linux/drivers/i2c/i2c-dev.c, branch v4.7

i2c: dev: use after free in detach

2016-05-28T15:37:42Z

The call to put_i2c_dev() frees "i2c_dev" so there is a use after free when we call cdev_del(&i2c_dev->cdev). Fixes: d6760b14d4a1 ('i2c: dev: switch from register_chrdev to cdev API') Signed-off-by: Dan Carpenter Signed-off-by: Wolfram Sang

i2c: dev: don't start function name with 'return'

2016-05-27T11:13:01Z

I stumbled multiple times over 'return_i2c_dev', especially before the actual 'return res'. It makes the code hard to read, so reanme the function to 'put_i2c_dev' which also better matches 'get_free_i2c_dev'. Signed-off-by: Wolfram Sang

i2c: dev: switch from register_chrdev to cdev API

2016-05-26T19:18:57Z

i2c-dev had never moved away from the older register_chrdev interface to implement its char device registration. The register_chrdev API has the limitation of enabling only up to 256 i2c-dev busses to exist. Large platforms with lots of i2c devices (i.e. pluggable transceivers) with dedicated busses may have to exceed that limit. In particular, there are also platforms making use of the i2c bus multiplexing API, which instantiates a virtual bus for each possible multiplexed selection. This patch removes the register_chrdev usage and replaces it with the less old cdev API, which takes away the 256 i2c-dev bus limitation. It should not have any other impact for i2c bus drivers or user space. This patch has been tested on qemu x86 and qemu powerpc platforms with the aid of a module which adds and removes 5000 virtual i2c busses, as well as validated on an existing powerpc hardware platform which makes use of the i2c bus multiplexing API. i2c-dev busses with device minor numbers larger than 256 have also been validated to work with the existing i2c-tools. Signed-off-by: Erico Nunes [wsa: kept includes sorted] Signed-off-by: Wolfram Sang

i2c: i2c-dev: sort includes

2016-02-20T22:33:38Z

I request this for drivers, so the core should adhere to sorted includes as well. Signed-off-by: Wolfram Sang

i2c-dev: Fix I2C_SLAVE ioctl comment

2015-10-23T21:26:44Z

The first part of the comment is wrong since November 2007, delete it. The second part of the comment is related to I2C_PEC, not I2C_SLAVE, so move it where it belongs. Signed-off-by: Jean Delvare Cc: Wolfram Sang Signed-off-by: Wolfram Sang

i2c-dev: Fix typo in ioctl name reference

2015-10-23T21:26:43Z

The ioctl is named I2C_RDWR for "I2C read/write". But references to it were misspelled "rdrw". Fix them. Signed-off-by: Jean Delvare Signed-off-by: Wolfram Sang

i2c: remove FSF address

2014-11-07T17:35:33Z

We have a central copy of the GPL for that. Some addresses were already outdated. Signed-off-by: Wolfram Sang

i2c: i2c-dev: Create 'name' attribute automatically

2013-09-30T04:02:31Z

The 'name' attribute is needed for all i2c-dev class devices, meaning it can be created automatically by pointing to it in the class data structure. This simplifies the code and reduces the probability for race conditions (the name attribute should exist by the time the device is announced to user space). Signed-off-by: Guenter Roeck Signed-off-by: Wolfram Sang

new helper: file_inode(file)

2013-02-23T04:31:31Z

Signed-off-by: Al Viro

i2c-dev: Add support for I2C_M_RECV_LEN

2012-05-30T08:55:34Z

As the bus driver side implementation of I2C_M_RECV_LEN is heavily tied to SMBus, we can't support received length over 32 bytes, but let's at least support that. In practice, the caller will have to setup a buffer large enough to cover the case where received length byte has value 32, so minimum 32 + 1 = 33 bytes, possibly more if there is a fixed number of bytes added for the specific slave (for example a checksum.) Signed-off-by: Jean Delvare Tested-by: Douglas Gilbert