talos-op-linux/drivers/pinctrl/pinctrl-single.c, branch v3.14

pinctrl: single: fix infinite loop caused by bad mask

2014-01-15T07:31:53+00:00

commit 4e7e8017a80e1 (pinctrl: pinctrl-single: enhance to configure multiple pins of different modules) improved support for pinctrl-single,bits option, but also caused a regression in parsing badly configured mask data. If the masks in DT data are not quite right, pcs_parse_bits_in_pinctrl_entry() can end up in an infinite loop, trashing memory at the same time. Add a check to verify that each loop actually removes bits from the 'mask', so that the loop can eventually end. Signed-off-by: Tomi Valkeinen Acked-by: Tony Lindgren Signed-off-by: Linus Walleij

pinctrl: single: fix pcs_disable with bits_per_mux

2014-01-15T07:30:08+00:00

pcs_enable() uses vals->mask instead of pcs->fmask when bits_per_mux is enabled. However, pcs_disable() always uses pcs->fmask. Fix pcs_disable() to use vals->mask with bits_per_mux. Signed-off-by: Tomi Valkeinen Acked-by: Peter Ujfalusi Acked-by: Tony Lindgren Signed-off-by: Linus Walleij

pinctrl: single: call pcs_soc->rearm() whenever IRQ mask is changed

2013-11-14T18:43:17+00:00

On OMAPs the IO ring must be rearmed each time the pad wakeup configuration is changed. So call pcs_soc->rearm() from pcs_irq_set(). As pinctrl-single is now an interrupt controller in some cases, we should follow the standards and keep the interrupts enabled constantly, and not just for wake-up events. The tracking of runtime vs wake-up interrupts can be handled separately for the automated runtime PM solution when we have it in the future. Signed-off-by: Roger Quadros Acked-by: Linus Walleij [tony@atomide.com: removed wrong comment, updated description] Signed-off-by: Tony Lindgren

pinctrl: single: Fix build when not built on ARM

2013-10-18T23:43:06+00:00

Looks like we need a little bit of arch specific handling with the generic IRQ. Fix the issue with an ifdef the same way as other drivers do. ARM needs things set to IRQF_VALID, which also then sets noprobe. Others seem to use just irq_set_noprobe(). Otherwise we can get: drivers/pinctrl/pinctrl-single.c: In function 'pcs_irqdomain_map': drivers/pinctrl/pinctrl-single.c:1750:2: error: implicit declaration of function 'set_irq_flags' [-Werror=implicit-function-declaration] drivers/pinctrl/pinctrl-single.c:1750:21: error: 'IRQF_VALID' undeclared (first use in this function) drivers/pinctrl/pinctrl-single.c:1750:34: error: 'IRQF_PROBE' undeclared (first use in this function) Reported-by: Randy Dunlap Signed-off-by: Tony Lindgren

pinctrl: single: Add support for auxdata

2013-10-10T22:30:47+00:00

For omaps, we still have dependencies to the legacy code for handling the PRM (Power Reset Management) interrupts, and also for reconfiguring the io wake-up chain after changes. Let's pass the PRM interrupt and the rearm functions via auxdata. Then when at some point we have a proper PRM driver, we can get the interrupt via device tree and set up the rearm function as exported function in the PRM driver. By using auxdata we can remove a dependency to the wake-up events for converting omap3 to be device tree only. Cc: Peter Ujfalusi Cc: Grygorii Strashko Cc: Prakash Manjunathappa Cc: Roger Quadros Cc: Haojian Zhuang Cc: linux-kernel@vger.kernel.org Reviewed-by: Kevin Hilman Tested-by: Kevin Hilman Acked-by: Linus Walleij Signed-off-by: Tony Lindgren

pinctrl: single: Add support for wake-up interrupts

2013-10-10T22:30:30+00:00

The pin control registers can have interrupts for example for device wake-up. These interrupts can be treated as a chained interrupt controller as suggested earlier by Linus Walleij . This patch adds support for interrupts in a way that should be pretty generic, and works for the omaps that support wake-up interrupts. On omaps, there's an interrupt enable and interrupt status bit for each pin. The two pinctrl domains on omaps share a single interrupt from the PRM chained interrupt handler. Support for other similar hardware should be easy to add. Note that this patch does not attempt to handle the wake-up interrupts automatically unlike the earlier patches. This patch allows the device drivers to do a request_irq() on the wake-up pins as needed. I'll try to do also a separate generic patch for handling the wake-up events automatically. Also note that as this patch makes the pinctrl-single an irq controller, the current bindings need some extra trickery to use interrupts from two different interrupt controllers for the same driver. So it might be worth waiting a little on the patches enabling the wake-up interrupts from drivers as there should be a generic way to handle it coming. And also there's been discussion of interrupts-extended binding for using interrupts from multiple interrupt controllers. In any case, this patch should be ready to go allowing handling the wake-up interrupts in a generic way, or separately from the device drivers. Cc: Peter Ujfalusi Cc: Grygorii Strashko Cc: Prakash Manjunathappa Cc: Roger Quadros Cc: linux-kernel@vger.kernel.org Cc: Benoît Cousson Cc: devicetree@vger.kernel.org Acked-by: Linus Walleij Acked-by: Haojian Zhuang Reviewed-by: Kevin Hilman Tested-by: Kevin Hilman Signed-off-by: Tony Lindgren

pinctrl: single: Prepare for supporting SoC specific features

2013-10-09T21:47:54+00:00

Let's replace is_pinconf with flags and add struct pcs_soc_data so we can support SoC specific features like pin wake-up events. Done in collaboration with Roger Quadros . Cc: Peter Ujfalusi Cc: Grygorii Strashko Cc: Prakash Manjunathappa Cc: linux-kernel@vger.kernel.org Acked-by: Linus Walleij Acked-by: Haojian Zhuang Reviewed-by: Kevin Hilman Tested-by: Kevin Hilman Signed-off-by: Roger Quadros Signed-off-by: Tony Lindgren

Merge tag 'v3.11-rc7' into devel

2013-08-29T07:46:30+00:00

Merged in this to avoid conflicts with the big locking fixes from upstream. Signed-off-by: Linus Walleij Conflicts: drivers/pinctrl/pinctrl-sunxi.c

pinctrl: Pass all configs to driver on pin_config_set()

2013-08-28T11:34:41+00:00

When setting pin configuration in the pinctrl framework, pin_config_set() or pin_config_group_set() is called in a loop to set one configuration at a time for the specified pin or group. This patch 1) removes the loop and 2) changes the API to pass the whole pin config array to the driver. It is now up to the driver to loop through the configs. This allows the driver to potentially combine configs and reduce the number of writes to pin config registers. All c files changed have been build-tested to verify the change compiles and that the corresponding .o is successfully generated. Signed-off-by: Sherman Yin Reviewed-by: Christian Daudt Reviewed-by: Matt Porter Tested-by: Stephen Warren Acked-by: Laurent Pinchart Signed-off-by: Linus Walleij

pinctrl: pinctrl-single: fix compile warning when no CONFIG_PM

2013-07-22T21:00:55+00:00

This warning has been introduced by the commit 0f9bc4bcdf4f pinctrl: single: adopt pinctrl sleep mode management Signed-off-by: Jean-Francois Moine Acked-by: Tony Lindgren Signed-off-by: Linus Walleij