linux-toradex.git/drivers/gpio/gpio-pxa.c, branch v4.9.131

gpio: pxa: Fix potential NULL dereference

2018-09-26T06:36:39+00:00

[ Upstream commit 9506755633d0b32ef76f67c345000178e9b0dfc4 ]

platform_get_resource() may fail and return NULL, so we should
better check it's return value to avoid a NULL pointer dereference
a bit later in the code.

This is detected by Coccinelle semantic patch.

@@
expression pdev, res, n, t, e, e1, e2;
@@

res = platform_get_resource(pdev, t, n);
+ if (!res)
+   return -EINVAL;
... when != res == NULL
e = devm_ioremap(e1, res->start, e2);

Signed-off-by: Wei Yongjun 
Acked-by: Robert Jarzmik 
Signed-off-by: Linus Walleij 
Signed-off-by: Sasha Levin 
Signed-off-by: Greg Kroah-Hartman

gpio: pxa: fix legacy non pinctrl aware builds

2016-04-08T08:10:08+00:00

In legacy pxa builds, ie. non device-tree and platform-data only builds,
pinctrl is not yet available. As a consequence, the pinctrl gpio
direction change function is a stub, returning always success.

In the current state, the gpio driver direction function believes the
pinctrl direction change was successful, and exits without actually
changing the gpio direction.

This patch changes the logic :
 - if the pinctrl direction function fails, gpio direction will report
   that failure
 - if the pinctrl direction function succeeds, gpio direction is changed
   by the gpio driver anyway.
   This is sub optimal in the pinctrl aware case, as the gpio direction
   will be changed twice: once by pinctrl function and another time by
   the gpio direction function.

Yet it should be acceptable in this form, as this is functional for all
pxa platforms (device-tree and platform-data), and moreover changing a
gpio direction is very very seldom, usually in machine initialization,
seldom in drivers probe, and an exception for ac97 reset bug.

Fixes: a770d946371e ("gpio: pxa: add pin control gpio direction and request")
Reported-by: Guenter Roeck 
Tested-by: Guenter Roeck 
Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: use gpiochip data pointer

2016-01-05T10:21:12+00:00

This makes the driver use the data pointer added to the gpio_chip
to store a pointer to the state container instead of relying on
container_of().

Cc: Daniel Mack 
Acked-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: checking IS_ERR() instead of NULL

2016-01-05T10:19:26+00:00

irq_domain_add_legacy() returns NULL on error, it doesn't return error
pointers.

Fixes: 384ca3c6a28d ('gpio: pxa: change the interrupt management')
Signed-off-by: Dan Carpenter 
Signed-off-by: Linus Walleij

gpio: pxa: fixes non devicetree builds

2015-12-22T09:09:57+00:00

The commit "gpio: pxa: change the interrupt management" should have
taken care of moving an ifdef to not englobe irqdomain related
structures anymore, as they are used now for all builds.

This repairs the broken builds where CONFIG_OF=n.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: add pin control gpio direction and request

2015-12-15T13:16:47+00:00

If a pin control driver is available, use it to change the gpio
direction. If not fallback to directly manipulating the gpio direction
register.

The reason to use the pin control driver first is that pin control in
pxa2xx architecture implies changing the gpio direction, even for non
gpio functions. In order to do it atomically, only one driver should
control the gpio direction, and if a pin controller is available, it has
to be him.

There is a small catch : if CONFIG_PINCTRL is selected, then a pinctrl
driver has to be probed. If not, gpio_request() will return
-EPROBE_DEFER as pinctrl_request_gpio() returns it in that case.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: change the interrupt management

2015-12-10T15:05:44+00:00

The interrupt management is changed by this patch to rely on chip data
instead of chained interrupts.

The main goal is to loosen the dependency on the global pxa chip
structure in favor of the passed chip data. The secondary goal is to
better show in /proc/interrupts the difference between interrupts for
gpio0 and gpio1 (directly wired to interrupt controller), and the other
gpios (wired onto a third line in the interrupt controller).

The last advantage of this patch is that the interrupt is actually
requested, so that another driver cannot steal this line, or overwrite
the handler.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: convert to devm_ioremap

2015-12-10T15:04:40+00:00

Use the device managed ioremap to simplify the probe function.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: convert to one gpiochip

2015-12-10T15:02:54+00:00

The pxa gpio IP is provided by one chip, which holds multiple banks.

Another reason the driver should register only one gpiochip instead of
multiple gpiochips (ie. 1 per each bank) is that for pincontrol and
devicetree integration (think gpio-ranges), it's impossible to have the
contiguous pin range 0..127 mapped to gpios 0..127.

This patch, amongst other thinks, paves the path to loosen the bond with
the global structure variable pxa_gpio_chip.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij

gpio: pxa: change initcall level second attempt

2015-11-19T08:24:38+00:00

This patch is a second attempt at what was previously in commit
6c7e660a27da ("gpio: pxa: set initcall level to module init").

The goal is the same : enable gpio & pinctrl driver to work
together. As pinctrl driver will be initialized at device level, the
gpio should be as well, so that the deferring mechanism is honored.

Yet this patch should also respect the legacy platforms, so the set of
constraints is :
 - in legacy platforms (ie. non dt), gpio_[gs]et_*() should be available
   for machine code => core initcall
 - in new platforms (ie. dt based), pinctrl will be available and no
   machine code => device initcall

In order to fullfill all these constraints, the initcall level is either
postcore for non devicetree platforms, and device for devicetree platforms.

Signed-off-by: Robert Jarzmik 
Signed-off-by: Linus Walleij