linux/drivers/misc/eeprom, branch v4.9

Merge branch 'i2c/for-4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux

2016-10-07T21:12:21Z

Pull i2c updates from Wolfram Sang: "Here is the 4.9 pull request from I2C including: - centralized error messages when registering to the core - improved lockdep annotations to prevent false positives - DT support for muxes, gates, and arbitrators - bus speeds can now be obtained from ACPI - i2c-octeon got refactored and now supports ThunderX SoCs, too - i2c-tegra and i2c-designware got a bigger bunch of updates - a couple of standard driver fixes and improvements" * 'i2c/for-4.9' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux: (71 commits) i2c: axxia: disable clks in case of failure in probe i2c: octeon: thunderx: Limit register access retries i2c: uniphier-f: fix misdetection of incomplete STOP condition gpio: pca953x: variable 'id' was used twice i2c: i801: Add support for Kaby Lake PCH-H gpio: pca953x: fix an incorrect lockdep warning i2c: add a warning to i2c_adapter_depth() lockdep: make MAX_LOCKDEP_SUBCLASSES unconditionally visible i2c: export i2c_adapter_depth() i2c: rk3x: Fix variable 'min_total_ns' unused warning i2c: rk3x: Fix sparse warning i2c / ACPI: Do not touch an I2C device if it belongs to another adapter i2c: octeon: Fix high-level controller status check i2c: octeon: Avoid sending STOP during recovery i2c: octeon: Fix set SCL recovery function i2c: rcar: add support for r8a7796 (R-Car M3-W) i2c: imx: make bus recovery through pinctrl optional i2c: meson: add gxbb compatible string i2c: uniphier-f: set the adapter to master mode when probing i2c: uniphier-f: avoid WARN_ON() of clk_disable() in failure path ...

at25: fix debug and error messaging

2016-09-27T10:43:35Z

The patch does the following: - fixes specifiers and removes explicit casting of the parameters - joins literals to one line - increases readability of the parameters Signed-off-by: Andy Shevchenko Signed-off-by: Greg Kroah-Hartman

eeprom: at24: check if the chip is functional in probe()

2016-08-22T06:19:58Z

The at24 driver doesn't check if the chip is functional in its probe function. This leads to instantiating devices that are not physically present. For example the cape EEPROMs for BeagleBone Black are defined in the device tree at four addresses on i2c2, but normally only one of them is present. If the userspace doesn't know the location in advance, it will need to check if reading the nvmem attributes fails to determine which EEPROM is actually there. Try to read a single byte in probe() and bail-out with -ENODEV if the read fails. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: tweak the loop_until_timeout() macro

2016-07-18T18:07:54Z

loop_until_timeout() replaced a do {} while loop in the at24 driver with a for loop which, under certain circumstances (such as heavy load or low value of the write_timeout argument), can lead to the code in the loop never being executed. Make sure that at least one iteration of the code enclosed within loop_until_timeout() is always executed. Suggested-by: Wolfram Sang Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: add support for at24mac series

2016-07-17T17:58:00Z

Add a new read function to the at24 driver allowing to retrieve the factory-programmed mac address embedded in chips from the at24mac family. These chips can be instantiated similarily to the at24cs family, except that there's no way of having access to both the serial number and the mac address at the same time - the user must instantiate either an at24cs or at24mac device as both special memory areas are accessible on the same slave address. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: support reading the serial number for 24csxx

2016-07-17T17:57:30Z

The chips from the at24cs family have two memory areas - a regular read-write block and a read-only area containing the serial number. The latter is visible on a different slave address (the address of the rw memory block + 0x08). In order to access both blocks the user needs to instantiate a regular at24c device for the rw block address and a corresponding at24cs device on the serial number block address. Add a function that allows to access the serial number and assign it to at24->read_func if the chip allows serial number read operations and the driver was passed the relevant flag for this device. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: split at24_eeprom_write() into specialized functions

2016-07-17T17:41:54Z

Split at24_eeprom_write() into three smaller functions - one for the i2c operations and two for the smbus extensions (separate routines for block and byte transfers). Assign them in at24_probe() depending on the bus capabilities. Also: in order to avoid duplications move code adjusting the count argument into a separate function and use it for i2c and smbus block writes (no need for a roll-over for byte writes as we're always writing one byte). Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: split at24_eeprom_read() into specialized functions

2016-07-17T17:41:54Z

Split at24_eeprom_read() into two smaller functions - one for the i2c operations and one for the smbus extensions. Assign them in at24_probe() depending on the bus capabilities. Also: in order to avoid duplications move the comments related to offset calculations above the at24_translate_offset() routine. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: hide the read/write loop behind a macro

2016-07-17T17:41:54Z

Before splitting the read/write routines into smaller, more specialized functions, unduplicate some code in advance. Use a 'for' loop instead of 'do while' when waiting for the previous write to complete and hide it behind a macro. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang

eeprom: at24: call read/write functions via function pointers

2016-07-17T17:41:54Z

The first step in simplifying the read and write functions is to call them via function pointers stored in at24_data. When we eventually split the routines into smaller ones (depending on whether they use smbus or i2c operations) we'll simply assign them to said pointers instead of checking the flags at runtime every time we read/write. Signed-off-by: Bartosz Golaszewski Signed-off-by: Wolfram Sang