summaryrefslogtreecommitdiff
path: root/drivers/pinctrl
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2016-07-26 19:16:01 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2016-07-26 19:16:01 -0700
commit1cd04d293c818687795b83cd8f2626bd4662feeb (patch)
treedcbaadd82c02204114b99c418bfae1ee57b2c4ca /drivers/pinctrl
parent9c1958fc326a0a0a533ec8e86ea6fa30977207de (diff)
parent224f9e6d538c4cfb2fa8dc4206fceb9431271388 (diff)
Merge tag 'gpio-v4.8-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio
Pull GPIO updates from Linus Walleij: "This is the bulk of GPIO changes for the v4.8 kernel cycle. The big news is the completion of the chardev ABI which I'm very happy about and apart from that it's an ordinary, quite busy cycle. The details are below. The patches are tested in linux-next for some time, patches to other subsystem mostly have ACKs. I got overly ambitious with configureing lines as input for IRQ lines but it turns out that some controllers have their interrupt-enable and input-enabling in orthogonal settings so the assumption that all IRQ lines are input lines does not hold. Oh well, revert and back to the drawing board with that. Core changes: - The big item is of course the completion of the character device ABI. It has now replaced and surpassed the former unmaintainable sysfs ABI: we can now hammer (bitbang) individual lines or sets of lines and read individual lines or sets of lines from userspace, and we can also register to listen to GPIO events from userspace. As a tie-in we have two new tools in tools/gpio: gpio-hammer and gpio-event-mon that illustrate the proper use of the new ABI. As someone said: the wild west days of GPIO are now over. - Continued to remove the pointless ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB Kconfig symbols. I'm patching hexagon, openrisc, powerpc, sh, unicore, ia64 and microblaze. These are either ACKed by their maintainers or patched anyways after a grace period and no response from maintainers. Some archs (ARM) come in from their trees, and others (x86) are still not fixed, so I might send a second pull request to root it out later in this merge window, or just defer to v4.9. - The GPIO tools are moved to the tools build system. New drivers: - New driver for the MAX77620/MAX20024. - New driver for the Intel Merrifield. - Enabled PCA953x for the TI PCA9536. - Enabled PCA953x for the Intel Edison. - Enabled R8A7792 in the RCAR driver. Driver improvements: - The STMPE and F7188x now supports the .get_direction() callback. - The Xilinx driver supports setting multiple lines at once. - ACPI support for the Vulcan GPIO controller. - The MMIO GPIO driver supports device tree probing. - The Acer One 10 is supported through the _DEP ACPI attribute. Cleanups: - A major cleanup of the OF/DT support code. It is way easier to read and understand now, probably this improves performance too. - Drop a few redundant .owner assignments. - Remove CLPS711x boardfile support: we are 100% DT" * tag 'gpio-v4.8-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-gpio: (67 commits) MAINTAINERS: Add INTEL MERRIFIELD GPIO entry gpio: dwapb: add missing fwnode_handle_put() in dwapb_gpio_get_pdata() gpio: merrifield: Protect irq_ack() and gpio_set() by lock gpio: merrifield: Introduce GPIO driver to support Merrifield gpio: intel-mid: Make it depend to X86_INTEL_MID gpio: intel-mid: Sort header block alphabetically gpio: intel-mid: Remove potentially harmful code gpio: rcar: add R8A7792 support gpiolib: remove duplicated include from gpiolib.c Revert "gpio: convince line to become input in irq helper" gpiolib: of_find_gpio(): Don't discard errors gpio: of: Allow overriding the device node gpio: free handles in fringe cases gpio: tps65218: Add platform_device_id table gpio: max77620: get gpio value based on direction gpio: lynxpoint: avoid potential warning on error path tools/gpio: add install section tools/gpio: move to tools buildsystem gpio: intel-mid: switch to devm_gpiochip_add_data() gpio: 74x164: Use spi_write() helper instead of open coding ...
Diffstat (limited to 'drivers/pinctrl')
-rw-r--r--drivers/pinctrl/pinctrl-xway.c5
1 files changed, 2 insertions, 3 deletions
diff --git a/drivers/pinctrl/pinctrl-xway.c b/drivers/pinctrl/pinctrl-xway.c
index a13f2b6f6fc0..b9375544dff0 100644
--- a/drivers/pinctrl/pinctrl-xway.c
+++ b/drivers/pinctrl/pinctrl-xway.c
@@ -1724,9 +1724,9 @@ static int pinmux_xway_probe(struct platform_device *pdev)
}
xway_pctrl_desc.pins = xway_info.pads;
- /* load the gpio chip */
+ /* register the gpio chip */
xway_chip.parent = &pdev->dev;
- ret = gpiochip_add(&xway_chip);
+ ret = devm_gpiochip_add_data(&pdev->dev, &xway_chip, NULL);
if (ret) {
dev_err(&pdev->dev, "Failed to register gpio chip\n");
return ret;
@@ -1749,7 +1749,6 @@ static int pinmux_xway_probe(struct platform_device *pdev)
/* register with the generic lantiq layer */
ret = ltq_pinctrl_register(pdev, &xway_info);
if (ret) {
- gpiochip_remove(&xway_chip);
dev_err(&pdev->dev, "Failed to register pinctrl driver\n");
return ret;
}