get_value().

For sysfs users there are few ways to solve the problem:

1. Call gpio_direction_input() in the gpio_value_show(). This isn't
   good because some GPIO controllers provide capability to read-back
   output pins. Using the FLAG_IS_OUT isn't good for the same reason.

2. Call gpio_direction_input() at the export time, if succeeded, set
   FLAG_CAN_INPUT. Then check that flag in the gpio_value_show().
   Viable.

3. Just check for .get != NULL in the gpio_value_show(). Most
   straightforward. This is implemented in the patch.

Signed-off-by: Anton Vorontsov <avorontsov@...mvista.com>
---
 drivers/gpio/gpiolib.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c
index 8d29405..4e4a498 100644
--- a/drivers/gpio/gpiolib.c
+++ b/drivers/gpio/gpiolib.c
@@ -236,6 +236,8 @@ static ssize_t gpio_value_show(struct device *dev,

 	if (!test_bit(FLAG_EXPORT, &desc->flags))
 		status = -EIO;
+	else if (!desc->chip->get)
+		status = -EINVAL;
 	else
 		status = sprintf(buf, "%d\n", gpio_get_value_cansleep(gpio));

-- 
1.5.6.3
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/