kernel/drivers/pinctrl/qcom/pinctrl-msm.c, branch linux-rolling-stable

pinctrl: qcom: msm: Fix deadlock in pinmux configuration

2025-11-12T12:41:34Z

Replace disable_irq() with disable_irq_nosync() in msm_pinmux_set_mux() to prevent deadlock when wakeup IRQ is triggered on the same GPIO being reconfigured. The issue occurs when a wakeup IRQ is triggered on a GPIO and the IRQ handler attempts to reconfigure the same GPIO's pinmux. In this scenario, msm_pinmux_set_mux() calls disable_irq() which waits for the currently running IRQ handler to complete, creating a circular dependency that results in deadlock. Using disable_irq_nosync() avoids waiting for the IRQ handler to complete, preventing the deadlock condition while still properly disabling the interrupt during pinmux reconfiguration. Suggested-by: Prasad Sodagudi Signed-off-by: Praveen Talari Reviewed-by: Bjorn Andersson Signed-off-by: Linus Walleij

pinctrl: generic: rename PIN_CONFIG_OUTPUT to LEVEL

2025-09-08T12:24:20Z

This generic pin config property is confusingly named so let's rename it to make things clearer. There are already drivers in the tree that use PIN_CONFIG_OUTPUT to *read* the value of an output driven pin, which is a big semantic confusion for the head: are we then reading the setting of the output or the actual value/level that is put out on the pin? We already have PIN_CONFIG_OUTPUT_ENABLE that turns on driver buffers for output, so this can by logical conclusion only drive the voltage level if it should be any different. But if we read the pin, are we then reading the *setting* of the output value or the *actual* value we can see on the line? If the pin has not first been set into output mode with PIN_CONFIG_OUTPUT_ENABLE, but is instead in some input mode or tristate, what will reading this property actually return? Reading the current users reading this property it is clear that what we read is the logical level of the pin as 0 or 1 depending on if it is low or high. Rename it to PIN_CONFIG_LEVEL so it is crystal clear that we set or read the voltage level of the pin and nothing else. Acked-by: Sudeep Holla Signed-off-by: Linus Walleij

pinctrl: qcom: make the pinmuxing strict

2025-09-08T12:23:24Z

The strict flag in struct pinmux_ops disallows the usage of the same pin as a GPIO and for another function. Without it, a rouge user-space process with enough privileges (or even a buggy driver) can request a used pin as GPIO and drive it, potentially confusing devices or even crashing the system. Set it globally for all pinctrl-msm users. Reviewed-by: Konrad Dybcio Tested-by: Neil Armstrong Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij

pinctrl: qcom: add infrastructure for marking pin functions as GPIOs

2025-09-08T12:23:07Z

Add a helper macro that wraps PINCTRL_GPIO_PINFUNCTION() for pinctrl-msm pin functions and assign the .function_is_gpio() callback in pinmux_ops. Reviewed-by: Konrad Dybcio Tested-by: Neil Armstrong Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij

pinctrl: qcom: use generic pin function helpers

2025-09-08T12:22:50Z

With the pinmux core no longer duplicating memory used to store the struct pinfunction objects in .rodata, we can now use the existing infrastructure for storing and looking up pin functions in qualcomm drivers. Remove hand-crafted callbacks. Reviewed-by: Konrad Dybcio Tested-by: Neil Armstrong Signed-off-by: Bartosz Golaszewski Signed-off-by: Linus Walleij

treewide: rename GPIO set callbacks back to their original names

2025-08-07T08:07:06Z

The conversion of all GPIO drivers to using the .set_rv() and .set_multiple_rv() callbacks from struct gpio_chip (which - unlike their predecessors - return an integer and allow the controller drivers to indicate failures to users) is now complete and the legacy ones have been removed. Rename the new callbacks back to their original names in one sweeping change. Signed-off-by: Bartosz Golaszewski

pinctrl: qcom: msm: mark certain pins as invalid for interrupts

2025-06-18T11:56:34Z

On some platforms, the UFS-reset pin has no interrupt logic in TLMM but is nevertheless registered as a GPIO in the kernel. This enables the user-space to trigger a BUG() in the pinctrl-msm driver by running, for example: `gpiomon -c 0 113` on RB2. The exact culprit is requesting pins whose intr_detection_width setting is not 1 or 2 for interrupts. This hits a BUG() in msm_gpio_irq_set_type(). Potentially crashing the kernel due to an invalid request from user-space is not optimal, so let's go through the pins and mark those that would fail the check as invalid for the irq chip as we should not even register them as available irqs. This function can be extended if we determine that there are more corner-cases like this. Fixes: f365be092572 ("pinctrl: Add Qualcomm TLMM driver") Cc: stable@vger.kernel.org Reviewed-by: Bjorn Andersson Signed-off-by: Bartosz Golaszewski Link: https://lore.kernel.org/20250612091448.41546-1-brgl@bgdev.pl Signed-off-by: Linus Walleij

pinctrl: qcom: switch to devm_gpiochip_add_data()

2025-06-09T08:16:45Z

In order to simplify cleanup actions, use devres-enabled version of gpiochip_add_data(). As the msm_pinctrl_remove() function is now empty, drop it and all its calls from the corresponding pinctrl drivers. Signed-off-by: Dmitry Baryshkov Link: https://lore.kernel.org/20250513-pinctrl-msm-fix-v2-3-249999af0fc1@oss.qualcomm.com Signed-off-by: Linus Walleij

Merge tag 'pinctrl-v6.16-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl

2025-05-30T16:46:28Z

Pull pin control updates from Linus Walleij: "An especially linear and sparse improvement and new drivers release. Nothing exciting. The biggest change in Bartosz changes to make gpiochip set/get calls return error codes (something we should have fixed ages ago but is now finally getting fixed.) Core changes: - Add the devres devm_pinctrl_register_mappings() call that can register some pin control machine mappings and have them go away with the associated device New drivers: - Support for the Mediatek MT6893 and MT8196 SoCs - Support for the Renesas RZ/V2N SoC - Support for the NXP Freescale i.MX943 SoC Improvements: - Per-SoC suspend/resume callbacks in the Samsung drivers - Set all pins as input (High-Z) at probe in the MCP23S08 driver - Switch most GPIO chips to use the setters/getters with a return value - EGPIO support in the Qualcomm QCM2290 driver - Fix up the number of available GPIO lines in Qualcomm QCS8300 and QCS615" * tag 'pinctrl-v6.16-1' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl: (82 commits) pinctrl: freescale: Add support for imx943 pinctrl pinctrl: core: add devm_pinctrl_register_mappings() pinctrl: remove extern specifier for functions in machine.h pinctrl: mediatek: eint: Fix invalid pointer dereference for v1 platforms pinctrl: freescale: Enable driver if platform is enabled. pinctrl: freescale: Depend imx-scu driver on OF pinctrl: armada-37xx: propagate error from armada_37xx_pmx_set_by_name() pinctrl: armada-37xx: propagate error from armada_37xx_gpio_get_direction() pinctrl: armada-37xx: propagate error from armada_37xx_pmx_gpio_set_direction() pinctrl: armada-37xx: propagate error from armada_37xx_gpio_get() pinctrl: armada-37xx: propagate error from armada_37xx_gpio_direction_output() pinctrl: armada-37xx: set GPIO output value before setting direction pinctrl: armada-37xx: use correct OUTPUT_VAL register for GPIOs > 31 pinctrl: meson: Drop unused aml_pctl_find_group_by_name() pinctrl: at91: Fix possible out-of-boundary access pinctrl: add stubs for OF-specific pinconf functions pinctrl: qcom: correct the ngpios entry for QCS8300 pinctrl: qcom: correct the ngpios entry for QCS615 dt-bindings: pinctrl: qcom: correct gpio-ranges in examples for qcs8300 dt-bindings: pinctrl: qcom: correct gpio-ranges in examples for qcs615 ...

pinctrl: qcom: switch to devm_register_sys_off_handler()

2025-05-20T21:41:29Z

Error-handling paths in msm_pinctrl_probe() don't call a function required to unroll restart handler registration, unregister_restart_handler(). Instead of adding calls to this function, switch the msm pinctrl code into using devm_register_sys_off_handler(). Fixes: cf1fc1876289 ("pinctrl: qcom: use restart_notifier mechanism for ps_hold") Signed-off-by: Dmitry Baryshkov Link: https://lore.kernel.org/20250513-pinctrl-msm-fix-v2-2-249999af0fc1@oss.qualcomm.com Signed-off-by: Linus Walleij